🔊┃audio

[tribal panther]

Never used Steam Audio before, but AFAIK you need to attach an Steam Audio Source to any object which produces a sound in other to use the Steam Audio systems

[hard gulch]

Your events should have a Steam Audio Spatializer though

[tribal panther]

How do I do that? Is there like a method to call?

[hard gulch]

Hmm not sure. It could be needed for reflections and occlusion and such

I have only used the HRTF/spatialization feature of steam audio

Which works with createinstance

[tribal panther]

(We want to use the Steam Audio so sounds take into account walls from rooms, as at the moment they ignore walls)

[hard gulch]

In FMOD, you need to replace the standard spatializer with SA spatializer

In the effects of the event you know

But theres a chance that you need steam audio source for occlusion

[hard gulch]

This is called occlusion btw

And possibly reflections

[tribal panther]

Oh, I'm not very familiar with the names of this effects, thanks

So for occlusion I can no longer use RuntimeManager.CreateInstance?

[hard gulch]

Not sure, you probably can, but you need to test/research yourself

[maiden quarry]

Hi all. I am using TTS for voicing on of my game characters, who is an AI gone wrong, so it fits nicely. But I still wanna add some filters to the audio to make it real ominous sounding. Any advice on what free software I can use to get this done? Thanks!

[hard gulch]

So they are premade audio clips, not generated on the fly?

You could either use the audio effects included in unity, or apply some effects with Audacity (free) or Reaper (not free but has unlimited trial)

[mellow finch]

heyup, anyone know if there's way to get FMOD audio to scrub correctly in Timeline? It seems to always play audio from the start as you scrub past events

[frigid harness]

can you guys check out this part of the main menu music im working on? any feedback is appriciated! :)

[tribal panther]

Using FMOD + Steam Audio, how can I play a sound by code? If I do it like I would do with just FMOD, it doesn't works as Steam Audio spatialization requires and Steam Audio Source component.

[orchid thicket]

Hey! I just ran into a bugg on 2022.3.32.
If you have multiple audiosources on a gameobject and call .stop() on one of them, it ends up muting all of the audiosources on that gameobject instead of only the targeted one

public AudioSource audio1;
public AudioSource audio2;

private void Start()
{
  audio2.Stop(); // this will also stop audio2 if they are on the same gameobject
}```

[delicate rivet]

double check audio1 and audio2 are not referencing the same AudioSource

[quasi hound]

Are you supposed to stop an AudioClip, or an AudioSource?

If you stop a source that's playing multiple clips, they are all expected to stop

"Stopping" a clip shouldn't really be a thing as far as I know

[orchid thicket]

Oh I just realised that my post was incorrect, it is suppose to be AudioSource and not Audioclip

[quasi hound]

I'd hazard a guess you're playing audio with PlayOneShot
Individual clips played that way cannot be stopped so you should use multiple AudioSources

[orchid thicket]

Iam using multiple AudiosSources, " audio1" and audio2" isnt audioclips but audiosources, I just named it wrong in my post

[quasi hound]

If so, there should be no reason that I know of why calling methd on one should be affecting the other
I'd triple check that you're calling the correct methods on correct classes and that the component references point to where you think they point to

[orchid thicket]

@quasi hound Ive already checken all of that, and it got solved when I changed one of the audiosources to a child object instead.

Like I said in my first post, this is a bug.
I was mostly wondering if others were aware of it.

[delicate rivet]

If you think its a bug you should report it https://support.unity.com/hc/en-us/articles/206336985-How-do-I-submit-a-bug-report. But from my checks in 22.3 you can call stop on one AudioSource and the other one keeps playing, they should not affect each other

[hard gulch]

Did you get it to work?
If not, show the settings of a "Steam audio spatializer" effect in your event in FMOD

Steam audio spatialized sounds work fine for me in Unity. I don't use SteamAudioSources, just regular FMODUnity.RuntimeManager.CreateInstance

[stuck arch]

Hi, I'm currently trying to set up playing multiple audio sources back to back seamlessly using dspTime and PlayScheduled. My current use case is:

• play an intro to a song
• loop the middle of a song
• detect when the audio source loops/changes and access the progress of the current source
  The main questions I have about this are with scheduling the start time. I understand the need to schedule in the future by adding to the current dspTime, but all the examples I can find online use an arbitrary float value (Example from unity: https://docs.unity3d.com/ScriptReference/AudioSource.PlayScheduled.html). My questions are as follows:
• Is there a better way to schedule audio than by adding a magic number to the current dspTime?
• If not, what's the typical range to play around in to play something as soon as possible without running into desync issues? Is it possible to completely prevent desync in high-lag situations?
• Should I be looking into something else entirely that better handles my use case?
  Thanks!

[kindred thicket]

What is your target build?
Unity currently uses 2 audio engines - Legacy FMOD for standalone apps and Web Audio API for WebGL build, so PlayScheduled behavior may differ.

[tribal panther]

I managed to kind of solve it by creating at runtime an gameobject with CustomStudioEventEmitter which inherits from StudioEventEmitter (and add some internal logic to set sounds) and SteamAudioSource instead of using FMODUnity.RuntimeManager.CreateInstance. But its quite annoying and complex.

My current intention is to make the Occlusion work, then I may add reflection and transmision.

[stuck arch]

Currently, I'm primarily aiming for a standalone app

[cosmic night]

Hey, is there a way to play audio at point in AudioSource component like if it was played by AudioSource.PlayClipAtPoint?

[quasi hound]

Not without moving it
Usually you'd use pooled audiosources
PlayClipAtPoint just creates a new one anyway

[spiral talon]

can i create audiorolloff presets or do i have to make new ones for every audiosource?

[kindred thicket]

I have some understanding about PlayScheduled in Unity 2019.4, just sharing my observations:

• Technically on internal FMOD side it's a delay (something like ChannelControl::setDelay), scheduled AudioSource will take 1 voice, so AudioSource.isPlaying will return true even if it doesn't make sound.
• AudioSource.PlayScheduled works best with Decompress On Load and worst with Streaming Load Type. Decompress On Load increases RAM usage, but since audioclip is stored uncompressed, you can instantly run AudioSource.PlayScheduled(dspTime) without latency. For Streaming type you need to add latency, i.e. 'magic number'.
• AudioSource.PlayScheduled has sufficient accuracy for sample-by-sample sync, but in my standalone tests there were rare cases of time shifting by 1024 samples = 21ms @ 48000Hz samplerate.
• AudioSource.PlayScheduled can break timing in Unity Audio Filters - Reverb Filter, Echo Filter, Chorus Filter.

Despite the issues, you can still create music audio system. I recommend '140' by Jakob Schmid, it is first game on Unity 3.5 that actively uses scheduled AudioSources.

[vapid coral]

Any recommendations for reliable royalty free SFX? I’m usually bouncing between Pixabay and Freesound but I’m struggling to find some specific fx

I figured this would be related enough to this channel so hopefully I’m not being annoying

[quasi hound]

Perhaps a little, but check out Sonniss GDC sample packs
Ultimately if you want something specific, or some accountability against stolen sounds, you'll have to buy some

[vapid coral]

So it says it’s got a bunch of SFX but everything is labeled per year and I have no clue what content they actually offer…

[quasi hound]

Each bundle has a tracklist link under its downloads

[vapid coral]

Ohhh

Yo it even says it has grunts and battle cries, those are what I’ve been working on 👍 thanks!!

[stuck arch]

Thanks so much for the pointers! The problem I was running into was in the load settings, I had Preload Audio Data enabled which I think caused the audio to not be done loading when I scheduled it. Disabling Preload Audio Data and loading it in the script fixed it!

[calm adder]

Has anyone solved the problem that when playing audio in a loop in a webgl build, there is about 1 frame of silence (gap) every time it loops?

[peak kiln]

does anyone know how to loop an event if it uses a parameter?

[feral geyser]

Hi all. I have a live audiovisual project and about to write a way to define bpm with midi pad, for it to then loop a broadcast event based on the rhythm

What i'm wondering about is the 1st part

Note: this is not for a beat press rhythm game

So i guess i'll use FixedUpdate, and determine if this frame has passed the supposedly next beat, and then update the "lastBeat" time by offsetting the 1/bpm value

I think FixedUpdate is good enough for this

Or even update, bcoz the resulting broadcast is just for visual trigger

[feral geyser]

Since this is for live, i'll need to calibrate the bpm somehow with just the midi pad press interval

[kindred thicket]

1. Do you want to create Tap Tempo like in Ableton?
2. How do you capture events from MIDI-device?
   Perhaps it is better to do this on MIDI-library/middleware side for greater time accuracy.
   DryWetMidi works well with Unity.

[feral geyser]

The midi input is sorted

Actually i'm already done implementing it

But since this is a custom audiovisual thing, i was asking others if they have experience building something like this, how they decide on the controls

Currently the midi pad is for manual trigger some fx
Ctrl pad and pressing it following a beat will start sampling the distances between beat, averages it, and sets it as the new beat length
It'll also add the pad to the list of pads to loop for each beat

Pressing the midi pad again without Ctrl, will trigger that fx then removes that pad from the registered list of pads that's being looped

[kindred thicket]

Sorry, it's too complicated for me

[feral geyser]

Haha sry it's fine. Just writing out loud maybe someone have something to say. Else it helps me imagine the design

[kindred thicket]

Doing audiovisual stuff is cool, it requires unusual approaches 🙂

[tribal panther]

Question about Audio Source. I know the point of Occlusion it to block sound from moving through solid meshes... but is possible to have a sound source inside a mesh and ignore that mesh (its owner) in particular?

[naive thicket]

@severe eagle There's no promotion in learning channels. If you have a Unity project this is attached to you may post it in #1180170818983051344 . Be sure to read post guidelines.

[severe eagle]

Noted sorry about that

[spice wren]

Can I have 1 audio source but 5 different audio positions?
I am try to make the 5 speakers in my scene all play the same audio clip and the position of the audio matters because I want spatial audio playing from each speaker.

I feel like duplicated 5 audio sources is bad for perf and writing a script sync them would be hacky

[kindred thicket]

How many channels are in AudioClip? Is it 5.1 or just mono?
I think there are several ways:

• AudioSource.PlayScheduled: sync playing of all AudioSources
• OnAudioFilterRead(): copying AudioSource audio buffer to others
• AudioClip.PCMReaderCallback: create streamed AudioClip for all AudioSources
• Unity Native Audio Plugin SDK: see Speaker Routing Demo
  In any case, you will have to create 5 sources.

[hollow osprey]

No

[hollow osprey]

No its not you could have 10000 audio sources and it would barly affect performance

Ok thats maybe a bit much audio sources, but you get what i mean

[kindred thicket]

Unity's AudioSource component provides some overhead to CPU, which controls audio backend parameters, calculates attenuation, doppler effect, audilbity, etc.

A long time ago I tried to play all 256 real and 4096 virtual voices and got significant performance loss on empty scene.

If you really care about audio system performance on standalone platforms, you should use audio middleware like FMOD Studio or Wwise, there is much less of such overhead there.

[hollow osprey]

True, but technically it like wont matter at 1 real and 1000 virtual devices

[kindred thicket]

btw, Unity had an alternative called DOTS Audio, where it was really possible to play 10,000 audiosources, but project was abandoned.

[whole pewter]

DSPGraph; was is really abandoned

re # of AudioSources: original question was about 5 of them, which would be anywhere near to pose real problem i'd assume

[whole pewter]

those are android (i suppose) specific - you should post in more relevant android channels/outlets
mobile audio has its own.. specifics, which also in case of android depend more on your hw/phone

[raw ridge]

anyone know how I can add button SFX (like clicking sounds) to an audio mixer to manipulate them with an SFX slider/checkbox?

[bitter granite]

Im trying to add SteamAudio on Unity 6 but i have an infinite loop while importing the package
I have follow the documentation but that not work

[bitter granite]

Anyone know how to get the steam audio occlusion working ?
When i enable them, there is not difference.. I have followed the documentation but nothing happens

[pallid marten]

hai im trying to use Unity's timeline to do a quick intro scene, first time using it. I have an animation clip added that moves a GO from one spot to another, and Im trying to implement a sound effect to follow but I can't hear anything. The .wav does play a sound tho. Also the game is in VR, not sure if that changes anything although even when the game isnt playing and im in the editor, the sound doesnt play. I tried attaching the audio to an empty GO and attaching it to the (NONE, audio source) track of the screenshot but that also didnt work. Any tips ?

[high sluice]

i have 3 audios but my 3th audio doesn't go of

    private AudioSource audioBouncy;
    private AudioSource audioHit;
    private AudioSource audioExplosion;

    private void Start()
    {
        // Get all AudioSource components attached to this GameObject
        AudioSource[] audioSources = GetComponents<AudioSource>();
    
        if (audioSources.Length > 2)
        {
            audioBouncy = audioSources[0];
            audioHit = audioSources[1];
            audioExplosion = audioSources[2];
        }
        else
        {
            Debug.LogError("Not enough AudioSource components found on this GameObject.");
        }```

my debug log does go off

            audioExplosion.Play();
            Debug.Log("explosion played!");```

[past igloo]

#1225882878831759430 message what do anyone think of this main menu music?

[mossy tiger]

Anyone know how to use fmod

[sour quarry]

Does playoneshot bypass audio filters? I have an audio source referenced from another game object, and I have the filters under the same gameobject as the audio source, but I'm not getting any filters. Does anyone know how to perform playoneshot with filters? Or do I have to create an audio source for every sound effects?

[rain canyon]

Was told to ask here, hope it's okay to forward it

[quasi hound]

I guess it depends what you mean by "stop", but an audio source is no longer processed only if Max Real Voice limit is exceeded and the audio is inaudible
https://gamedevbeginner.com/unity-audio-optimisation-tips/#pause_audio_sources

[rain canyon]

So is there tangible benefit to manually disabling out-of-range audio sources? I know that website says yes but as a beginner it's hard for me to evaluate the accuracy of that. For instance, I've been told distance calculations are extremely expensive, which is what the proposed solution on there is using.

[quasi hound]

As with all optimizations, you can't exactly know until you test

Is the number of out of range audio sources a tangible drain on resources to begin with?

Often if your game levels are sprawling, you'd be loading in the objects gradually by area anyway which makes this less of an issue in particular

[rain canyon]

Potentially - with my current approach every one of 50-100 units would have an object with an audio source attached. What that means in practice is a bit beyond me.

But like you said those units would come over time.

[quasi hound]

The Real Voice Limit already guarantees a certain threshold for audio culling

Your tests should reveal if that's enough or not

[rain canyon]

I'm also curious if Audio Sources "do" anything if they're not playing sound, do those take up resources just for being in the scene?

[quasi hound]

They take up memory, but little to no processing power as far as I know

[rain canyon]

I guess I'll have to sit down one on one with the profiler again properly and figure out how to properly use it.

Thank you for your help it's been really useful, I think I understand how it works now a bit better.

[quasi hound]

Also, it's bit of an old adage by now to say that distance checks are awful, from a time with much weaker hardware
Still, those can be optimized
Distance comparisons (using squared magnitude) are much cheaper than distance checks

Comparing to a squared distance tells you just as well if you're closer or further

[rain canyon]

Thank you, I'll look into those as I haven't used those before

[silver coral]

Where do you guys source sound for your games ive been looking for weapon sounds but they are all either not what I need or they are asking for some stupid subscription fee for something they 100% stole from somewhere else anyway

Ive got a gunshot sound I recorded myself but I dont have access to an AK atm and wont for a while

[spiral trail]

Humble Bundle almost always has something on. For example, https://www.humblebundle.com/software/gamedev-market-best-2024-software has a bunch of sfx and music packs in it.

[silver coral]

Thanks mate ill take a look

[inland raven]

Don’t tell anyone but https://sonniss.com/gameaudiogdc

[cosmic night]

Is there a way to play audio via AudioSource in stereo like it's in AudioSource.PlayClipAtPoint?

Like you hear the sound from the point it's been played

[quasi hound]

You place the gameobject with that audiosource there when you want it played, and ensure it's 3D spatial

PlayClipAtPoint doesn't do anything fancier than to create a gameobject with an audiosource, play it, then destroy it

[foggy karma]

has anyone here used Steam Audio (https://valvesoftware.github.io/steam-audio/) and know if it works for Consoles?

[inland raven]

Steam Audio supports Windows (32 bit and 64 bit), Linux (32 bit and 64 bit), macOS, Android (armv7, arm64, x86, x64), and iOS platforms.

[foggy karma]

rip thanks

[rain canyon]

im audio pooling in unity 2022.3.26f1 and i was wondering, how does audio pooling work with random pitch looping for things like footsteps? i know in 2023 and up there is an audio random container and that prob helps but im scared footsteps themselves will take up basically all 32 audio sources

from what I heard in unity 5 there are also custom limits for audio sources, but for this project updating is not an option (its a group uni project, so the version is the same as on the in-class computers)

[quasi hound]

If it's your own pooling system, it'll be responsible for setting/resetting pitch and other properties as needed

[swift crown]

Hello. Something weird happened to my working setup. After I made a clean install of Xubuntu 24.04 and installed Unity Hub with Unity Engine and opened my working project I got errors from FMod plugin basically claiming it couldn't initialize audio (ERR_OUTPUT_INIT Error initializing audio device). Tried refreshing banks, rebuilding fmod project, reinstalling (with purge) alsa and pulseaudio packages, tried even installing another linux distro (Manjaro), but the error persists. The audio is actually seems to be working fine - I can listen to music or watching videos or playing games without any problem. The problem is only in this unity project. Unity version is 2022.3.52 (LTS). Any suggestions on where to dig?
P.S. On windows the project opens and works fine.

[rain canyon]

My question was poorly worded, I was more so concerned with grabbing the footsteps from the pool every time I want to play a different pitch of the sound and how much of the 32 audio sources that could take up, but I figured it out now.

I was wondering if anyone has any other ideas for what these don't cover: I have a "OnClickSound" script that handles sounds that play when something is clicked. I have a "PlayOnAwake" script which handles static objects like waterfalls, buildings etc that just need to play a sound either once or on loop. Are there any other uses that I haven't covered in these scripts that you can think of?

One is random pitch looped sounds, for example footsteps

Music will be a separate subsystem

[spark cove]

yo anyone got an idea for a theme song its for a sad game

[earnest pasture]

Is there any reason to use AudioSource.Play instead of AudioSource.PlayOneShot in a scenario where you don't care about stopping the sound as its playing ?

The only advantage I found is the ability to cut off the current clip, at the cost of making using the audio system really awkward (from what I understand)

[quasi hound]

Probably not

[hexed falcon]

Thought this seemed like the best place to ask this,
but is anyone else having a Vorbis decoder failure when using Unity 6 (6000.0.23f) and Wwise 2024.1.0,
where for some reason all audio that's been converted to Vorbis in generation, cuts off with that error after pretty much exactly 2 minutes.

I don't use Zero Latency like Wwise mentions as a probable cause, and there isn't any errors that come up during the generation itself,
so Im pretty stumpted on what I could try here. Wwise 2023.1.7 works fine in comparison.

[fierce ibex]

Hey, does anyone know how I can have music with an intro loop back to a certain part of the song, so that after the intro it loops the main part?

I cant find any good tutorials

[flat spruce]

{
    int frequency = originalClip.frequency;
    int startSample = (int)(startTime * frequency);
    int samplesRequired = (int)(duration * frequency);
    
    float[] data = new float[samplesRequired];
    originalClip.GetData(data, startSample);
    
    AudioClip newClip = AudioClip.Create("Section", samplesRequired, 1, frequency, false);
    newClip.SetData(data, 0);
    
    return newClip;
}

that help?

[fierce ibex]

this will allow me to set the looping points in the editor?

[flat spruce]

I don't know

but actually you should just use unity's built-in loop points?

[fierce ibex]

oh I didn’t know unity had its own

[flat spruce]

https://docs.unity3d.com/ScriptReference/AudioSource.html

[fierce ibex]

thanks

[flat spruce]

{
    public AudioSource musicSource;
    
    void Start()
    {
        // Set the loop points
        musicSource.loop = true;
        musicSource.loopStart = 20f;  // Start loop from 20 seconds
        
        // Play from the beginning first time
        musicSource.Play();
    }
}```

my first answer is from jippity, second from claude. claude seems to know some tricks.

[rain canyon]

any idea why I can still hear the sound despite being over max distance away from it?

unity 2022.3.26f1

[rain canyon]

Spatial Blend is set to 1

also another issue - tabbing out of the editor while in playmode ends one of my sounds and puts them back into the pool, as if "isPlaying" gets set to false?

[quasi hound]

Logaritchmic rolloff reaches zero very slowly
That means if you decrease max distance it'll simply cut off earlier at a higher minimum volume, become easier to hear over distances rather than harder

So you'd grab the furthest right curve key and set it to zero at where you want the volume to be zero

So probably 200

[rain canyon]

that seems really weirdly implemented... how can I adjust max distance properly at runtime then? or is there no way

also in the screenshot it looks like it's at 0, it could be very close to 0 but not quite there, but I could still hear the sound clearly

[quasi hound]

Can't be quite 0 if you're hearing it

[quasi hound]

Logarithmic follows a specific curve so it's not very good for runtime adjustment, unless you also get the last curve key and set it to 0 in code
But that also causes issues if the range varies wildly
Linear works better, but there's not really obvious best solution
In what situation are you adjusting it

[rain canyon]

im using audio pooling and setting audio source attributes to what they need to be when they're pooled, so I'm setting rolloff mode at runtime, but if I want 2 different logarythmic rolloff sounds to behave differently, then I'd have to pass that with the other attributes

[quasi hound]

You'd probably want to save the curve too

Or just all component values wholesale

[rain canyon]

yeah most likely, is there any way to get that onto a scriptable object in a reasonable way or?

[quasi hound]

Probably, but one one that I know of off the top of my head
Could also keep a reference to the prefab, but I don't know if it's worse to keep those prefabs in memory rather than just arrays of properties

[rain canyon]

i don't know if that would work with how ive got it implemented but ill try exploring around see if i can find anything

[abstract relic]

is there a free audio software? i use free use youtube video sounds.

[spiral trail]

Audacity

[abstract relic]

Thanks!

[ember geode]

Anyone know anyone good tutorial/guides and resources on where to start with making really nice sounds for your game? Currently need some nice footstep sounds, jumping sounds, and a dash whoosh sound.

[inland raven]

Well I do not have tutorials but all good sounds start with recording actual sounds.

Or using pre-built libs

[ember geode]

Are there any specific softwares I should be using?

[inland raven]

Audacity is a nice one to learn

[fast beacon]

• Make sure your sound effects have a similar volume relative to eachother. You can amplify the sound effect’s volume in post.
• Reverb makes things powerful.
• Make sure your Mic doesn’t clip.

[uneven badger]

How would u fix the constant audio crackling (krlrklrkrlrkrlrkrkrktrkrkrkrrrtk krkrr klkkktrrt kkrrr kr krr KRR so u know what i mean) because of a lot of 3d audio sources in the scene with the same priority. They all got a custom pretty short max distance tho so idk why that should be a problem. In the profiler the audio cpu usage seems very high like 150%, idk if thats why it happens.

[odd path]

I need some tips on trimming small short sfx. currently I'm separating footsteps sound into 8, the thing is when I cut it where I think it starts and ends, the exported audio didnt produce any sound even though it sounded fine in audacity. I have to add a "what i think is empty" leeway in the beginning, and determining how big it is annoys me

perhaps there is a setting in Audacity for editing small audio clips so what i heard in the editor is what i will get in exported audio ?

[odd path]

alright so apparently some video player buffer for a bit at the start while playing the media. so if a media is so short it ends before the buffer finishes, it doesnt get played.
they still works fine in unity, at least in my case.

[obsidian pivot]

Hi if anyone is familiar with Wwise need some help, we're adding an RTPC to control pitch in the following code

[Header("Sound Settings")] 
    public string engineSoundID = "Speed";

    public float truckSpeed = 0;
    public float maxTruckSpeed = 100;
    
    public void UpdateSpeed(Component sender, object data)
    {
        if (data is not float speed) return;

        truckSpeed = (speed / maxTruckSpeed) * 100;
        
        Debug.Log(truckSpeed);
        AkSoundEngine.SetRTPCValue(engineSoundID, truckSpeed, gameObject);
    }```

we were using a single main bank and it worked properly

we have now switched to seperating the banks by category, so now a Vehicle Sound bank is attached to the truck where the AKAmbient components are as well, the sound plays but the RTPC no longer changes the sound and we don't know why

So to reiterate, the only thing we changed before it stopped working was making seperate banks and this one is dedicated to the Vehicle sounds

[noble hatch]

How do i get my audio soruce to play in the play mode? i have it to play one awake but it doesnt play in the actual play mode

[formal depot]

Does it have an AudioClip assigned to it?

[noble hatch]

Yup

If I click the speaker on the edit mode it plays but still not the play mode

[hard gulch]

Game view has its own mute button, make sure thats not enabled

[noble hatch]

it isnt i already checked that a couple times

[hard gulch]

Do you have an audio listener in the scene

And what are your audio source's settings?

[formal depot]

I recently spent like an hour trying to help someone with this and it turned out to be something really silly

but I don't remember what it was...

it might have been the game-view mute button

[autumn hazel]

Someone here knows to I can emulate N64 reverb and delay?

And no,I can't just download and use GameVerb.

Not only I'm broken but the dollar is high as hell in my country

[urban lake]

Hello1 Does anyone have any info on audio improvements included in Unity 6? I know they added a new way to randomize audio clips, but I was wondering if any changes on attenuation shapes and occlusion has been made.

[urban lake]

Never mind... found my answer 😦 .... bad, bad, bad Unity.

[meager storm]

Audio DSP should be something like 15% not 150% .. did you find the problem yet ?

[meager storm]

Unity has a roadmap doc with some audio improvements planned

[uneven badger]

I solved it by adding a script to all 3d audio sources that constantly checks if the player is within its max audio distance, and disabling the source if theyre too far away

Kinda feels like a weird solution tho, feels weird id have to do this and idk how much it affects my performance, i will end up with probably more than 300 audiosources... so if u got any better ones or info about this let me know

[formal depot]

Audio sources will still play even when you're past the max range (and still use a "real voice")

Unity supports 32 real voices by default, I believe? I know that's the number in VRChat

Anything past that will caues the lowest-priority, then quietest, audio source to be evirtualized

meaning that the audio is no longer processed, but the clip's time still increases

so when it becomes real again, it sounds like it was playing the whole time

[formal depot]

Do you have any audio filter components in the scene?

[uneven badger]

Oh alright, well theres gotta be some good way to solve that tho , cuz i wouldnt say 32 audiosources in a scene is that much, and if u up the limit it starts krkrkekrkring right

[uneven badger]

I do yeah

[formal depot]

That could significantly increase the amount of audio processing that's needed

[uneven badger]

guess u could maybe also change the priority in a script so it would keep playing from where it left off but there should be some inbuilt thing right

[uneven badger]

Would reverb zones solve it

[formal depot]

I'm not sure what you mean by this. When you run out of real voices, the least important audio source (by priority, then by volume) gets virtualized -- it keeps playing, but it's not actually producing any audio

Are you asking if you can make it pause when virtualized?

Because you could certainly do that, although I don't know how you would correctly unpause it later

[formal depot]

That'd help make sure you only have one reverb filter at a time, at least

[uneven badger]

I guess yeah? Or like when the player gets out of range from the audiosource, that source gets its priority set to low instead of disabling the audiosource

[formal depot]

That's already roughly what happens. If every source has the same priority value, then the quietest one gets virtualized first

[uneven badger]

Shouldnt there be some in built thing for this tho cuz i feel oike this would be common

[uneven badger]

Hmm yeah ur right

[formal depot]

It sounds like your problem is that, with 32 real voices playing, the audio system is getting overloaded

[uneven badger]

Mhm

[formal depot]

Try disabling all of your filter components and looking at the audio cpu usage

You should be able to find all of them in the scene by searching the hierarchy

[uneven badger]

Alright ill try that soon

[formal depot]

hm, that's annoying: there is no parent class for the audio filters

they're all just Behaviours

so you can't just search for t:audiofilter or whatever

[meager storm]

Your problem is the reverb filter on your audio source

Apply reverb at the audio mixer or using reverb zones

That will fix your problem

It’s possible to use reverb filter on an audio source but that breaks down at scale exactly like you are describing your original problem with a lot of short sounds

In project settings, there’s a bool in the audio tab that will keep filter effects on the audio source active. The default is disabled false.

This audio engine feature shuts down an audio source when the clip is not playing. However, this behaviour conflicts directly with adding the Reverb filter because the Reverb filter is not complete. So you get the sound of the clip cutting off.

[uneven badger]

Aha.. yeah I tried it and im pretty sure u guys are right, turning off the filters seem to do the job

Thank you

[meager storm]

experiment with removing that code you created to check if the audio source should play .. that's what the underlying audio system does for you in the background already .. that system is highly optimized and 'just works'

for now keep all your sound effects in the same priority setting in audio source .. you'll want music and other sustained 2D sounds to be a higher priority , the numbering is arbitrary .. so setting 2D and music to priority "10" works because it's higher than the default "128"

set to a higher number means your audio source will not be sent "virtual" once the playing voices reaches 32 which is the engine default limit

hope that makes sense

[uneven badger]

Yeah it makes sense nice info, and i did try to remove the code and it seemed to still work, prob gonna test it again when i get home to make sure tho

[viscid fjord]

Has anyone been faced with a import error saying "Errors during import of AudioClip {path/to/file.mp3}: FSBTool ERROR: Internal error from FMOD sub-system." - my project does not use FMOD, I guess Unitys built-in importer does, and it happens every time I try to import 1 specific mp3 file, which is odd because it was exported out a DAW the same as every other mp3 file, I can load it in Audacy, play it in a browser, etc and the file itself is not corrupt, all its byte data is there, any ideas what the issue could be or how to fix it? I tried looking online which basically suggested "its a known problem and is already fixed" with many replies saying "its broken again in later versions" but the cause or solution isnt clear (as this is not a problem with any other 2K+ audio files), im using 2021.3.x LTS but I dont think the version should have an impact on this problem (as it seems to be a outlier from my investigation)

[whole pewter]

• unity uses fmod for audio (it's their own customized (and older) version)
• fmod itself (as in 'normal' !unity specific version) has issues decoding some mp3/mpeg files (*
• updating unity might help, but i'd rather
• reencode the file - use different bitrate/quality etc
• if the file is somewhat large, split it (but IIRC in that case the error could be different one)
  (fsbtool is just editor time audio importer)

(* some mpeg files can't be properly played even w/ latest FMOD provided engine sample
it got properties of the format wrong and played it as twice as long, with 2nd part being silent; other audio players were able to play the same file normally
(i had a workaround in my asset to detect this, but it's BS heuristic to do anyway - i removed it since)

[viscid fjord]

That helps a lot, thank you for that explanation!

[formal depot]

Is there anything particularly interesting about the MP3? Is it very long?

[viscid fjord]

Nothing really interesting about it, its < 3MB and about 2 mins long, just a looped audio track for background music, many other music files in the project are around the same 2 - 5 mins long, under 10 MB in worst case, and all work fine except this one file, all exported with the same settings in-DAW

[dense topaz]

Do you think this could work for some kind of castle?

[meager storm]

@viscid fjord The expected workflow is importing asset audio clips into Unity as full resolution .wav then using the import settings to compress. If you import .mp3 you're loosing a lot of quality becuse the engine will compress a second time.

[whole pewter]

yea it does i forgot about this; depends on track character though - you gain very little for looping ambiences / background tracks except for several times larger assets
but it's probably the easiest way to avoid all the above issues with importing..

[violet mist]

how should one approach complex runtime audio generation in an ecs/dots project? is there any guidance on interfacing between jobs and OnAudioFilterRead?

or is there perhaps some other means of outputting generated audio with ecs/dots than OnAudioFilterRead

[violet mist]

tested out DSPGraph but unfortunately i'm running into domain reload getting stuck which seems to be similar to this https://discussions.unity.com/t/domain-reload-stuck-when-using-multiple-audio-apis-in-82349/1501927

[whole pewter]

yea it's probably far from being usable yet (..still)

if you search forums you should be able to find github project of someone who used the whole stack in their ECS framework can't recall how it was named now

[violet mist]

if you mean this one then it's the same person as the above bug report thread https://discussions.unity.com/t/dspgraph-current-limitations/937491/5

i'm able to get audio playback to happen through DSPGraph but unfortunately after playing any audio through it, any code change will cause the editor's domain reload to get stuck

which makes it impossible to use in practice since i have to restart the editor for every code change i make

i'm perfectly happy to use OnAudioFilterRead as well but i don't know how to synchronize it with jobs since it's called on the "audio thread", i wonder if there's any documentation for that?

[whole pewter]

find out how to pass data from a job to main thread in ECS - this should be mentioned somewhere.. - you should be able to get it into audio cb identically (if not, i.e. job will complain/not work, you will have to use main thread as intermediary)

[violet mist]

what i'm thinking of is if there's something that lets me schedule jobs in an optimal manner wrt latency

i can of course just output things from jobs into a large enough circular buffer, read from that same buffer in OnAudioFilterRead and hope for the best, but i'd like to understand the system better

though i suppose if i want to really go low latency then i'll have to just bypass unity and use platform interfaces

[meager storm]

All that DOTS era audio stuff is abandoned

[violet mist]

so it seems... very unfortunate

the dspgraph interface seems like a very nice start

[viscid fjord]

Really? Interesting, to my understanding, I had thought its best to have different formats depending on the usage in-engine because of the file type compression, where games usually use mp3 for music, and ogg/wav for sounds and voice files, I also thought the compression settings were mostly a engine advantage and get decompressed at runtime/in a build - is the standard having ALL audio files being wav regardless of how they will be used in-engine?

[whole pewter]

^ well yes if you care about quality (which might be non issue..) - when you import an audio asset unity recompresses it, depending on import setting
but note that some platform use hw decoder for some formats (iOS has AAC, possibly mp3 hw playback... ) but this should be in documentation ..

[whole pewter]

with some limits i.e. playback limited to one file/asset so it's suitable only for long running stuff such as background/ambience not general sound produces by app -- but i'm not entirely sure if unity uses this now tbh hah

[whole pewter]

/ but i'd say the likelihood unity doing absolutely nothing about hw supported playback on some platforms+formats is nearing 1oo% since it doesn't allow to indicate this anywhere in import settings so you can probably disregard the above

[whole pewter]

oh and one more thing: IIRC fmod can decide to use platform hw decoder automatically when playback is issued up to some limit - but im not entirely sure about this....

[meager storm]

Yes @viscid fjord it doesn't say anywhere in the docs but the standard is to add all audio files as .wav to Unity engine and use the import settings to compress

there is no standard compression format for sfx or music in games . the import settings are critical to understand as they affect performance at runtime

for example in most cases you would set music to stream

[viscid fjord]

Interesting, I havnt noticed any performance issues but I also have pretty good specs and we havnt added all our sound files to the project yet, ill keep that in mind when we have the files to profile, from previous games ive seen, I noticed a lot of their files were in ogg or mp3, (although some of them where early 2000s era games), im guessing in other engines this is likely not a problem and they may have chose for formats out of preference?

[meager storm]

rest assured that .wav is the source and .mp3 or .ogg are compressed destination formats

if you choose the wrong unity import settings and you have 3000 audio clips then that's where you run into perf' issues .. really depends on the project scope

[viscid fjord]

Ah I see, thanks for the insight

[terse haven]

What's the right way to loop audio? Just checking the Loop in an audio source causes a small gap of silence to appear before the loop happens

[regal needle]

Just putting it on loop is fine, my sounds loop super fast.
It's probably your sound itself. Can you show the sound wave graph of the thing your trying to loop.

[terse haven]

As you said there's a short silence at the start and end, which is not there when I open the source file in audacity for example

[regal needle]

Well at least we found the problem, now we need to find a solution

[terse haven]

Do you know the solution? I don't know what to search for this

[regal needle]

No, I usually just pay people for my sounds, never had this issue.
You can check if this is a Unity issue by just playing the sound outside of unity and see if it also has the 0-bands at the end and start

[terse haven]

It does not

[regal needle]

Well there doesn't seem to be some sort of Unity import setting for Audio that creates the 0-bands as far as I can see

Can you upload the sound here?

[terse haven]

Here

There is a tiny fadeout at the end to minimize clicking on the loop

but still you don't hear the short silence when playing on loop anywhere

[regal needle]

In VLC the sound waves really do move down at the end.

So it seems like an Audacity export setting then.

Anyhow, I don't know, perhaps someone else can figure it out, but the 0 bands at the end and start are your problem, don't know how to fix.

[terse haven]

That's weird. I'll try some stuff later

[sly nexus]

Hey can someone help?, all my audio sources in my scene doesn't play/work I had this problem for a month and I couldn't solve it

[formal depot]

You don't have the Game View audio muted,d o you?

[safe totem]

my audio clip is just a bit too long for it to sound smooth (its a landing sound clip) can I somehow just remove the first few milliseconds?

[formal depot]

Use something like Audacity to edit your audio

[safe totem]

there has to be a other way, Unity internal

it would be so dumb, having to export multiple audio files until you get your desired length

[formal depot]

unity is not a DAW, a photo editing suite, or a 3D modeling program

You can certainly fade the volume out on the audio source, I guess

[patent tartan]

I dont know if it was reported, but setting time longer than audio length in audio source, gives internal error instead of standard unity error.

its like there is no check inside the unity code and it tries to set it (mostly likely)

im not sure if it can break something at runtime/in build.

[sly nexus]

so my linear rolloff is broken in my audio source which i think makes the min and max distance broken, my audio sources dont work unless i put it on costum rolloff, still my min and max distance is broken can you please help

[formal depot]

show the inspector for an audio source that's not working right

[sly nexus]

ok

its all of my audio sources in my scene btw

[sly nexus]

any idea how to fix it

[formal depot]

This has a max distance of 2

[sly nexus]

im kinda talking abut my audio sources NOT EVEN PLAYING IN SCENE so yeah

[formal depot]

yes, and the max distance is 2

if the audio listener is more than 2 units away from the audio source, the volume will be 0

[quasi hound]

Why multiple times? You could cut it to the right length the first time
Or pause/fade the clip in Unity if the playback length is not precisely known or varies

[sly nexus]

no its like

if i put it on like 10 or wtvr

so like the whole scene could here the audio

its snot evenplaying

like at all

[tall oyster]

hello, I'm trying to add a background music to my game and then a slider to control it. But when I hit play, I don't hear the music. Can you help me? I join a picture of my AudioSource. I tried to remove the output value (with wich i found that a sound was displayed), the music is good because I tried to play it inside the editor and it worked. Can you help me please

[formal depot]

start by setting the spatial blend to 0

this shouldn't be a sound that's coming from a specific place

also consider clearing the "Output" property so that it isn't going through a mixer

you might have a problem with the mixer in the built game

[sly crystal]

im trying to make a game where the gameplay causes the tune to evolve, anyone got any idea how to do that? lol

not evolve as such but add more layers etc, ive been looking at rez as an example

[quasi hound]

!collab

[karmic bone]

:loudspeaker: Collaborating and Job Posting

We do not accept job or collab posts on Discord.
Please, use Discussions to promote yourself as job-seeking, advertise commercial job offers, or look for non-commercial projects to participate in:
• Collaboration & Jobs

[quasi hound]

You'll need a whole procedural music system in that case
A middleware will help because from what I hear managing and syncing audio to that degree is kinda tough to do completely on your own

[sly crystal]

yes its difficult for me to even get the concept right in my head

[quasi hound]

Conceptually you'll need to play the instruments directly from scripted sequences that can be altered or rearranged

So you'd first have some kind of midi or tracker player, then implement a system that lets the player interact with it as per your game design

There's quite some demand for systems like that so I'm sure whole third party assets exist for that

[sly crystal]

yeh i dont know of any, fmod sort of looks relevant

[quasi hound]

Fmod and Wwise are the most common middlewares for precise audio control and effective management, but there are others also like Sonity
I assume none of them do the kind of music instrumentalization out of the box, but give you the tools to build that system

[sly crystal]

Thanks Spazi, u seem knowledgable about this 🙂

[quasi hound]

Appearances can be deceiving!
Most of what I know about audio is from observing others, I've had no need for anything fancier than Unity's own audio sources and mixers
But it might be enough that now you have a better idea where to look for the information you need

[meager storm]

@sly crystal what you're talking about is called "dynamic music" there's a bunch of info out there about how it can work. It's non trivial. It requires and advanced understanding of music production and audio design for games. You can however do simple layer cross fade in unity using the mixer but you would need the music created in separate tracks as .wav

So a very simple version of dynamic music is possible using native Unity audio. I would not recommend looking at FMOD or Wwise unless you already have a strong foundation in audio for games. Those tools are very advanced

[lone forge]

my fault, collab would be sick but i guess i'll just focus on my own project

[glacial raft]

I feel like this is the best spot to ask this
Does anyone know what this sound effect is called (not the ticking sound) or where it comes from? I recorded this from a really old youtube video

[hard gulch]

Orchestral hit

[glacial raft]

@hard gulch THANK YOU!!! That's exactly it

found a good one on pixabay

[grim storm]

I have the max distance of this audio source set to 1.01 yet I can still hear the audio from across the map. How do I fix this?

[quasi hound]

If you want to place a hard limit on the maximum audio distance, use linear or custom rolloff rather than logarithmic

[whole pewter]

^ the source is 2D

[quasi hound]

True that's the bigger problem

But still if you cut off the logarithmic curve with a low max distance, the volume at that distance will be above 0 which means it'll be played at that volume beyond the max range up to infinity

[whole pewter]

unity(fmod) will eventually virtualize the sound if it's not audible - at least it shouldn't be audible across the map, and problem seemingly solved - but it's still resources being used yes

[quasi hound]

Just to be sure we're talking about the same thing
A logarithmic rolloff audiosource with a very low max distance will never be inaudible

1.01 would play it at 99% volume everywhere in the scene

[whole pewter]

👍 lol i haven't even looked past that 2D blend on the source

[quasi hound]

And I didn't glance above the rolloff settings in turn ^^
This logarithmic curve quirk is just something I wish I had noticed earlier and see a lot of people trip up on

[whole pewter]

it should attenuate to 0 past max distance cutoff if source has full 3D blend set IMHO
do you happen to know if there is any opened/closed report / by design, for this ?

[quasi hound]

I think the logic is that regardless of min and max, the logarithmic option still gives you a real world accurate logarithmic rolloff
Since it's basically a physical constant

Using a custom curve instead with a similar ramp over a shorter distance is very simple alternatively

I haven't seen any official report about it though

There's no indication it's not working as intended, it's just an easy detail to miss that the tail end of the curve starts rising up from zero as distance is decreased

[whole pewter]

so you think min/max distance for logarithmic rollof curve defines it just partially and function is applied regardless of the distance ?
that looks at least slightly confusing to me tbh
it might be just some fmod quirk they didn't bother to fully implement/document but idk..; at least the names of parameters sound not entirely descriptive to me 🤷‍♂️

[quasi hound]

I wouldn't say partially because the function is not separate from the curve
The actual rolloff follows the curve, but the curve's shape doesn't change
Rather it flatlines where it gets cut off

[whole pewter]

yea that's what partial def. is - i.e. defined only on some domain interval - but it's not 1 function, but two in this case ( "log" + constant past 'max distance' )
somebody should bother to file report for them at least to clear up the params names 🌚

[quasi hound]

You best do that if you want to see it done ^^

[whole pewter]

i, too, wish so ]

[meager storm]

that log curve behaves totally different than a custom attenuation or liner attenuation .. to see what I mean choose linear then adjust the min and max distance

custom curves cannot use min distance

it takes some experimenting to find a setting that works for your design needs

I use custom curves all the time because I have a different method for defining minimum size using simple colliders to define the minimum size area

[quasi hound]

Colliders? How does that work in your case

[meager storm]

set to trigger , the audio source follows the player using the collider as the boundary , then as the player leaves the boundary audio source attenuation kicks in

[olive pasture]

I tried making a button that plays a click sound when clicked but it is not playing the sound on first click however on it works on second and later clicks. This issue remains after not clicking anything for a while. Should I optimize the audio file or use another method?
Setup:
audio listener to cam,
audio source to button itself
audioSource.Play() to button click

Some buttons sets game object and itself as inactive when click on which plays no sound 😦

[formal depot]

well, yes, if you turn off a game object, an audio source on it will be disabled

don't do that

Perhaps you want to use a single audio source to play all of your UI sounds

[mild geyser]

hey guys, anyone here use wwise? would you say its tough to learn? I would like to incorporate it into my project, as i havent delved too deep into audio yet. i've heard wwise is free too. thanks!

[cobalt kraken]

🫰

[naive thicket]

@cobalt kraken There's no self promotion in help channels.

[pale timber]

There's lots of free resources to learn it. It is not too hard.

[pale timber]

https://www.audiokinetic.com/en/learning/learn-wwise

No need to pay for the certs unless you really want them. Just go through the courses.

[pale timber]

You need to offload the AudioSource to a different object so that it doesn't get deactivated.

You might try something like this. Might be more complicated than you need, but it's a pretty nice design like most content from git-amend.
https://www.youtube.com/watch?v=BgpqoRFCNOs&ab_channel=git-amend

[fiery marsh]

Heya, where do ya'll get gun sounds beyond unity store? Looked through the variety, but it's kinda limited, i need gun sounds that fit older guns, 1940s era

[formal depot]

That era should sound pretty similar. You've got a mixture of full-power rifles (that'll sound close to modern carbines) and submachineguns (matched by modern pistols)

I guess that's just the actual firing sounds, though

Are you looking for other sounds, like working the bolt?

[fiery marsh]

Yes, and also sounds for certain japanese weapons such as Type 99 LMG, Type 14 pistol and Arisaka rifles

[formal depot]

A lot of little gun-handling sounds can be made by something other than the actual firearm

i've gotten good use out of the sound of me putting the paper tray back into my printer, haha

https://freesound.org/people/chemicalcrux/sounds/530960/

[pale timber]

I got a ton of use out of the Triune Hollywood Guns pack. It's nice that the sounds are already designed to be beefed up in case you don't want to spend a lot of time just using them as source to design a new sound.

https://www.triunedigital.com/products/hollywood-guns-sfx

And as Fen mentioned for gun handling you can make custom stuff using household items.

Otherwise you can look for gun Foley packs. Triune has a nice one of those as well.

https://www.triunedigital.com/products/gun-foley

[naive thicket]

@pale timber Please remove referrals

[pale timber]

Huh? Like delete the links?

[naive thicket]

referral id is for kickbacks etc on asset sale

[pale timber]

Okay my bad. Thought this would be fine since the question is about gun sfx

[naive thicket]

You can edit it out from the link

[pale timber]

Ohhh gotcha

I just copied the URL after doing a google search

I will see if I can get links without referrals

[formal depot]

you usually just chop off everything after a ?

that's why amazon links are always 30 pages long

lots and lots of tracking tags for things like "how did they reach this page?"

[naive thicket]

Only the second link had one as well

[pale timber]

Fixed

Learning moment. Thanks

[distant dune]

did anyone here ever see a crash in UnityPlayer!FMOD::OutputWASAPI::recordUpdate+0x1ae? some of our players are having crashes and the dumps report memcpy on an invalid ptr there as the reason for them but google doesn't have any results for this and we don't have any reliable reproduction

(posting it here cause the place here is the unity audio engine)

[whole pewter]

you should be checking if the rec device you started on is still present, and stop/restart on different one possibly, if not
but this crash might not be completely avoidable since it might depend on user hw (+how the version of fmod unity uses accesses/uses it..)

[distant dune]

i think it's the output device

cause of the output

and unity just called the method for playing more audio record for some reason

and the people that have this told me they weren't doing any unplugging or anything

[whole pewter]

you can record from 'output' on windows under special circumstances if it's configured to have its loopback available for recording
but regardless if it's available then the recording should just work - so it's specific hardware in this case most likely and w/o UT being able to repro this they probably won't be able to do anythign about it...

[hushed summit]

Hey, guys! I would like to know if there is a method to actually prevent a sound effect from passing through walls and obstacles like audio absorption or something, because I have an audio source and it has 3D spatial blend.

Is there anyone available to give me some suggestions or whatever?

[delicate rivet]

nothing built in to unity really, 3rd party tools like steam audio has some occlusion stuff https://valvesoftware.github.io/steam-audio/doc/unity/source.html its been a while since I used it, I am not sure if you can make walls completely stop sound coming through

you could roll your own, simple thing like using Physics.Raycast, seeing if it hits an object tagged with "wall" or something before it hits the listener then turn down the volume on a source. This is pretty naive, only takes into account direct path sound. but maybe its enough to do what you want.

[formal depot]

That is roughly what I've done

along with using a navmesh to figure out the length of the path between the audio source and the listener

(I need to try out Steam Audio -- pretty sure it supports ARM now...)

[spiral trail]

There’s a paid package called AudioBounce that does this. You could take a look at their demo and see if it’s something worth investing in.

[hallow marlin]

Anybody got a sound libary they could recommend?

Mainly looking for sound effects

[inland raven]

Steam audio simulates occlusion as well https://github.com/ValveSoftware/steam-audio

[inland raven]

https://sonniss.com/gameaudiogdc

[hallow marlin]

thank you

[limber oar]

does anyone have general advice on how to get started making music/SFX for games? im completely new to making audio in general and have a very very basic understanding of music theory. if anyone has any apps to recommend/resources to learn to get started, that would be awesome. Thanks!

[pale timber]

Do you have an understanding of common implementation approaches? If so it’ll help you know how to best prepare your assets.

For example it’s fairly common to have layered music, so if you render out your music not only as a master mix but also with separated-out stems, it would make your asset pack more valuable with only a small amount of extra effort

Also for music, including the meter, tempo and key in the files name is nice

[limber oar]

I'm not entirely sure. I get what layered music is (mostly based on the name lol). Idk what other "common implementation approaches" would be though

[pale timber]

A simple example of layered music would be if you have a main melody/harmony layer and a percussion layer. During exploration gameplay you only play the first layer. Then the percussion layer gets faded in and out as the player enters or exits combat

The percussion layer then adds to emotional excitement to enhance the sense of being in combat/danger

While keeping the loops and transitions seamless

[limber oar]

that makes sense

[pale timber]

If you’re totally new to all this you can try an online school I did years ago when I was getting started.

https://school.videogameaudio.com/apply/

Otherwise you can learn a lot through YouTube for free

[limber oar]

thanks!

[long owl]

when ever i play a sound many times in a row, the bass gets randomly boosted

how can i fix this

[hard gulch]

Are you sure it isnt just played multiple times at once?

[long owl]

like using the same audio source component

[long owl]

ok i changed it to make a new audio source foreach audio played and that worked

[cosmic pendant]

is it possible to add a sound to a an output event on die in vfx graph?
You can use the otput event for sound but this can not be added to a trigger "on die" event
I have fireworks and when they die it starts the event trigger of explosion.
If i joint the output event to the fireworks trail(main event) the sound stras at the beginning.
VFX graph does not allow to join the output event to the GPU Event (trigger event)

[kindred thicket]

I have no idea how VFX Graph works, but I saw only one way to get callback from GPU using AsyncGPUReadback
Check this: https://youtu.be/e-hTTVr_pDI?t=426

[tacit spindle]

Im pissed off… how do I get audio to not desync when using Unity recorder package with Timeline? Sound effects too near the start of the TL are playing desynced. I already set the sounds in import settings to preload

[raw ridge]

anyone know why when I press the search icon, the event option does not have a dropdown indicating my events created in fmod? i'm using the EventReference field but I cannot locate the events I created in fmod. anyone know why?

[pale timber]

Was it working before? Are you sure you have your FMOD integration set up correctly?

This indicates that it has no data from banks to display

Do any errors show up in the console when you try to browse for an event?

[raw ridge]

nothing at all, the integration and all was all correct but somehow it still didn't work. I switched to wwise and got that to work though lol

[pale timber]

Okay. Note that Wwise does not support web builds

[pale timber]

I’m actually working on a package for Unity to handle a lot of the common inconveniences that I’ve run into with FMOD. Hoping to have it on the asset store in the next couple months. A big part of it is it parses the FMOD project’s xml data to auto-generate scriptableobjects for events, banks, parameters and snapshots, so you don’t need to fiddle with the EventRef browsers and use string fields for so many things

Makes it more similar to working with the Wwise integration

I’ll post it here when it’s up on the store. Hopefully people can find it helpful.

[raw ridge]

nice! I might buy it because I really like the fmod interface compared to wwise and is probably more suitable for my project

[formal depot]

i also wound up writing a sort of "adapter" class

mostly for my bootleg spatial audio system

[pale timber]

Curious, when you say “spatial audio” are you referring to ambisonics/HRTFs, or are you referring to propagation like Wwise’s rooms and portals?

[formal depot]

The latter

i need to look into Steam Audio

[pale timber]

Oh nice. That was gonna be the next thing I was gonna try to tackle for FMOD/Unity

Kinda tricky when I imagine it though because I don’t know how to programmatically just apply things like HPF/LPF and reverb send per event instance like Wwise does under the hood

[formal depot]

I would've started with Steam Audio, but I'm pretty sure it didn't work on ARM for a while

but I believe it does now?

[meager storm]

@pale timber that's an extremely complex task as I am sure you are aware

I would caution you against that because it's better to just use the tool that already exists = Wwise. The effort required to write an effective Unity implementation for FMOD that would get similar results is not a one person job 🙂

[pale timber]

Haha. I mean, I use the Wwise Spatial Audio tools all the time at work. Trying to make a similar framework for FMOD is more about personal challenge and potentially making some side money

But yeah I’m aware it is not an easy task

That’s part of what makes it exciting and interesting though

[meager storm]

Sure it's about the challenge being interesting I get it

[pale timber]

The other thing is I would like to have at least a simplified version of propagation like that for FMOD because I use FMOD for game jams since it supports web

[meager storm]

But think about how little exists in the current Unity / FMOD implementation. And also how events are instanced. You would need to do so much.

[pale timber]

That’s what got me started on making my own framework that’s led to this scriptableobject generation stuff

[meager storm]

I mean that's a really valuable and effective effort . We use scriptable objects here as well.

[pale timber]

The theory behind it doesn’t seem too complex — move the room emitter to a gate/portal and recalculate its attenuation to simulate diffraction , but yeah I’m sure it is quite a challenge, especially to do it with good performance

[meager storm]

for every sound instance

every game frame

heheh

[pale timber]

Right.

[meager storm]

Anyway yeah it seems like there a huge gap in Unity native and even in FMOD where spatial is not considered at all.

[pale timber]

Yeah it is a really premium feature to have just bundled with Wwise

I should look at the integration code Wwise uses :3

I have looked at it somewhat out of necessity on the UE side but not Unity

[meager storm]

I'm currently comparing the different spatial options from different sources and really Wwise is coming out way beyond the rest of the options

[pale timber]

Yeah they really have done a great job with it

And once you get used to that system it really reveals how primitive more traditional positioning is

[meager storm]

yes that's exactly my point

[pale timber]

Which is why I want that for my FMOD 😅

[meager storm]

FMOD and Unity are using tech that's 25 years old

[pale timber]

My main concern before really starting it is that doing enough ray casting to make it work will be too much of a hit to do in C#

[meager storm]

its too simple now to support how sophisticated game are looking and how much we know about acoustics etc

[pale timber]

But I’ll find out when I get there

[meager storm]

Yeah Owen I think you will find there's a reason Steam and others have aren't using C#

[pale timber]

Have you tried doing a room/portal system with native Unity audio (or FMOD) at your studio?

If so I’m curious to hear what sort of issues you guys may have run into so I know what to look out for

[meager storm]

Yeah we have been working on if for about 5 years in very small steps

We can set something up if you want just DM me

I'm in EST Canada

Owen I think we've talked on game audio slack a few times

[pale timber]

Yeah we have

I saw your video about grouping together audio sources. Looked pretty cool

[pale timber]

Cool. I will take a rain check on this and hit you up when I finish the current package and get it up on the store. Then the bigger challenge will begin 😄

Thanks Joel

[whole pewter]

fmod uses/supports/has google resonance for full 3D
unity itself should consume google resonance properly as well as its spatialization plugin

in both cases it's just positional source, i.e. w/o 'sound materials' / geometry

[pale timber]

Yeah I’m less interested in ambisonics and ERs. It’s the geometry and propagation that I’m interested in having. Things like ambisonics and ER acoustic surfaces are cool and all but aside from VR IMO it doesn’t really have much impact on the experience. Simulation of occlusion and diffraction on the other hand does make a big difference regardless of the perspective of the listener

[whole pewter]

oh resonance has sound volumes
these are driven from C#/Update since it uses Unity physic colliders/triggers, i think this was limited to box colliders (+ it has 'room properties' i.e. different filters) that's all IIRC

[jolly widget]

are BBC sound effects good?

[fresh ore]

hey guys, I have this pouring coins sfx, but when I build and install the game on my phone, it sounds like an empty can.

My BGM sounds okay.

Do I need to change something here?

[brave frigate]

Hey everyone!

I’ve been exploring an app called Emergenyc Sirens And Horns and noticed a really cool feature that I’d love to implement in a similar app I’m working on. Specifically, I’m curious about how the manual siren mode works.

Does anyone have any insight into how this functionality is achieved? Could it be handled through a specific plugin?

I also noticed the app uses MonoBleedingEdge along with the following DLLs:

mono-2.0-bdwgc.dll
MonoPosixHelper.dll

Any guidance or pointers would be greatly appreciated! Thanks in advance.

[pale timber]

Seems like you might need to synthesize the sound… although in Unity you can also give the pitch a negative number and it will play in reverse. Not sure which approach they took here

Using synthesis would probably be a better/more flexible overall implementation.

[pale timber]

You talking about speaker difference or does it sound different between the two platforms when listening through headphones? Looks like you’re not really compressing the sound much so there shouldn’t be a reason for audio itself to be different

[formal depot]

all of the "mono" stuff is just part of the Unity runtime

so that's a red herring

[tall arch]

If anybody could help me with this, it'd be much appreciated 🙏 #💻┃unity-talk message

[tawny lotus]

How do I set up a 3D audio zone in Unity so that a sound gets louder as I approach a floating plane and fades as I move away?

[pale timber]

Distance attenuation

I guess in Unity they call it “Volume Rolloff”

Expand the 3D Sound Settings on your AudioSource.

[pale timber]

Might not be exactly what you’re looking for since you specify that you’re trying to learn implementation, but still could be worth seeing what PowerUp Audio did on the FMOD side and playing around with your own audio. They released the Celeste FMOD project publicly a few years ago and you can generate your own banks and LiveUpdate connect to the game to do redesigns.

https://www.fmod.com/docs/2.00/studio/appendix-a-celeste.html

I haven’t really looked at it myself, but I’m guessing you need to keep the names of existing events, banks and parameters (possibly some other things like VCAs depending on how they did things) the same since those references are in the Unity project. But I imagine that aside from that you can redo the events themselves pretty much however you see fit.

[pale timber]

Otherwise I’d say just get any Unity example project or any free projects that YouTubers put onto GitHub.

Or do a game jam

[tall arch]

Thanks for the leads! I had heard of Celeste some time back but completely forgot about it... Could be a decent option.

I'll have to scavenge around the YouTubes and Reddits a bit more to find some 3D oriented projects as that's been more so my focus. I found a Unity community driven project that was discontinued called Chop-Chop which looked lovely art-wise, but will have to figure how to get it updated and running on current Unity versions.

I'd for sure love to participate in game jams but I've just always found it so daunting on how to find a team...

[stuck skiff]

Hey!

I’m experiencing an issue where the sound on my audio source is popping and glitching. However, as soon as I remove the rigidbody from the same prefab, everything works perfectly fine. Since these two things seem unrelated, this is very strange.

[pale timber]

I don’t think I’ve ever encountered that. Is the object moving? What does the “popping and glitching” sound like? Could you provide a video?

[stuck skiff]

It seems from this post I’m not the only one having this issue happening:

https://discussions.unity.com/t/audio-cracking-popping-regardless-of-everything/911742

Wether it’s moving or not the issue happens, but even moving without the rigidbody fixes it

I’ll make a video once I get home

[kindred thicket]

In provided link it doesn't sound like “popping and glitching”, it's more like pitch changes.
It seems rigidbody quickly moves game object which triggers doppler effect in attached AudioSource. "Doppler Level" controls the pitch speed.
Try to disable it in project settings or set it to "0" in AudioSource settings.

[sinful cradle]

How to do audio and sounds? What's the good and recommended way to produce sounds like for movements and axe throwing and breaking objects in a 3D game?

[hazy oxide]

Hi, I have 360 video with four-channel audio. It's ambisonic video. What is the 2025 path to ensuring the directional video plays accurately in Unity? Google resonance is often mentioned in old 2017-19 articles. Surely there is something newer?

[half sonnet]

A quick glance at the docs suggests you can also use the oculus spatializer. Maybe try importing the meta sdk so that you can use the plugin. From there you should be able to add the spatializer plugin to the Unity mixer and reference that from an audio source

[silver hornet]

Hello all! Been trying to build the essentials project to WEBGL but having issues with audio building

Basically everything is fine with the 3d audio in the play zone but when built to webgl it goes 2d and fixed volume

I have tried using some spatializer plugins like Resonance Audio without success, also haven't found the solution online

is it a known issue? I'm using 6000.0.41f1 version for the project

[versed sail]

There's some limitation on audio when importing to web. I don't know the exact details, but I know some Unity audio features much like the mixer do not work at all, so perhaps that's related to your problem.

https://docs.unity3d.com/Manual/webgl-audio.html

[kindred thicket]

1st order ambisonics is a super basic format, I wouldn't worry about playback accuracy.
Old articles are still relevant, there have been no significant audio changes since Unity 2017.1

[kindred thicket]

It's not an issue, WebGL audio system is based on Web Audio API, so only basic functions are supported.

[silver hornet]

I see!

I've seen some games using some kind of 3d audio in unity play, is there a workaround or does it mean the game was not necessarily built with unity?

[hazy oxide]

Thank you, I’ll try that. I did find the Meta decoders, and pointing the VideoPlayer at an audiosources. I’ve attached Meta plugins to it. Next I’ll look for the audio mixer Spatializer plugin.

[hazy oxide]

Thank you. I’m not yet knowledgeable about Ambisonic formats. If that sector of audio hasn’t changed much for a while, I’ll try Google’s Resonance plugin as well as testing Meta’s components.

My client’s original files have third order ambisonic with up to 16 channels.
In Unity I must research if I need to set those 16 up manually to multiple audio sources placed around the scene as ‘tracks the user can control’, or just one track pointed at one audio source, and the decoder works out what the sound should be at any vantage point. These are 360 videos so the viewing location is immovable in any case.
Clearly I have reading to do. 🤓

[half sonnet]

Either way all you should really have to do is enable a couple settings when you import one of the plugins

[balmy hawk]

Heyho,
I've been experiencing troubles with audio synchronization in WebGL builds and now I want to include a system to show the subtitles for a dialogue, as soon as the connected audio file is loaded, how would I do that?
I only know of clip.LoadAudioData(); but that isnt like a async function that I could use in a coroutine, how would I do that?
Basically I expect to load audio data and exit a waiting loop as soon as loaded = true or smth like that.

Thanks for your help!

[slender flame]

Hi Everyone, we are using Unity and Meta XR Audio SDK and Wwise and have issues getting it to run, does anybody have experience with this? It seems like a conflict with the Meta XR Audio SDK. Thanks a lot.

[heady orbit]

hi, for sounds, known as SFX, it's good to start with Free Sound Library. Recording, editing, mixing sounds is a skill in itself and takes time. I'd recommend you to get good foundations of sound library that fits your game feels, and learn to implement one-shots.

[pale timber]

This seems more like what you’re looking for https://docs.unity3d.com/6000.0/Documentation/ScriptReference/AudioClip-loadInBackground.html

https://docs.unity3d.com/6000.0/Documentation/ScriptReference/AudioDataLoadState.Loading.html

You can write your own Coroutine to check on the AudioDataLoadState.Loading value and only proceed if/when that’s false

[pale timber]

Something else must be happening that isn’t supported in WebGL. If by “3D audio” you just mean it pans left and right based on position relative to the listener and it gets louder/quieter based on distance from the listener, that should work fine, along with the mixer AFAIK

On the other hand something like Resonance or ambisonics decoding for “spatial audio” probably is not supported for web. Can’t say for sure because I haven’t tried it.

Dunno why you’d really want to have that level of detail for a web build anyway. Those technologies’ best use cases currently are for VR and it’s unlikely you’d really want to be running a VR game from WebGL

[spiral talon]

Im making a game on webgl, and im facing an issue where when i try to mute a audiosource when the game is paused using the audiomixer or doing it directly it just does not work, and it has to wait until the game gets unpaused to get muted, anyone know a fix?

[pale timber]

Well this implies that the audio mixer is dependent on Time

I don’t know that for sure but for example if you used this same pause method (by actually stopping time) then you wouldn’t be able to animate something while the game is paused

This looks like that type of scenario

https://discussions.unity.com/t/how-to-have-audio-mixer-unaffected-by-timescale/865542

But hmm in that thread it sounds like it’s just the fade that was dependent on time, which would make sense. But yours is just setting a value directly.

[spiral talon]

that's the weird part, it seems to be a webgl quirk or something, even muting is tied up to timescale, im honestly lost on what to do here

[pale timber]

It might be that WebGL generally runs single-threaded. Not sure if that’s the cause but it might be. You may need to pause in a more sophisticated way if that’s the case

[echo light]

can i post as audio file or does it have to be unityassets

[pale timber]

Maybe this is what you need:

https://docs.unity3d.com/ScriptReference/Audio.AudioMixer-updateMode.html

[spiral talon]

yeah tried this before it no workie ;-;

[frank spear]

Has anyone here messed around with Steam Audio? It's a bit confusing and I can't seem to make it work properly. + I can't find any actual tutorial, there only seems to be people showcasing their projects and the only tutorial out there is from 8 years ago and everything is now different.

[whole pewter]

audio + its interfaces haven't changed (almost literally) in 8 years

[median lynx]

im wondering why I dont have the AudioSource option in my Boolean? This is my uni assignment and its too late in the night to ask my lecturer

im following my lecturer's video 1 to 1

[whole pewter]

you have to assign gameobject/component in scene to the receiver, not an asset

[median lynx]

Ohh thank you so much, works now

[hazy falcon]

Hello! I have a few questions regarding sound. If anyone has a little bit of time, it would be super nice!

I’ve just started mixing audio in Unity, and I want to understand a few basics. Is it true that your sound should never go above 0dB, or else it could cause distortion and unwanted crunching sounds?

Also, one of my problems is that when I keep my sounds at 0dB, some of them sound really low. So, should I boost them but avoid going above 0dB and use effects like a compressor or EQ filter to add more dynamics and improve the overall sound ?

[pale timber]

You want the overall mix to stay below 0 dB to avoid clipping distortion (and destroying peoples’ ears). That of course means any one sound should be well below that and in order to control the mix and guarantee it won’t actually clip at the output you can have a limiter on the master mix, but generally speaking that should only be catching occasional extreme jumps in dynamics and you shouldn’t rely on it to keep your mix at an acceptable level.

To keep mixing simple, you should either render your sounds all normalized and then subtractively mix them in engine, or try to render them at an appropriate loudness level before importing so you only need to make minor changes in volume in engine. Up to you which is preferable.

Either way you standardize something to simplify the process. What you don’t want is for everything to be random and have a big mess to deal with

[minor geyser]

I'm a Japanese school student and not good at English. If you detect any unnatural phrases, please inform me of them! I want to improve my skill of both English and unity.
I want to make a music game. In this kind of games, it is really important not to delay the audio, because players will feel uncomfortable if they tap notes and effect sounds at the just same time. However, in my experiences, it have been difficult to prevend sound delay. Please teach me the way I can do it. I wonder how many music games do it. I'm sorry if my English is not good.

[whole pewter]

use best latency audio settings and make sure the clips being played are preloaded and uncompressed in memory (in AudioClip settings, no vorbis/adpcm codec to avoid decompression at runtime..)
there is also PlayScheduled for sequences (if there are any)

but note that unity might not be the best tool for this overall..

[minor geyser]

Thanks! I'll try it.

[chilly fjord]

I actually saw a stream from some developers of a rhythm game in Unity and they had explained how they got minimal delay I think. I’ll see if I can find it.

Why did my message @ them?

My fault.

Steam doesn’t save past broadcasts it seems.

[sinful cradle]

Yeah! I did. Thanks

[mossy jetty]

I need to implement a system involving 10 parallel bandpass filters which are modified by the player's movement. What tools might be able to aid with this? I've been looking at FMOD but it would really difficult to have to create a set of parameters for each individual filter. I bet I could do it in RNBO or Pure Data but I'd prefer a traditional code based solution. I'd also like to avoid writing native code if possible...

[raw ridge]

You could use an audio mixer and a custom DSP using OnAudioFilterRead and write your own filter logic. It’s quite minimal. I don’t think there’s a way to do it without writing some code though.

https://docs.unity3d.com/6000.0/Documentation/ScriptReference/MonoBehaviour.OnAudioFilterRead.html

[mossy jetty]

This is fine actually! It's not native code

[raw ridge]

Oh, I missed the word native lol. Glad that works though

This will give you low level control over the audio system though which is a lot easier and more direct than using FMOD

[mossy jetty]

Yessss thank you! Kinda shocked it's this easy to break into the DSP chain

[silver pollen]

What is the best practice for implementing audio? Should each audio-emitting entity have its own "audio controller" script component, or should it all be centralised into an "audio manager" gameobject?
(or both)

[raw ridge]

Try using the same audio source if you can for objects and have a culling system for looping spatial audios that culls the component so the audio isn’t running in memory when in an inaudible distance, which consumes CPU

[proper bone]

How do you turn all current audio off, and then play new audio.

Say a voiceline is being said, and you reach a new area, a new voice line is said.

How do I turn the old one off, so the new voiceline doesn't overlap the old one.

Thanks

[raw ridge]

you could use two audio sources and disable the old voice line or you can run audioSource.Stop() before playing the new one

[proper bone]

How do I do a command which stops all audio

@raw ridge ❤️

This code doesn't work for me, the voices still play and overlap

Ignore the bracket I deleted that XD

[raw ridge]

All audio? You can just disable the listener

[pale timber]

The unsatisfying answer is: it depends

In most cases I’d say both. You have the audio manager handle global things while the individual things can handle their own audio behaviors

Basically the larger in scope the project is, the less centralized you’ll want the audio behavior to be, as with most other code.

The other point to remember is that any positional sound will need it’s own emitter (AudioSource), but how you handle that is up to you. A common method is have a global audio manager which can provide a prefab instance with an AudioSource. The individual behaviors then request an emitter instance and pass in the configuration they want (clip, pitch, volume, bus, etc.) and the instance gets attached to the requesting object (or placed at a specified point in the world) and either destroys when it’s done (okay) or returns to a pool (better).

[pale timber]

On the other hand if your game is simple enough that it doesn’t really benefit from positional sounds, you could do everything through the audiomanager and just have it cycle through a collection AudioSources on itself so there can be multiple voices but everything plays from one place.

[tame jungle]

What happens if I try to import FMOD to an existing project... with hundreds of audio sources?

Is there a way to **not **have to redo all the audiosources?

[pale timber]

Write a script to automate it. That’s about all you can do other than doing it manually.

[proper bone]

I need a way that when the charecter takes damage it plays a random voice line -

Idk the best way to ask for this except ask really nicely of anyone can call and go through it with me, and I'll show u what I got and what I wanna try and do.

I'd rlly appreciate some help and idk if this is more of #archived-code-general thing.

Please let me know

[wooden hill]

While I'm yet to do audio in a Unity project, here's how I'd go about this in any game project:

I would separate logic for playing sounds so that I can call it from wherever

It should have a method implemented that would take in the sound to play and the position where it should be "rendered" if applicable

This is where it would make instances of AudioSource

[wooden hill]

Does positional need much more extra than the position of where the sound would originate in?

[wooden hill]

Have the random voicelines all predefined in a list or array and then just generate a random number between 0 (start index) and list/array length - 1 (since index is 0 we want to subtract the 1 to avoid an outofrange exception)

Then simply when triggering the sound to play you get that random number and get the element with index of random number from that list/array of voicelines

It should be pretty simple

[proper bone]

I've done it now, thanks for your assistance

[wooden hill]

Nice

[pale timber]

Whether positional or not you’ll usually want to pass in at least volume and pitch and have some sort of way to slightly randomize pitch and ideally also randomize clips (select from a collection of similar assets so it sounds more random and natural)

But yes you potentially could pass in a lot of data

[wooden hill]

Yeah of course

[pale timber]

There’s also less obvious things like reverb send level and volume roll off

[wooden hill]

One could use something like this but along with clip identifiers there could be various weights for parameter randomization

[solar fox]

Hey yall, I need some help regarding implementing player footstep sounds

So I'm currently going with a method that just plays a loop of a long sound clip for footsteps (which of course can be easily stopped when needed)

However I can't figure out how to make this sound good, I'm thinking i should crossfade the sounds so it goes from walking > running or vice versa more smoothly. (Currently, the sounds transition instantly, which causes some unpleasant 'pop' sound on each transition)

Implementing this crossfade though is another story, I can't figure out how to make it work for me - I'm using a soundmanager system, so I call the soundmanager functions on the object that I want to play the sounds on.

What's a good way to handle this situation? I have a soundmanager script, and a functioning audiomixer

I've tried looking it up, but a lot of the scripts seem to be overly complicated (and cause more issues), or it's some other kind of implementation that isn't intended for crossfading

[pale timber]

The unfortunate reality is that using Unity’s native audio means any AudioSource can only own one clip at a time. Therefore to do any cross fade you need to fade one source out at the same time another fades in. A common approach for this is to make a cross fade coroutine.

That said though with your current setup, you would want the cross-fade to be extremely short, basically just to prevent the pop. Otherwise you’ll get weird flubbing from overlapping the different loops that’ll make it sound like your player has more than two legs during the cross fade

It’d be better to loop via retrigger on single footsteps and select them randomly from an array, rather than running a long loop of footsteps. Then you don’t need to worry about the weirdness caused by a crossfade between different speeds. Instead just increase the trigger rate based on speed.

And also then you more effectively reduce sense of repetition. With running one long loop if the player runs in a straight line long enough they’ll hear the loop… unless of course it’s ridiculously long like 1 minute +

[solar fox]

Ohh aha i see, i wanted to avoid that because my teacher mentioned that it could become grating, but randomly selected sounds would prevent that in a way

And i could rework the cinemachine camera I have, as it currently works with a continuous loop of noise, simulating footsteps

Could have the same functionality in the cinemachine camera with the impulse function, thanks!

Just to clarify, what i mean by this is:

My current implementation, I have the camera doing a bobbing motion using a continuous noise function

• so I thought of implementing the footstep sounds in a similar way for consistency (so looping a long sound)

But with the sound retriggers, I could make the camera also "bob" with each footstep sound effect, instead of the long loop method - which would make the camera and footsteps align easier as well

[pale timber]

Ahh gotcha

[solar fox]

Many thanks though, I'm very curious how this will turn out

[tame jungle]

does anybody know how to adjust an entire FMOD's bus's pitch via unity?

[tame jungle]

anybody...?

[whole pewter]

you obtain masterchannel group of system and set pitch on it
that is, supposedly , for FMOD being used directly since that is not how it's done in Unity itself and we don't have crystal ball to read your mind about what you're doing

[tame jungle]

Slightly rude for some reason. But if that's how I can adjust the global pitch in a unity script then yeah that's what I need.

[whole pewter]

Slightly rude for some reason
no

[pale timber]

Couldn’t you also just add a parameter to the Pitch property of the bus in FMOD, then set that parameter value in Unity?

[tame jungle]

Their idea worked, with some tweaking. I don't know if your idea would necessarily be global tho.

[pale timber]

Parameters can be global or local, and busses are global, no?

[tame jungle]

I mean I'm VERY new to FMOD, so I'm probably wrong lol

[pale timber]

Dunno what you’re trying to achieve but the benefit of using a parameter is it simplifies things in Unity

Like in FMOD you can give it a curve and interpolation without needing any math

[tame jungle]

I was using it to adjust all audio pitch to the timescale

[pale timber]

Most things in FMOD can be controlled/modulated with parameters

[tame jungle]

This code worked btw: just use masterGroup.setPitch

had to find this online lol

[nimble bane]

I'm searching for ak, m4, fn fal, pistol, knife audio files. Any idea where i can find some quality ones?

Can i have your opinion on how this sounds?

[pale timber]

First one sounds okay but it sounds suppressed and it sounds like you’re just repeating the exact same audio clip and without any volume or pitch variation

Ah looks like it is suppressed. That one sounds alright then. Ideally you should have some slight variation though to avoid typewriter effect

Second one (unsupressed) is kinda shit

Sounds like a recording from inside a firing range

[nimble bane]

Okay thanks for the feedback, yeah i need a better soundtrack for all of my weapons. But i can't find a good website for it.

[pale timber]

For a shooter game you should use a layered approach for gunshots

That way you get a lot of variation so the sounds don’t get stale

https://youtu.be/_J56n496u6k?si=QlDwKp9QJHKOihLb

https://youtu.be/-Cw36iIGfns?si=gdXDB9pHFJD9mSZY

In shooters the player will hear gunshots more than any other sounds in the game, so you should put in extra detail and effort on those sounds so they don’t get tiring and hopefully are satisfying

The typical approach (like you’ll see in those videos) is to have things layered out as:

• Transient/body (midrange, tight, characteristic of weapon)
• Punch (bass/sub-bass range, very tight and quick, more “felt” than heard)
• Mech (sounds of mechanisms moving on the weapon when firing)
• Tail (delayed long reflection tail following the shot, changes depending on environment)
• Bullet/shell casings hitting the floor (changes depending on surface)

When the player fires a round, all of those get layered together to make one “sound”

Within each layer you have several audio clip variations and also can subtly randomize pitch and volume to get even more variation

When you do that then the amount of variation grows exponentially, so it masks any sense of repetition

But for the gunshot sounds coming from enemies you can reduce the layers and detail

And also should change the clips being used, especially as they change distance so you can simulate the way gunshots sound very different when far away vs. close.

In general for good sound libraries you should check out https://www.asoundeffect.com/

[nimble bane]

Nice thanks 👍

I'm going to work now, will watch all of it after the shift. Thanks, i send you a dm too. @pale timber

[burnt thunder]

With an Audio Source Component, if I play a oneshot clip on that audio source, does the clip that plays respect 3D Sound Settings (specifically volume rolloff based on distance)

[burnt thunder]

i did an experiement. and it seems it does.

[tardy pawn]

is there something that im missing about max distance? it seems that even when my listener is beyond the max distance, i can still hear the audio playing.

i tried gradually lowering the max distance value but it had no effect on the percieved volume of the audio that was playing

all other audio source component settings seem to work as expected, though. the only things that don't seem to be working properly are min and max distances

[raw ridge]

It’s because you are using a logarithmic rolloff which has a very long decay. If you want audio to fade when the listener is farther than the max distance i’d use linear rolloff

[quasi hound]

The logarithmic curve is predefined, so changing min or max cuts off the curve at whichever value rather than resizing it to the new range
You can manually change the curve points to keep it similar to a logarithmic curve while ensuring specific values at start and end

[cedar galleon]

"FSBTool ERROR: Encoder for chosen format is not supported on this platform."
"im having this error while trying to import a sound, please help

[cedar galleon]

I'm on 2017.1.2p3, building for the Wii U platform

[cunning skiff]

Is there any reason a car models audio file being choppy for when you give it gas? I’m not sure if it’s a script, or audio issues or what

[steep raft]

my audio isn't quite working, the audio isn't playing after i implemented the out of breath stuff, i attached the inspector and the script, can someone lmk what's wrong?

i have already made sure the audio sources are properly linked and have a good sound file playing

[raw ridge]

show the PlayAndFadeOutStaminaSound() method

[raw ridge]

I see two potential issues. you should be using one audio source for your player and use PlayOneShot instead. Play() returns one audio clip per audio source at a time

[steep raft]

oh i see

that fixed it 👍

thx

[heady orbit]

https://www.youtube.com/watch?v=xu7SsP0znm4

Recently made this soundtrack for a short film contest, it's supposed to be dark, ambient, suspenseful and anxious.
Also, I wonder if the visual is the right fit for the music. I am not good at artwork. What do you think?

[pale timber]

IMO the mini piano doesn’t work well. Makes me think of things like this:

https://youtu.be/A3tOXBx-OVg?si=KV3eJut51u04PFa7

Otherwise yeah it reads as “something bad might happen soon”

[heady orbit]

I see your point. Using a minipiano/toy can evoke a sense of childish feel.
Also, I did not know what instrument to use to fill the midrange and the highs in a suspenseful context. What would you use ? So I may think of it for the future

[pale timber]

I’m not really a composer. I would just keep trying different samples/VSTi plugins until I found something cool. The mini piano might even work if it was more washed out with DSP — e.g. reverb/LPF/wobbly pitch, etc.

It’s just the (mostly) clean mini piano has a strong emotional tie to child play. Maybe the game is a horror game that somehow involves that theme and in that case it would fit fine.

[heady orbit]

Ah I see, I will think about it next time I will write music in that genre, thank you!

[cunning skiff]

Hello! I have a go kart, when I start driving it plays an engine sound but it’s very stuttery and sounds bad quality , how can I fix that?

[whole pewter]

this code is good quality, though advanced and rather invasive esp. when used as an ad-hoc solution..
you should look into how things are being done in FMOD studio/projects and be using that as OwenCooper mentioned - that is if you need to be using FMOD Studio at all for some stuff it is good at, like dynamic audio changes at runtime - otherwise Unity audio is not as capable, but it works often well (and it OOTB)

[tame jungle]

I did originally look at the FMOD studio and project but nothing had a **global **pitch as far as I could tell. Also I'm just tryna learn FMOD in general, first time use, I'll get better lol

[pale timber]

Need to see code to help

[stuck frost]

    public void UpdateVolume(float amount) {
        musicVolume = amount;
        musicMixer.SetFloat(StringConstants.VOLUME, Mathf.Log10(amount) * 20f);
    }   

Guys I have this function that runs for OnSceneLoaded. However, on the first scene, it doesn't change the volume of the audio mixer? Only when I change to another scene

[dark atlas]

oh you didn't say anything about OnSceneLoaded before. did you check to see if that event is triggered in general? and shouldn't this be updated whenever volume is changed rather than on scene loaded anyways?

[whole pewter]

depends when you are doing it, this might be related : https://discussions.unity.com/t/audiomixer-setfloat-doesnt-work-on-awake/579429

[cunning skiff]

hello everyone! im trying to smooth out this audio, sometimes it is ear piercing like when trying to apply the gas again when driving it sometimes gives a loud screetch and the audio is choppy.

[cedar pagoda]

Well, test it out at different values and see what combo of values produces that effect. Then just adjust the max values for this not to happen. Assuming this is just the parameters issue.

[cunning skiff]

thank you very much again @cedar pagoda

for anyone else, ive actually added in an idle and running sound and then lowered the max engine sound

and its phenominally better.

[autumn totem]

Guys, I need help. What would cause an AudioSource's time to get stuck at 0? The clip plays the first time, but when trying to play it again, it gets stuck at 0. I can see isPlaying becomes true, after calling Play() again, but time remains at 0 during Update() test.

Nevermind, I found the problem. My script was setting pitch to a defaultPitch property in OnDisable() but missing defaultPitch = pitch; in Awake().

[quasi hound]

No
The files you import should be uncompressed (or losslessly compressed) because Unity's internal formats don't support their compression methods
Any compression artifacts you introduce by compression will be preserved and take space just as much as the original clear audio

[static gorge]

i'll compose for free 🐢

[viscid fjord]

A bit of an odd question, but is there a less tedious way of handling long audio files? I have voice lines in my game where the actor recorded several takes in 1 file with a bit of "air" between some lines, so ive setup code to play at certain timestamps in the audio files, the tedious part is going through and marking each start/end timestamp for every take of every line, I was thinking I could maybe write code that will look for the "air" moments but some lines intentionally have emotional pauses in it or the actor put their own style on it where the pauses are inconsistent (but sound good) - is the only real option here to go through the tedious process of timestamping every part of these files or is there maybe some faster way I could do this?

What I mean with the "air" is for example, the first 2 waveforms on the left side of the white line are 2 different takes with "air" - and the 2 wavefroms on the other side are 1 line with a intentional pause - the long "air" in the middle is actually just air, with the other 4 smaller ones being 2 different takes with intentional pauses - so I cant think of a way I could tell code "look for air but not intentional pauses" to automate this process a bit and I dont really have any other ideas

[pale timber]

Why not trim them and make them into separate assets?

Are the “several takes” literally different takes? Like different performances of the same line? Or are they different lines?

I guess either way, trimming these up into different assets in a DAW would probably be quicker than trying to timestamp everything. Also would be far less brittle if you decide to replace assets.

Like let’s say you do another VO session for Exerts. Well, your new exerts assets have totally different time stamps than the ones you painstakingly made for the previous iteration, unless you place them into the same spots in the long file

Whereas if you had Grizzly_Exert_01, _02, _03, etc. you can just render out the new assets and replace them and change nothing over in Unity (other than those assets being replaced)

Additionally then with this approach if you decide you want to exclude one of the takes, you just remove it from whatever array is storing the clips, rather than needing to know the timestamps of the particular take you want to remove

As for automating where to trim it, any decent audio-editing software will have a tool for doing a “dynamic split”, where you can set parameters for the signal to detect whether or not to place a split on the audio region, so this also would remove the need for you trying to code this functionality yourself.

You’ll still need to manually confirm and correct the splits, but it can take care of a majority of the work.

[viscid fjord]

Thanks for the detailed response - the several takes are both different performances for the same line and different lines related to the same game event (ex: "group up", "over here", "on me", and however many takes they decide to record for those 3 lines, sometimes they only do 1 or 2 takes, sometimes they might do 7+ takes)

I was considering chopping them in Audacity (as thats the only DAW I have), but I feel like it would be equally as tedious and time consuming going through the waveforms to chop, name and export each part myself - ill look into "dynamic splitting" though, maybe thatll help for this project - we are getting ready to release this month so we are kinda in "pencils down" mode, though you mentioned a lot of good points, and I think for our next project with VOs it would probably be better having them split and name each take for all their lines when they record and maybe send everything as a large zip

[pale timber]

If really you need to stick to long files with time codes, you might look into something like this https://discussions.unity.com/t/solved-how-to-read-wav-markers-in-unity/784051/10

So you could embed the markers in a DAW and then read them in Unity

[quasi hound]

!collab

[karmic bone]

:loudspeaker: Collaborating and Job Posting

We do not accept job or collab posts on Discord.
Please, use Discussions to promote yourself as job-seeking, advertise commercial job offers, or look for non-commercial projects to participate in:
• Collaboration & Jobs

[quasi hound]

Audacity has "automatic labeling" by sound volume
You can then export them as separate clips with "export multiple" option
There's no system that can tell the difference between pause between takes and an artistic pause, but creating and modifying labels by hand takes about a second per take because you only need to click and drag basically

Trying to play raw recordings by timeskipping sounds like a nightmare

It's inviting playback problems, and file storage problems since you'd be storing all the background noise between takes, not just the takes

Coding a system that analyzes audio does not sound like a good use of limited time, even if such a tool didn't already exist

But then you'd also have to code your own meta data for audio labels, and name and tweak those labels by hand using a text editor I'd assume

Even if Audacity didn't already have the exact kind of tools you need, designing, developing and debugging such tools would not be guaranteed to be faster than copypasting the clips out of the recordings

Raw recordings usually need to be mastered and cleaned up anyway, the splitting part is not the biggest timesink

[pale timber]

Yeah I’m of the same opinion that trying to make the existing timecode setup robust would take just as long or longer and would still have more cons than just splitting up the assets.

[pale timber]

From a quick search it looks like the dynamic split option in Audacity is “Beat Finder”

https://youtu.be/zqgaMUsEoP8?feature=shared

Dunno about the next step of actually splitting

But also Audacity is to audio what MS paint is to images. You should use something better

Reaper has a perpetual trial license as a fully-featured and highly customizable DAW. The only downside to the trial license is seeing a pop up asking you to buy a paid license every time you start the software

And even a paid license is very cheap

I paid $60 for it once, 15 years ago, and my license is still active for the most recent versions

But even if you did it by hand it is still better than relying on manually-entered time codes.

[heady orbit]

https://youtu.be/nofVws6L70U

I just learned this pitchshifting technique I have overlooked (or didn't know).
Take one sound, pitchshift it, negatively in this case, and it gives you different wonderful results
It might be useful for those who didn't know as myself

[quasi hound]

It's "automatic labeling" as mentioned
Under "analysis" where you also find beat finder
"Labels" are a data track that runs parallel with an audio track and contains labels in the form of start, stop and name
You can select a part of an audio track and hit ctrl + b to create a label for that range
Using automatic labeling automatically creates a label track, which you can then modify as needed

[viscid fjord]

Good points - at the time, this approach was the simplest for both the VOs and editing 1 file with filters sounded easier than applying filter settings to multiple files, and didnt think timestamping would be much of a roadblock, fast-forward a year later and maybe that wasnt the best idea lol - though the screenshot above is a editor I made (roughly over 2 days of work, mostly doing Editor API research), my strongest skills are in programming so I figured coding a "quick and dirty" solution might be fastest for me, but I think Audacitys "automatic labelling" approach sounds like itll be my best bet for this project with minimal manual work, so ill give that a shot - thanks again to both you and Owen for all the insight

[viscid fjord]

I certainly learned a lot from working on this project and with hindsight, if I had the time to basically nuke it and start over, I totally would (theres a lot, not just time splitting, that "if I knew then what I know now", would have done things differently lol) - and thats a good comparison with Audacity, an audio engineer had told me "Audacity is for destructive audio" as in, when you apply filters, it modifies the waveform (destructively), where most DAWs treat filters like Photoshop treats layers, I just dont have the money or experience with DAWs, though I might look into Reaper, it sounds like a good suggestion

[quasi hound]

Also, even if Audacity falls short in professional use it's plenty capable
Most people don't realise how many tools it has

[viscid fjord]

It does have a lot for free, which is very nice

[quasi hound]

No
Unity cannot play typical audio file formats, they always must be decoded, then encoded into internal formats
Just the same way image formats must be converted into textures

What is the absolute need?

There may be other solutions that serve the same goal, whichever it is

[quasi hound]

You can specify a compressed audio format for an audio clip in unity and lower the quality, which is what you're doing anyway on the outside of it
https://docs.unity3d.com/Manual/class-AudioClip.html

I guess then it's worth a try to get the file externally
It'll still be decoded and re-encoded so you lose the compression ultimately, but you have the benefit of being able to unload them too

WebGL is different and awkward in a lot of ways, I don't have any technical insight on its audio

@stuck frost I guess it supports Resources.Load which would be simpler than WebRequests?

[meager storm]

@quasi hound is correct . You must import full resolution source files ( .aiff or .wav ) into Unity first. Then you can use the Import settings to make the asset (clip) take as little RAM as possible. See this example the compressed file is 1.6% of the original size.

Use this method if you are NOT streaming. With streaming you can get much higher quality because you're not loading anything up front.

Here's a streaming example

You still have to include the music in your build. As another option I'm pretty sure there are asset store solutions for streaming music in from a URL so you don't have to include it in your build.

[whole pewter]

the UnityWebRequest is for 'streaming from an URL'

[whole pewter]

depending in its current non/buggy state of its 'streaming' flag for AudioClips, that is (it used to vary between versions, often)

but i'm not sure about its usefullness here, since it's WebGL --- might be as well as useless for this

compared to other asset types audio is actually rather well supported as 'external asset' at runtime, also because the mp3 spec is freely available also on standalones now, (not only on mobiles like previously...)
proper integration w/ AudioSources/Unity audio takes bit of effort - for asset preprocessing/fixed sets of AudioClips, preparation in Editor is better, for sure

[quasi hound]

What do you think of the plan to try to save as much disk space as possible by loading in an externally compressed audio file, in this case when the tracks are very long and quality isn't the priority (nor RAM)

In my test I can compress a 170 MB raw audio into 3.5 MB with Unity's heaviest vorbis compression
3.0 MB with quality 0 ogg in audacity
1.0 MB with 48-85 kbs mp3

So if there's a lot of those tracks and you need only one in memory at a time, it seems to be reasonable to import them at runtime

Streaming audio directly from the server that also hosts the webgl application would likely be a much better solution though

[quasi hound]

Mp3 may have a better compression ratio than that 2.06%, but it also looks like decoding it leads to some pretty bad audio artifacts
Better be a pretty bad radio for it to be worth it

[pale timber]

Also mp3 notoriously do not seamlessly loop

So if they’re meant to loop, then ogg should be preferred

In theory it sounds like you could have the files in the StreamingAssets folder, compressed externally however is needed and load them in as needed at runtime. I know I’ve done that with video on WebGL.

Haven’t tried it with audio clips and audio sources though

[whole pewter]

if VideoPlayer does indeed work with filesystem like StreamingAssets folder in WebGL then AudioSources will work / it's almost the same media handler of UWR and VideoPlayer needs audio decoding almost always /

[kind holly]

can unity use vsts?

[whole pewter]

no
btw can unreal/does something exist for it which can ?

[ivory tusk]

Getting an error on WebGL build:

Trying to get length of sound which is not loaded

I have preload checkbox checked in the clip properties.

And I play it with:

  AudioSource.clip = MyClip;
  AudioSource.Play();

The clip is chosen in the editor.

The audio plays but I think it throws on WebGL build and the function after that doesn't run.

[whole pewter]

i think it might be decoder failing - try uncompressed audio to see if everything works first

[kindred thicket]

It depends.
Actually seamless loop is supported in LAME Project encoder, it adds additional metadata for sample-accurate playback (MP3_ACCURATE_LENGTH, ENC_DELAY, ENC_PADDING). But other legacy decoders (non-LAME) will ignore this metadata and play with gaps as a regular MP3 file.

[kindred thicket]

wait, but you researched VST support?
#🔊┃audio message

[kindred thicket]

btw, web-browsers support this LAME metadata (at least Chrome and Firefox), this can be useful in Unity WebGL for seamless looping.

REAPER DAW uses a LAME codec, so it can export and import MP3 files with this metadata.

[whole pewter]

with what FMOD provides only, but it was hit and miss - i managed to load some VST2 plugins, but overall compatibility and stability was not good
this needs something more robust and stable, and VST3 support at least (FMOD doesn't - or didn't at the time but i doubt anything changed - support anything but v2)

[whole pewter]

(chances of revisiting this are not very high so i think this is left for the next generation of researchers for the moment .))

[pale timber]

How do you get the mp3 to not have the pop when looping in Unity then?

I use Reaper and render mp3 with LAME. I’m pretty sure last I checked it still has the pop

[kindred thicket]

Sorry I'm confused, what FMOD are you talking about?) FMOD as Unity's internal audio-backend or external FMOD Studio?

[whole pewter]

[https://fmod.com/docs/2.03/api/core-guide.html] it can load a VST(2) as "" plugin, this is part of Studio ( which it itself uses it )

unity fmod is heavily embedded in engine i don't think there's a way to access its normal API easily

[quasi hound]

This server is not for self promotion
!collab

[karmic bone]

:loudspeaker: Collaborating and Job Posting

We do not accept job or collab posts on Discord.
Please, use Discussions to promote yourself as job-seeking, advertise commercial job offers, or look for non-commercial projects to participate in:
• Collaboration & Jobs

[limpid shell]

oh sry

[ivory tusk]

How to properly play and wait for the audio clip to finish playing?

Like when it's in WebGL when I do AudioSource.Play(); clip might not even be loaded yet.

And if I try to check AudioSource.clip.isPlaying, it throws an error
Trying to get length of sound which is not loaded
because clip isn't even downloaded yet.
So what is the proper way of playing the sounds and waiting for it to finish?

[pale timber]

Sounds like you need a callback for when the asset is loaded, not when the AudioSource is done playing

How are you loading the asset?

[ivory tusk]

It's a list in a MonoBehavior:

public List<AudioClip> AudioClips;

And then I just play them by name:

    public void Play(string ClipName) {
        if(AudioClipDictionary.ContainsKey(ClipName)) {
            AudioClip AudioClip = AudioClipDictionary[ClipName];
            if(AudioClip.loadState != AudioDataLoadState.Loaded) {
                Debug.Log($"Trying to play not loaded audio clip {ClipName}");
                return;
            }
            AudioSource.clip = AudioClip;
            AudioSource.Play();
        }
        else
            Debug.Log($"Trying to play non-existing audio clip {ClipName}");
    }

Well I also pre-load them on Awake:

        foreach(AudioClip AudioClip in AudioClips) {
            AudioClip.LoadAudioData();
        }

[pale timber]

Add a null check to your function

I assume you’d get a NRE on the line checking loadState is the clip is null, but I don’t know exactly how that function works.

But before you check loadState, you should make sure the clip isn’t null

And for sanity after you confirm it’s not null you can print info about the clip retrieved

This is all just to make sure that you even have a clip being assigned to your AudioSource. Because your current error seems to be saying that there’s no clip

Also you should not name your local AudioClip variable “AudioClip” because then it’s unclear whether you’re using static functions/properties from the class or functions/properties on the instance.

[kindred thicket]

Unity Editor/Standalone Platform can't play LAME-encoded mp3s without gaps because it uses internal FMOD backend that doesn't support LAME extra metadata.
However, seamless looping is technically possible in Web(WebGL) Platform, where Unity uses Web Audio API as audio backend (but still requires LAME support from browser-side).

This is example of loading local mp3-file from StreamingAssets folder

void Start()
{
    StartCoroutine(GetAndPlayAudioClip());
}
IEnumerator GetAndPlayAudioClip()
{
    var url = Application.streamingAssetsPath + "/FILENAME.mp3";
    using UnityWebRequest www = UnityWebRequestMultimedia.GetAudioClip(url, AudioType.MPEG);
    yield return www.SendWebRequest();

    if (www.result == UnityWebRequest.Result.ConnectionError)
    {
        Debug.Log(www.error);
    }
    else
    {
        audioClip = DownloadHandlerAudioClip.GetContent(www);
        audioSource.clip = audioClip;
        audioSource.loop = true;
        audioSource.Play();
    }
}

[kindred thicket]

I seem to have missed disclaimer - this is a technical possibility, not a call to use this format.

For the WebGL platform, I recommend using Vorbis(OGG) format, it is finally supported by all modern browsers, including Safari.

[rich mirage]

How does everyone handle volume levels per audio asset?
When sourcing audio from various sources it's difficult to get them to be the right levels.
Do you hard set levels from the bounce via Audacity or Logic etc? I feel like with the iterative nature of tuning levels it's not a great approach.

[kindred thicket]

Unfortunately, Unity 6.x does not yet allow you to choose a compression format other than AAC, but you can use a same solution with StreamingAssets folder for Vorbis.

private AudioType audioType;

void Start()
{
    if (Application.platform == RuntimePlatform.WebGLPlayer)
        audioType = AudioType.AUDIOQUEUE;
    else
        audioType = AudioType.OGGVORBIS;
    audioSource = GetComponent<AudioSource>();
    StartCoroutine(GetAndPlayAudioClip());
}

IEnumerator GetAndPlayAudioClip()
{
    var url = Application.streamingAssetsPath + "/FILENAME.ogg";
    using UnityWebRequest www = UnityWebRequestMultimedia.GetAudioClip(url, audioType);
    yield return www.SendWebRequest();

    if (www.result == UnityWebRequest.Result.ConnectionError)
    {
        Debug.Log(www.error);
    }
    else
    {
        audioSource.clip = DownloadHandlerAudioClip.GetContent(www);
        audioSource.loop = true;
        audioSource.Play();
    }
}

[kindred thicket]

From a technical point of view, normalized audio source (I mean lossless WAV-file) will give higher quality when converted to a lossy codec than silent audio source. Volume levels should be adjusted in game engine or audio middleware, not in the audio asset.

[quasi hound]

Audio should be mastered for the correct perceptual loudness after recording / during authoring, before it's exported into an engine / into a format where the data for doing that accurately is lost

In engine there's no limit to how much quieter you can make a sound, but you can't make it any louder than the original loudness without causing issues

Also better for quality to use all the available headroom

[rich mirage]

From what you and @kindred thicket are describing it's both. You just said 'during authoring', but using all the headroom is best quality & mixing during runtime.

Currently I've made an SO that handles mixing levels with a simple float 0-1 and an audio clip. One of the sr devs said the 'correct' way is to master them all to be the desired levels. However we're sourcing these assets from anywhere and everywhere, and there are hundreds of them. I've come into this project relatively late and the backlog of trying to retroactively master all of these assets seems like a waste of time compared to the method I've implemented, but they absolutely need some mixing.
The other problem is it's an iterative & destructive process - reducing the volume of an audio file means there needs to be version control, saving the original full volume version incase the levels ever need to be raised again.

Also a simple normalization does not always mean equal loudness across assets, it just means the peaks are set to whatever db you normalized.

[pale timber]

This is an interesting point that I hadn’t really considered before, and I would definitely agree if implementing directly into Unity that normalized is better since you can’t easily add gain to audio assets. However I’ve heard the case made commonly from industry professionals lately at least when using middleware that it’s good practice to render assets at their intended loudness (LUFS or RMS reading, not peak) so that the amount of gain correction in middleware is minimal and the game can be closer to an appropriate mix faster because of it.

That is something I’ve been trying lately, as opposed to in the past just rendering at a normalized peak (e.g. -1 dB) and subtracting gain as needed in middleware. But the point about quieter assets likely having a greater amount of artifacts after conversion is one worth considering

In Wwise or FMOD it’s of course very easy to add gain to a signal and without any noticeable artifacting as a result, but yeah if the quieter sound is compressed and becomes more lossy than if it were rendered louder, then that gain boost will also boost those artifacts

[quasi hound]

So then is mixing for desired loudness, even if quiet, better than mixing for maximum lossless loudness and reducing volume at runtime?

[rich mirage]

That seems to be the $$$ question.
I can't say I know the best answer, but I'm implementing the latter - good quality mix/master full volume level assets that get mixed at runtime, with a default level modifier set by a scriptable object.
Making sure I'm not completely making a fool of myself by doing so 🙂

[gilded stream]

Did anyone ever encounter a problem like this? Audio being really really choppy progressive with time at the point where I can't even hear the sound anymore

Also here are the project settings and audio source settings

Tried a lot of things but can't figure out why this happens

[meager storm]

@rich mirage Sound designer here with over 10 years experience working in Unity native audio. Also many other engines and middleware. The "best" way to do this depends on what works for you in this particular case.

Most of the time you'll want a sound description like your SO that can pass settings down the the audio source playing the clip. The down side to this is you will have to create a SO for each clip. As motioned before you cannot turn audio source up pst 0 attenuation, so there is no making a sound louder from Unity audio source. As an option you can turn mixer groups up +20.

In the example you gave where a SR dev told you "the 'correct' way is to master them all to be the desired levels" is misleading and not good advice. You cannot know how loud an asset needs to be out of context and you cannot set the volume correctly when you are creating it. There are generally accepted guidelines audio professionals use when they deliver source. That's typically done at relative levels where ambience is quiet and explosions are loud.

Because I create all my own source I typically have used that relative approach. This works fine for me because I am constantly updating the source clips as I work , changing the content or tone or volume to suit the game better.

There is another approach which is to send all assets into the game loud then turn them down in engine. This approach works better with Unity built in audio source priority system which in games where max voices is important is a very powerful tool.

Unity audio source priority system makes quiet sounds go virtual when max voices is reached. This system doesn't function correctly if you feed it clips set to relative levels because priority looks at the volume , attenuation and mixer group to determine "Audibility" . This is undocumented but you can see it here in the profiler

Basically you have to mix the game in engine becuase you cannot control the source. Your scriptable object idea is good but this means you'll need an SO for each clip .. yikes !

If you want to take full advantage of the priority system you'll need to batch process your audio source to hit a loud peak target. I suggest something like -3 dBuFs . There are lot's of audio batch processing tools out there.

The downside to this "make everything loud before you import" approach is that auditioning clips in the Unity preview is EXTREMELY LOUD

[meager storm]

the audio in this video capture is too quiet I cannot hear the problem .. if you repost I can listen to it and see if I can help

[rich mirage]

Thank you for the very informative response.
To continue what you were saying about the realVoices I do have concerns about reaching that number as I'm planning on implementing an ambience biome of sorts that dynamically plays based on game state - terrain etc. Is it really as bad as some forums mention to increase that number? I wouldn't need to max it out by possibly double it (~64). Of course I wouldn't really know until I profile it - but testing only goes reveals so much until the actual systems are built.

I have editor code to automatically assign and name the AudioAsssetScriptableObjects so adding one to a folder & at one click of a button it's all assigned. This is also good for collections of sounds with slight variation like foot falls, weapon attacks, etc. Those are generally cut from the same cloth and get the same treatments as far as levels are concerned. One nice thing I just made is a shuffle for an asset collection that's static across instances, no random streaks of the same clip firing even when there are multiple of the same entity calling the same audio. Most importantly is the context that allows me to filter them to the correct channels and adds some security & convenience in the editor that a musical asset has to be assigned as a musical asset - you can't accidentally assign the wrong type, and the search box is less cluttered.

[rich mirage]

Yes, this is not a fun experience.

[meager storm]

Rather than increasing your max voices to 64 I suggest to keep it at 32 and learn to use the priority system that’s literally what’s it’s there for

Again it’s making voices virtual based on audibility

Priority system can work very well to solve the common problem of “too many sounds are playing”

Here are some tips

• priority ‘0’ was bugged so t use that number just use 1 instead. Set only critical long loops to this high priority . Like music or other looping sounds that cannot be stopped.

• use the default priority 128 for all medium priority this should be a very large group of sounds like most of your sfx

• for detail sounds you want to be dropped set them to low priority like 256

Consider adding priority to your scriptable object.

Is your game 3D ?

[rich mirage]

No it's 2D

[rich mirage]

That is something I have been considering. There's also some inherent priority built in that I could utilize, music for instance should not cut in and out, takes priority over ambience type audio.

Lots to consider with this new information.

[meager storm]

OK, so in a 2-D game my approach would be to use even less than 32 max voices. This is because you want the priority system shutting down sounds that are not important. I also recommend to increase the size of your third bucket the low priority 256 groupso less sounds are in the middle priority of 128.

Try 16 Max voices once you have the priorities assigned into high medium and low

[rich mirage]

As a musician I'm all about using limitations as creative guidelines. Haha, this however... I see the value in what you're saying, but it's definitely a challenge.

[autumn totem]

What causes the audioSource to add a tick sound at the end of playing the sound? This doesn't happen with my other sounds, but this one in particular (which is seemingly no different as it is just another sound from the same collection of sounds I purchased).

[raw ridge]

Could be the format or compression you’re using. I know .wav should overall have better re-playability if you are looping the audio clip but if you are already using .wav i’d try a different compression or uncompressed just to see if that resolves it. Could also be a specific sound in the audio recording too so i’d play it outside unity to see if it reproduces that issue

[autumn totem]

Okay, I solved it. It was the wav file, because it works when I switch to a different sound. So, I edited this broken one and added a 0.0001s fade-in at the start, and 0.0001s fade-out at the end. Also changed pitch from 100 to 99.

[raw ridge]

Were you using .mp3 before?

[autumn totem]

No, it is WAV.

[ornate maple]

hi, sometimes when i switch back to the main menu of my game (separate scene) no music and sound can be heard (happens in editor and build). when loading again the scene music and sound can be heard. Any idea what might cause this? Unity 2021.3.45f1 thank you very much

[raw ridge]

Either you don’t have a proper state management set up or a script is disabling your audio source/audio mixer group (if using one)

[ornate maple]

thank you very much: audio sources and my audio mixer group are active. What do you mean by state management? Is this a specific setting?

[raw ridge]

No, meaning that your audio source is destroyed when switching scenes because you aren’t persisting it across

Or the object that plays your audio source (if any) is destroyed

[ornate maple]

no everything is there. as said, this happens only sometime. most of the time the main main is loaded and everything work. I try to get better insights

[raw ridge]

Then i’d really check if they aren’t being disabled/destroyed/losing reference in the hierarchy. Run your game and keep testing with the hierarchy open

[obsidian bramble]

Hi, I have this sound and I'm trying my best to make it loop but I can't seem to remove that click sound every time it restarts. I really don't have any audio knowledge, does someone know a good way to easily loop sound because I can't find anything.

Also, why is it so hard to find looping audio online?? Don't all games need this kinda stuff? Like sounds that loop continuously? I want this so I can change the pitch with code.

[raw ridge]

You’re better off using .wav for looping clips and probably using a PCM compression. Though it still won’t be 100% perfect so you can try using a script to crossfade the volume and pitch towards the ending (lerping out) then lerp in the values as the clip starts. That should provide a smoother loop

[obsidian bramble]

Thanks I converted it to .wav and now the click is gone, I also happened to cut it in the right spot with another software so it loops better.

Sadly higher file size tho. I guess it doesn't need to be 30sec if it's looping

[raw ridge]

Yeah .wav is a lot bigger than .mp3. You should only be using it for looping audio and shorter clips

[hallow marlin]

what are some good sound libaries?

[obsidian bramble]

Pixabay has a decent selection for free!

[pale timber]

You can use .ogg It compresses to almost as small as mp3 and can seamlessly loop.

MP3s notoriously have a click at the loop point.

You also can just import to Unity as the full quality WAV and use Unity to compress it to Vorbis (ogg).

[quasi hound]

So audibility is based on variables only and ignores the waveform itself?
A lot of good tips!

[quasi hound]

Any tips for implementing stereo clips that also have distance attenuation?
It seems normally to get distance attenuation even without directionality the clip must be 3D which forces it to mono
And doing distance attenuation by controlling volume with a script feels tedious when the audio engine is already doing distance calculations and then volume variable may need to be set from multiple sources at once

[meager storm]

Spatial blend set to 3D makes a stereo source mono. You have to create your own distance attenuation for the audio source or the mixer group using parameters. Attenuation should be based on distance from listener.

Why would you set volume variable based on multiple sources .. that sounds backwards to me .. ??

[meager storm]

Yeah audibility doesn’t know how loud the data inside the clip is so that’s not part of the calculation

[pale timber]

Does full spread not make the source become stereo with SpatialBlend set to 3D? I don’t work with native Unity audio enough to know off hand, but that was my assumption.

Like, full 3D spatial blend + full spread would essentially make it have distance attenuation without positioning

[meager storm]

Spread is super weird

It’s hard to describe

Basically spread 1 inverts the clip channels

I might have to do a quick video explainer on it if I have time this week

[meager storm]

Actually here’s a great breakdown on spread but it doesn’t directly answer what happens with stereo source or.. but worth reading

https://gamedevbeginner.com/10-unity-audio-tips-that-you-wont-find-in-the-tutorials-2/#spread

Johns tutorials are great I’ve gone back to these over the years because they are so thorough

[pale timber]

Hmm interesting. I’ll have to play around with it some time.

[quasi hound]

By sources I mean anything that could want to control a variable
So the script that handles volume by distance has to accomodate another script trying to change volume of that specific audio source by a gameplay event

[meager storm]

Ok @quasi hound I see what you mean . Sorry I don’t have an easy answer for that. I don’t write code at all. One way to approach conflicts with volume is to use a mixer group volume with parameters to do the distance attenuation then your other scripts can set volume on the audio source

[kindred thicket]

All audio files located in Assets folder (except for StreamingAssets folder and tracker music formats) are upconverted to internal uncompressed format (like PCM), then again re-encoded using selected AudioClip Compression Format (Vorbis, ADPCM, PCM).
So is always better to use WAV (uncompressed PCM) files to avoid quality loss due to double conversion.

[meager storm]

This how I understand the engine to work as well

[rich mirage]

.ogg is vorbis and is no different than unity compressing wave to vorbis, no?

[quasi hound]

I believe oggs get re-compressed on import

But difficult to know how or if that actually happens

If you import a quality 100 ogg and compress as quality 100 vorbis, the compression ratio is still like 88%
If you import a very low quality ogg and compress as very low quality vorbis, the compression ratio is above 100% so the file actually gets bigger

Whether that's lossy kind of double conversion or not I can't tell

Also, importing the super low quality ogg and compressing as very high quality vorbis will get almost the same ratio as if compressing as very low quality vorbis
So at least it's smart enough to not re-compress everything naively

[kindred thicket]

You can easily check this using tools for extracting assets from Unity build. (tldr; imported and exported Vorbis files will not be the same).

[rich mirage]

Vorbis does still compress .ogg files. I am seeing this >100% in some files, though it doesn't have to do with the original quality.

[pulsar gull]

whats the best way to make music for game

[agile light]

any fmod users
how do i control the speed of a clip using a parameter i looked everywhere online i cant find

[meager storm]

Do you mean pitch ? What is speed ?

[agile light]

Depending on how fast my character moves I want the clip to loop faster to sound like he’s going faster

I couldn’t find a way to control the loop

[pale timber]

You probably don’t want to loop using a loop region if that’s the behavior you want. I’ll circle back in a bit after opening up FMOD to provide you an alternative suggestion

[agile light]

Thank you I’m a total beginner and I’ve been trying to find an alternative. I saw people mention to not loop footsteps but instead call them per x time in unity?

[pale timber]

Yes that would be better

But if you were constrained to looping in FMOD, one alternative would be to use a Scatterer instrument rather than a Multi instrument. Then place a sustain point somewhere along the instrument’s region in the timeline. You actually could still use a loop region if you like how that looks more (I prefer the sustain point for something like this). It should behave the same way. You can automate the Spawn Rate with a parameter.

You’ll just then want to make sure to set the min/max scatter distance to the minimum value. For the min/max spawn interval you’ll want to make those the same value (unless you want some trigger rate variance) set to some value which feels appropriate. If you have it all the way at minimum you’ll be able to trigger the sound way faster than you’d ever need, so you’ll just want to find a decent value.

But IMO you should retrigger it in Unity rather than doing something like this. I have a placeholder MonoBehaviour that I use for looping footsteps when setting up new Unity projects with FMOD. You can just make a simple timer and have it play the event each time the timer completes, and make the interval a public value that can be controlled by the game.

Feel free to use

[agile light]

Wow thanks for the in-depth explanation! I’m not gonna lie it’s my first day using fmod so what you just said made almost no sense to me 😂😂😂😂😂

I will do more extensive research & probably go with what you mentioned with recalling the footsteps in my scripts

(Actually the video you provided makes sense) and the script is pretty simple will def use!

[frosty meadow]

What is the best way to make all soundeffects (< 2 seconds) the same "loudness" ? I tried LUFS, Momentary LUFS, RMS, Peak volume in Audacity. None seem to do the job perfectly well.

[pale timber]

By ear

Otherwise LUFS Momentary Max is probably your best bet for automating

[gray grotto]

I'm trying to add a footstep sound to my character using a third-person controller, and I don't know why the audio plays when I walk, but then when I'm still without moving, it keeps playing. I'm using the onFootstep function in the script, and in the inspector, I've passed the MP3 audio of my footsteps, which is 1 minute long

In my game I'm using a general audio manager that handles background music and other sound effects like attacks, I've tried using that to add the walking sound effect by activating the sound in the move() function but the same thing happens to me

[eager phoenix]

Do you have something that stops the audio when the player stops moving?

[gray grotto]

No, I think.
The function is an animation event:

private void OnFootstep(AnimationEvent animationEvent)

{

        if (animationEvent.animatorClipInfo.weight > 0.5f)

        {

            if (FootstepAudioClips.Length > 0)

            {

                var index = Random.Range(0, FootstepAudioClips.Length);

                AudioSource.PlayClipAtPoint(FootstepAudioClips[index], transform.TransformPoint(_controller.center), FootstepAudioVolume);
                

            }

        }

    } 

[eager phoenix]

Shouldn't be a surprise then that it keeps playing

Typically you wouldn't have a one minute footsteps clip, you'd have a clip of one footstep and a system that makes sure there's a delay between playing the clip

[pale timber]

The best way is to have a collection of oneshot footsteps that are randomly selected and one is played each time the event fires, and like Nitku mentioned, some way to enforce a minimum duration between sounds to prevent spamming (e.g. during animation blends).

If you want to use a 1 minute clip as a loop as a sort of lazy way of implementing footsteps, you do not want it to come from an animation event. You’d want to look at velocity or animation state

With your current script you’re constantly saying “start this one minute clip every time I take a step”

[gray grotto]

Yea, its true, So the easiest thing is to find a one footstep audio, to be able to use the animation event correctly right?

I've been trying to use that audio too by looking at the speed and animation state, with a condition in Move function, but it didn't work for me either

[pale timber]

Split the one minute clip up into individual steps and put them into your AudioClip array. It’s already an array with random selection, but you don’t have the right asset(s) for that functionality

[gray grotto]

Okey, I will try this, thanks

[stark lily]

Is there another way to implement a proximity voice chat in my game rather than vivox ?

[agile light]

in FMOD I know if you want to loop a 3d sound you need a StudioEventEmitter Component but if I want to play like a button sound in 3d, cant I just add a spatializer and just Play One Shot and it works? Any downsides?

[pale timber]

The only downside is PlayOneShot doesn’t cache a reference to the event instance, which means you can’t do anything to it other than play it.

The StudioEventEmitter just also holds a reference to the event instance, so that allows you to set parameters, stop, etc.

PlayOneShot also is technically less efficient, but not in a very substantial way. What it does under the hood is it plays an instance of the event and then immediately marks it for release once it’s done

[agile light]

a lot of my sounds like levers are just like a one shot sound but i guess other more complex soundds with parameters ill need an emitter right?

[pale timber]

Yeah you need a cached reference to the event instance that you want to act on. You can also do it manually by using the RuntimeManager to create and play the instance, and at the point cache it for acting on later. The StudioEventEmitter just packages that functionality for you to make things convenient

The way I like to work most of the time is to do it manually so I can have more custom control of everything, but the StudioEventEmitter component is just fine for quick setups.

[agile light]

Okay thanks!

[pale timber]

Technically you can PlayOneShot for loops but it’s not a good idea because you don’t have any (easy) way to stop it once it’s started.

[grave nest]

Am I doing something wrong, or can you not use the Audio Random Container with PlayOneShot? If so, is there any reason why?

[frigid venture]

How do I make the assessment if I want to use FMOD in my project or not?
I know nothing about FMOD beyond that many unity games use it.
I am not developing a "music game" and ive never used FMOD before, will it be overhead/overkill for a 'regular' projects needs, or is it indispensable?
I'm lacking the data to make a judgement call and dont know if I should implement it now or try to implement it later if somehow I discover my sound needs to be robust
Will it cost me financially as well? The package is free so I assume releasing a game using it isn't

I am not a sound engineer and the idea of having to do a lot of set up work in a totally separate piece of software is daunting
I think I have talked myself out of using it, the scope of what it does seems way above my project, even if my project did go commercial audio isnt really its main focus
but I feel like I cannot make that assessment because ive never done an audio heavy proj so I am torn up inside about making the right move and not fucking myself over in the long run by making the wrong choice now

[pale timber]

Middleware makes a lot of thing easier with audio. It also makes it easier if you have other people contributing who can’t code or who can’t code very much

You can just use native audio but if you want any more complex behaviors for the audio you will need to build them yourself

[quasi hound]

Really the only way to make the assessment with certainty is to study its features compared to unity's audio and to try it
It's very project-dependent and dev-dependent as well
Though I haven't used it, as I understand the main appeal is to reduce coding that would be required for things like dynamic audio, audio zones, layering, and to speed up audio implementation in general
Iirc it has some type of DAW included too
But if you do code and don't use a lot of audio, I think you can do fine just by what Unity gives you

It's not free though, similar business model to Unity where you start paying as you gain revenue

[pale timber]

FMOD licensing is very generous though

I think it’s free below 200k revenue

Need to double check that number

https://www.fmod.com/licensing#support

Under 600k “Free or $2000” I’m not sure what condition makes it 2k

IMO there’s way more pros than there are cons to using audio middleware, but I am biased as an audio person. But I am also a competent game programmer and still feel that way about it.

[quasi hound]

Chances are you use and need audio more than average, and it makes the choice easier to be familiar with the tools already

[pale timber]

One big thing that is often overlooked is it makes debugging audio stuff way easier

And optimizing per platform

[quasi hound]

I'd assume the biggest hurdle is to have to learn and deal with a whole another system, rather than anything technical

[frigid venture]

Hmm okay thats a lot to consider.
Yeah the fact there is an external FMOD studio tool I have to learn is the biggest hurdle I see right now.
Currently I don't have many sound use cases beyond playing X sound at Y location for SFX and cross-fading music tracks

[quasi hound]

If the music tracks don't have to stay synced when crossfading, it should be a very simple thing

Syncing is the kind of challenge I'd look for middleware for

[pale timber]

Yeah anything beyond just single music file on loop -> fade to next single music file on loop is way easier with middleware.

But if that’s all you need your game to do with the music and you only need simple “play this clip at this location” for sounds then introducing middleware might not be worth it

Even then, if you want to define a custom curve for how the fade should happen, its way more of a PITA to do it with native than with middleware (where it's a built-in ready to use feature). Lots of small things like that add up.

But yeah ultimately you should try to define, maybe on paper, what functionality you think your game might need from the audio. IMO anything beyond just bare bones you should go with middleware. The time spent learning how to use FMOD is likely a lot less than the time you'd need to spend on cooking your own solutions for things the software already provides.

[iron atlas]

Anyone have any ideas with a c000005 virtual memory access privelidges error

Last in call stack is unityplayer.dll!audiomixer::getsnapshotindex(class audiomixersnapshot const &)

Fmod is in project

[wet mist]

what're some good ways, or the best way, to handle numerous of the same sound effect playing at the same time, or maybe very close together...causing amplified volume and a harsh sound?

[pale timber]

You need to enforce some sort of instance limiting — how you do that best is going to really depends on your needs

In any case you will need some condition, a way to track that condition, and a behavior for what should happen when that condition is met

E.g. three sounds of the same type cannot play simultaneously within a radius of 2 units of each other.

You could get really fancy and track how close certain sound-emitter types are together and if they’re within a certain range they share a single audio source

Alternatively you could use an AudioSource pool and enforce just a global limit (via the object managing the pool) where only a certain number of them with a particular clip are allowed to be used at once. Then define the limit behavior — I.e. should new ones steal old ones or should they be prevented until the old ones are done?

Again there’s a lot of possible ways to approach it and it’ll vary on what’s best for your project

[quasi hound]

@pale timber You have experience with audio middlewares, do you know if they have features to support those three techniques or make them easier to implement?

[quartz lantern]

FMOD and Wwise both make it very easy to limit concurrent sounds. They give you a lot of flexibility for how you want to do it too. For the other features (like pooling/clustering) you'd probably have to write the logic in Unity yourself.

[pale timber]

Yeah like Toby said, Wwise and FMOD both come with easy to use instance limiting. They do things a little differently though. FMOD only has a global limit per event, while Wwise has a global or per-object voice limit per event. The per-object limiting can be extremely useful. FMOD has a cooldown setting per event, while Wwise does not. IMO the Wwise concurrency features, like most other part of the software, scale better with larger projects than FMOD. But FMOD’s GUI is nicer to use and the software allows for a lot more creativity in how you execute things.

Because of FMOD’s lack of per-object concurrency limit I may try to make that a feature in an add-on/framework I’ve been developing.

[lucid ermine]

People here very likely understands audio better than me. Working on a voice chat system which is going quite well, and working on implementing "filters" to the audio, allowing you to modify it was wanted. Wanted my first filter to be a low pass filter, as I thought that'd be an "easy" task. Damn was I wrong.

I got a sound now that i like from it, and it's working mostly as intended, however, as the cutoff goes lower (effect gets more agressive) I start hearing a horrible crackling sound. I believe the audio visualization shows it as well as these sudden walls, but i'm not entirely sure what i'm doing that's causing this. Any input is much appreciated!

public override void Process(ref ArraySegment<float> samples, int frequency, float strength)
{
    float cutoff = Mathf.Max(_def.cutoff, 300f);
    int taps = Mathf.Clamp(frequency / Mathf.RoundToInt(cutoff), 1, 64);

    var buffer = samples.Array;
    int offset = samples.Offset;
    int count = samples.Count;

    for (int i = 0; i < count; i++)
    {
        float sample = buffer[offset + i];

        _history.Enqueue(sample);
        if (_history.Count > taps)
            _history.Dequeue();

        float sum = 0f;
        foreach (var s in _history)
            sum += s;
        float smoothed = sum / _history.Count;

        float mix = strength * 0.9f;
        buffer[offset + i] = sample * (1f - mix) + smoothed * mix;
    }

    AudioDebugWaveform.SetSamples(samples);
}

[kindred hedge]

Hello guys. Can someone tell me what is this technique called? I want to trim the songs that i like, to add my game. I tried to search triming the notes of a song but i couldn't find any answers.

[quasi hound]

Can you describe what "technique" you mean? In the video samples are played in sequence, nothing that looks like trimming a song

[kindred hedge]

so basically, i want to take a song and i want to have every note as a different .mp3 file.

[pale timber]

You will need to do that manually

[quasi hound]

That's not something Unity can do, and not something any external tool really can do well either

At least not if the sounds overlap at all

[lime crater]

Audio bug. How can I fix it?
3D Render pipeline URP
6000.0.43f1
Spacial Blend - 3D
Volume Rolloff - Linear

[zinc notch]

I'm trying to speed up a timeline that includes an audio track. I want the audio to play faster, but without increasing the pitch.
What’s the correct way to do that?

I tried routing the audio source through an audio mixer and applying a pitch shifter effect, but the output sounded really bad.

I have signal tracks that go with audio so I would like to see the waveform in timeline

[pale timber]

Sounds like it’s pitched up. Could be due to Doppler because I don’t see a pitch value higher than 1 in the Inspector in the video

Ah yeah you have Doppler set to 1 so just to rule that out try setting it to 0

[hard gulch]

Could be a sample rate mismatch

[unborn bolt]

Augh..... this is deceptive.... it does NOT play on Awake() like the label says, it actually plays in OnEnable(). Why you do this to me Unity? I just wasted time assuming I couldn't replay a sound by turning an object on and off because of how they labelled this 😛

[quasi hound]

Could report it as a bug and see if they change the label after all these years
...Or explain why it's actually meant to be that way

[whole pewter]

the label is correct
Awake is not called on disabled GameObjects (it is called on disabled component, if its GO is active though)

[unborn bolt]

Weird. How come it runs everytime an object enables too? That is not Awake behaviour

[unborn bolt]

Heh, ah doubt it would be a priority

[whole pewter]

wdym "runs" ?

not sure if weird but Awake runs when GameObject is enabled, if it was disabled at instantiation, yes

[unborn bolt]

I mean that if you toggle a GameObject's Active state from false to true, the AudioSource will play its sound, then if you toggle it to false and then true again, it plays a second time

from Unity's docs on Awake(): "Unity calls Awake only once during the lifetime of the script instance. A script's lifetime lasts until the Scene that contains it is unloaded"

So this would lead me to think that I would not be able to toggle a GameObject's Active state repeatedly in order to replay the source, yet I can

But their documentation for AudioSource.playOnAwake: "Enable this property to automatically play the audio source when the component or GameObject becomes active.". That's not how Awake() works though