#🎬︱avatarfx-streams-share
1 messages · Page 2 of 1
I have uploaded the image you posted and it has accepted it... when it detects that it is a person, three options appear to modify it (make it realistic, make it anime style or make it 3D style)
When it detects that it's someone very real, it lets you make changes...
That's strange.... I tried it a good fifty times, over several days. But it was weeks ago. Maybe they changed their copyright infringement monitoring system.
As for characters being too real, I've had it with a character. But it was a child depicted in it. I know other people had the same issue. If it tells you it's detects a real person and offers to make changes in the style, it is most likely just a problem of the characters looking real and them wanting to avoid deepfakes. 🙂
Hmmmmm. It's weird that it works for you and not for me... 😦
It doesn't give you the option to modify the image?
No. Because it detects a copyright infringement. It won't let you do a character that ressembles a Disney character, regardless of the style.
I tried to upload an image of the US president and the same thing happens, with no option to modify it... But I haven't had any problems with the princess.
Whereas the Rapunzel image is refused as copyright infringement for me... That's strange...
I've tried it with Micky Mouse and it also detects the copyright... Anyway, if you want, I'll give you an online page where you can change the style of that image a little... you can even go from a drawing to a real one.
Thank you. I'm always happy to look at the page. I have been regenerating new images. They are not bad. But I must admit that that one was exactly what I wanted. So it's annoying that it won't accept it. 😅
Options, anyway it doesn't give me problems because I use the file generated in Discord... could it be that the file you use (the original) has some protection?
I tried the file generated through Discord after I read your message and it didn't work. It is still not accepted. I wonder why it works for you and not for me.
i don't know how to feel about this
The mouth movements came out pretty good
Well, I was seeing that if it detects that it's too realistic, it lets you modify the image, but if it detects that it has copyright (Mike Mouse, President of the USA, President of China, and President of Venezuela), you have to choose another one directly. 🙂 It's funny how they've implemented security systems.
Hey, could we talk about what modifications to the image description prompt have worked for you to make the video perform actions or movements.
I‘m currently working on a „tips and tricks for avatar.fx“ if someone might be interested 🙋♀️
Yes, the voice makes sense (although it's complicated) because of the commands... but trying to get it to perform actions by creating them in the description is something that depends more on luck..
and also from the initial image, add lighting effects or background motion... but only if the image seemed to have them at first.
It's 9:45 p.m. in Spain, do you want us to do what that job would be like... if you're not busy?
@pearl swan are you around?
I apologize, but it came from another channel, it was about how to generate anime or 3D cartoon images in avatar fx... If it detects a person, and it's not a politician or something... you can generate caricatures.
The 3D one looks a lot like him
Yes, it's very hit and miss.
and you can only see the final result, and a new generation can have different results even if you don't change anything.
🤷♂️ Well, that's also the fun of generative AI.
i kinda want to do Snape just to see how he looks like on Jujutsu kaisen style
it didn't let me
i had to use his younger version
Do you really want it?
grrrr
In the end, I created a black magic witch version of Rapunzel
Yes I am busy, sorry.
I’m working on a guide for avatar.fx
You can prompt quite fine what can happen in the animations.
I love her with music
and yes I promoted the way she should move and act
I'm sending you the animation because I have no problems uploading that image.
Oh, thank you! It looks great!
Which is better, A or B? The thing is, if Captain Price is detect, is blocked, so in the end, he can't be the one singing 😭
A
B
C
No, you're welcome. 
I'd say A.
Although, in A, it looks like the guy doesn't want to be there.
In B, it looks like he doesn't know the lyrics.
C looks better with regards to character expressions, but the fact that the guy at the front is completely static until almost the end of the video looks weird.
I think John Price's character is copyrighted... I had to lower the quality a lot because as soon as it detects it you have to modify the image. 😭
I don't know who these characters are. But yes, C.ai doesn't want to get sued so they have a very protective stance.
And soon @hard umbra will do a tutorial with her experiences in Avatar Fx.... Under Pressure 😆😆😆
Not sure if this rubs me the wrong way 🤨
But you don’t have to listen to anything I say 🤷♀️
I‘m getting amazing feedback for my videos on my socials.
I just need some time to get it written, others have asked me and I gladly share with those
The movements and the air effect are difficult to achieve if you don't include them in the image description... and even then, I don't know if I'd end up doing what I wanted. I don't speak English, and I don't know what tone to use. I should do it as an action outline or do it with if-then conditions so that it makes movements organized and not haphazard.
Or if you can describe that one of the characters has a deep voice and the other a high voice to make them speak in turns, or that one is mute and the other speaks... there are many things to try.
Because at the moment, apart from the creation of the script (voices) there are not many instructions.
At the moment you can’t make two characters speak.
I managed to get two animated independently,
but the voice won’t „split“ even with an audio with two voices
I've uploaded a moving gif to see how it describes the movements... but it doesn't explain much. 🤷♂️
A young girl with short brown hair is winking and playfully pointing forward.
If you have C.ai+ you are still limited to 5 avatar fx per day (or 10)
10
ok, five more 🙂
Description of a moving gif... let it upload and it stays moving, 🤷♂️ does it interpret the animation? 🤷♂️
A character with long hair is standing and using a glowing object in a dynamic pose.
A uniformed officer is walking forward and appears engaged in a task.
These are the descriptions you make when uploading animated gifs... any conclusions on how to indicate actions in the description?
Upcoming experiments... Will a timeline script work in the description?
Don’t you edit them yourself? 👀
I was testing to see an auto-generated description of a moving image, a gif.
So, the dialogue section... where the audio is generated... is actually where it has the least influence on the movement because it only serves to generate the audio?
Nope, combination of both
From my impression 
I practically don't use it, I upload already generated audios 🤷♂️
If want you can send me one of your image prompts.
As I said, I get my results between 4-8 tries 🤷♀️
What I do is rip an audio, run it through a karaoke editor and increase the volume of the vocals and lower the instrumentals (I actually use Avatar FX to make video clips)
I talk about the image description 🤗
I don't usually save them, but I have two that I saved because I couldn't get the image to respond the way I wanted and I did them several times.
A group of animated soldiers is standing together, equipped with firearms and wearing military attire. He starts singing seriously, staring at the camera and looking tough., the backlights flicker creating a strobe effect.
The background lights flicker, creating a strobe effect; green lights project laser-like beams of light. A group of four soldiers stand, armed and in military uniform. They sing seriously; the first soldier sings without gesturing; the rest sing exaggeratedly, looking at the camera with a rude air.
I usually describe some light or smoke effect to see if the AI detects it. 😆 🤣 😆 🤣
Special effects are luck. You’ll need some already in the image.
I tried with fire and amber.
Also background characters won’t get animated, at least not in this state of the feature.
Or you have to be very explicit with your prompts.
https://character.ai/share/v/e1394ce7-c932-4ccd-860a-86ffaba593e4
The ending came out really nice. I wanted the pet to pop out of the device to hug the viewer. But instead, the viewer is the one reaching for the device.
The fire and smoke effects come out quite well.
Camping girl
The question is, is the dialogue section. Is just to create the audio and the AI acts on that audio?, or if you put notes... those notes also affect the character's movement? 🤔
Notes tipe: [[|jumping|] etc...
Finally dared to make one of my favorite Bot ❤️
Was the second try and I like it just the way it is ☺️ fits him perfectly
Questions .... 20 seconds of music... is it considered intellectual property or is it just too short?
• We run the dialogue that users write through our safety :). The 🙂 screen for content that would violate our policies.
• We use industry-leading tools to block videos from being generated using photos of minors, high-profile politicians, and other notable figures.
• For other uploads of human photos, we use AI to change the image so that it is not recognizable person.
• We apply a watermark to the videos to make even clearer that they are not real footage.
• Users must agree to a robust set of terms to use the test feature. The terms prohibit impersonation, bullying, deepfakes, and use of protected IP without permission. They also require users to agree to use AvatarFX in line with our Terms of Service, Privacy Policy, and Community Guidelines. And they articulate a strict one-strike policy for violations.
I wouldn't put notes directly in the audio because it has happened to me that the voice generator doesn't recognize my instructions and instead reads the notes inside the brackets.
Also, I ended up using Avatar FX for fun instead of...using it for character introductions. So I ended up having the characters sing 😄 😄
P.S. This can't be done with opera songs because the high, sustained notes are not identifiable as words.
The good thing about audio generation is that you can recreate it several times, unlike the video render.
Mmm... buying a holographic fan to have the sensation of a 3D FX avatar would be too much... or would it just be for the interest of technology... 🥰
Hmmm. You are confusing several principles of law. And laws will depend on your country, so there is not one single rule:
- Intellectual Property infringement is the use of the work of others without their consent. This is the principle at play when the system recognizes Rapunzel from Disney and refuses it to avoid copyright infringement. Or when you use a recent piece of music or images on which you don't have rights (stuff you find on internet is not necessarily copyright free).
- The Right to one's Image is the ability to restrict the use of your own image by others. In some countries, images of a person cannot be used without that person's consent (unless they are a public person). This is why real characters are refused or kids.
- Then there is defamation which causes damage to a person's reputation. This is why some public personas are refused.
- There is also personal data, in Europe for instance. A person's voice and image are considered to be personal data and are protected as such. So you can't go around and process them any way you'd like.
Some countries will be more strict than others with certain aspects. But C.ai has to take the most restrictive approach to ensure their system complies everywhere.
Even if you use 20 seconds of a music, you are infringing on intellectual property if the song has not fallen in the public domain. Now, will Cold Play come and sue you for using Viva La Vida in a 20 seconds clip? Most likely not as long as you don't make a commercial use of it and as long as it's for fun. However, can you get banned for it? Probably as you are infringing on the agreed terms and conditions for the use of the tool.
It's just so fun... 😭 seeing how you can create a luxury mini-musical video with no expense (in theory)....
I suppose there will be artists who use their own material (c.ai users) such as illustrations or musical compositions... They can create Vtubers of their 20-second songs. ❤️ or if they are good at editing, compose several 20-second video clips for his entire song.
That is certainly a good use for people with original content. 🙂
Not like this, ha ha ha
Rickrolled
Bonus question: What if we taught kids how to use AI instead... instead of using it as a digital babysitter? (Error 404: Logic not found.) ||😭 It's not mine, it's from a bot...||
😭 Why is it that while the guy in the back is singing, he's making faces like he doesn't like what he's hearing? 🤔
Videos with chorus don't work well since they AI doesn't yet work on multiple character. At least, not in a consistant way.
Yes, you have to enhance the lead voice
That one is really funny
How's the Avatar FX tutorial going? @ornate pebble @hard umbra .... Have there been any discussions with developers since it was launched to ask questions to developers?
You asking US if we heard anything from the dev‘s? 😅
No....
Sometimes we usually have open channels for conversations with developer to talk directly about technical things about the bots... Have you done the same with Avatar Fx that
Done what?
I think you misunderstood Kagomek's message. First, I am not writing anything. Second, Kagomek is just doing a guide graciously, during her free time to help people. It is not something official that she is doing. And she is not doing it with the developers. 🙂
😭 the mobile translator is not very good
I started, half way through.
If anyone wants some tips or has questions, my dm are open.
I’m not that active here, but I don’t bite 🤷♀️
It's a personal guide manual (not Official), I named Moony because has experience in the Avatar Fx.
- They sometimes held events with developers to explain technical concepts such as writing bots, {{char}} {{user}} tags, and long definition limits.
Have they done any interviews about Avatar Fx? ()Developer answering user questions in chat)
@ornate pebble you having secrets 👀
For example, I have never tried using sound with dialogue generation... I used dialogue generation once but I find it more fun to upload audio with voices directly.
Please don’t spread rumors 😅
@atomic pike you could you please confirm that there are no Chat channels with devs for the video feature?
Thanks and sorry for ping
Before the events with users experienced in editing bots, they got hold of the C.ai staff..
There are no direct channels with the devs. The devs did interact here when Avatarfx was new, and they were also experimenting with the feature during that time.
I am so sorry I’m also a non-native speaker …
What exactly is your question?
What do you want to know?
feedbacks are dropped in #📝︱avatarfx-feedback and are consolidated and submitted to the team..
I'm now on a PC, not a phone.
And well, this chat channel isn't very good either because of the wait times between messages. Should we continue on another channel?
At the beginning of C.ai, when more than writing development we wanted to know how many things a dialogue could remember or how much the capacity of the long definition was, sometimes events were held where you could speak directly with the team staff (it was not common), now it is done with the bot creators to find out what tricks and how to create bots with charisma.
They weren't channels, they were events... but it was a long time ago.
For a limited time you could ask questions and speak directly to the stuff. When everything was very new and people didn't know the limit was 3200 instead of 32000, or when everyone was asking for an increase in chat memory for bots when they started being used for RPGs instead of funny character replies.
I thought something similar had been done with Avatar FX.
I'm not using the service from that long. But as Ershin said you can provide your feedback or questions on the #📝︱avatarfx-feedback , maybe they'll answer your question or submit a ticket if you're facing an issue.
There's no doubt about it. Kagome was going to compile her experiences with Avatar FX to keep as a memoir for herself (some people are organized and do that, others go crazy and don't write anything down).
Let's give her time before she is ready to post her methods.
Let's not ask her again and again.
It's so we can share things and tricks about how we've been using it...
For now, Avatar FX is more a matter of trial and error.
In theory, the tool is simple: add an image, create a dialogue, and it will automatically sync with the audio.
But, as with any new development, some of us want to experiment with it to see its limits.
Here, i did another one with this. Its on my feed if someone wont believe it. But i might delete it becuase its an actual song! Its only to proof that it works
@daring matrix
What I have noticed is that the special effects (backgrounds and lights) tend to activate almost at the end of the animations.
In that animation, the continuous rain effect is activated at approximately the 10th second mark... the background shows the rain moving spectacularly. And how did you manage to keep the voice of the song from making him speak when he's silent...? 🤔 Do you process the video after creating it to add the second audio track?
Yeah I’ve noticed too with my Flame Nymphe.
As I said, special effects depend more on luck.
But you can prompt movements quite fine
Ohhh I just notice the rain falling in the background after a few second ! nice
In this video, for example, the rain looks a little better near the end of the video... as if the raindrops were being blown on, but not from the beginning.
.
Even the sky cries for him 🥲
I didn’t really put much effort in that one 🤷♀️
Just a quick proof that I’m not telling stories!
I didn’t uploaded my music videos in my c.ai feed because I don’t want copyrighted music there.
I made these videos for my socials 👌🏻
But it works
And I think some have found in their animations that the drawings that are more realistic have better movements than those that look like cartoons.
A user uploaded these two animations comparing them... (but now I'm not sure if it's because the resolution of the second one has worse movement)
This is MY Honeymae 😆😬
These are my videos 👍
I posted both to show the difference
Funfact: these two videos had the same image description and the same audio
I also have tried one - posted somewhere earlier here- with 'cheap doodles' to test that and I was surprised, it did animated it well on first try without any particular prompts. (Except those with busy background, that I agree seems to not be recommanded, at least not my messy abstract ones)
Mmm... maybe over time we'll find tricks, like making the objects we want to move stand out, apart from the description that helps identify the image, maybe kinetic line effects or using blur effects on objects in the foreground will help interpret them as movement...
In this video, a barely perceptible hand appeared and animated it...
.
The hand was in the original image but I didn't even notice until the animation showed it moving.
Fact: If an image is detected as a person, and you generate a realistic one... If you download that new image to use later, it will detect it as a person again.
Fact 2: If you like the realistic version it generates and want to use it for future animations, it will probably be re-generated with the original image.
Fact 3: If you want to recreate or recycle an old character and don't want to redo the dialogue, you can download the video, rename it to mp3, and use it as audio for the new character.
Fact 4: You can always make an AI work for you.
Make a text-to-speech program transcription using commands to emphasize the text, {example commands [|pause|] [|happy|] [|sad|] [|angry|] [|exclaim|] [|exclaim|] [|whisper|] [|serious|] [|calm|] [|loud|] [|soft|] [|dramatic|] [|fast|] [|slow|] [|shout|] [|mutter|] [|question|] [|command|] [|plead|] [|encourage|] [|quiet|]} text to be transcribed: I've seen things you people wouldn't believe. Attack ships on fire off (the) shoulder of Orion. I watched C-beams glitter in the dark near the Tannhäuser Gate. All those moments will be lost in time, like tears in rain."
Oh wow. I wish I had known that sooner!!! Thank you!
It happened so often that I ran out of credit and had to do it all again the next day when I had a perfect audio the day before.
What the actual 🤨
It’s so hard to find good actors these days 😑
It has a cinematic finish... have you ever tried making two videos with two characters sharing the background and then... making a montage as if they were having a conversation, switching scenes (montage of the two videos) between the two characters???
I myself did not.
But i would say those on the c.ai blog are made like this, montages.
I prompt it 💡
In this silly one I promoted that his expressions are not exaggerated.
I had the problem that at the end of the video he was just … too much. Just write what you expect from it
Are the images you use generated in Chatgpt with prompts?
I generate the images with ChatGPT. I write the prompts.
Once you figured out how to write them it’s not that difficult to get the animations work.
You can tell me what you write and I can have a look at it.
I usually leave the description generated by Avatar FX and only add lighting or background effects, although if it doesn't detect them, I don't think it will follow the instructions. I think hand movements and the image mood and humor also work if you include them in the image description.
I don't use voice generation much, but if I had to... I'd probably pass the following prompt to an AI to format it:
Make a text-to-speech program transcription using commands to emphasize the text, {example commands [|pause|] [|happy|] [|sad|] [|angry|] [|exclaim|] [|exclaim|] [|whisper|] [|serious|] [|calm|] [|loud|] [|soft|] [|dramatic|] [|fast|] [|slow|] [|shout|] [|mutter|] [|question|] [|command|] [|plead|] [|encourage|] [|quiet|]} text to be transcribed: I've seen things you people wouldn't believe. Attack ships on fire off (the) shoulder of Orion. I watched C-beams glitter in the dark near the Tannhäuser Gate. All those moments will be lost in time, like tears in rain."
Ah wait, you generate an audio.
Now I get it
I told you yesterday that „special effects“ can’t really get promoted and animated.
That’s more of luck.
But behaving and movements, that’s something you can influence
Fire, smoke, and lights are generally well generated in animation. and... reflections in the mirrors 🤯
I focus on the movements tho I want „bring the characters to life“. So i use most of the tokens for that
Hey guys. Just wanted to share the video I made for Tiktok (the basis is an Avatar FX video). I was rather proud of myself. Although I am not sure if I should me. 😅
It’s cool! I love this glowing frog 😍
Thank you. I would never have thought of editing my videos if it wasn't for you. 🫶
Since its detected as a human, let me make his anime version.
and its 3D drawing version
What did you do to my man? 😭
Eeeeeh... nothing
It’s okay.
The first is nice
He he, I used an uncrop AI to generate the bottom part.
He is my unreal engine style version of an anime character 
I just made him that way to fit the AU roleplay

The good thing is that... now they can make it into a sticker
Ahem. Maybe it's best to not take other people's material without asking if they are ok with it. Just, you know. She spent a lot of time on her graphics, like all of us. And she might not take it well that you are using her work to play with it. ^^
Okay nothing against you,
But could you please stop?
This bot means a lot to me and it kinda hurts
That's the thing when people invest time in something, they tend to be a bit touchy about their creations. I'm like that too. When you take a lot of time on something, you don't want anyone to alter your work. Especially when you are happy with it. 🙂
If you put THIS into the sticker contest … sorry but I’ll report!
I Better go offline, sorry
No, I'm not participating in the contest.
😭 It's just that I detect it as human and I wanted to see what his anime and 3D versions were like.
Peter Pank is a bot. He detects images as human or not human. 😭 😆
Anyway, best to try and stick to video discussions here and use the general chat for the chitchat. 🙂
Does anyone know why, even if you tell them that in the image their head is fixed, looking straight ahead at a fixed point, they always look to one side?
My new video. I had to edit it because I couldn't get a single take in which she was not dangling her hand weirdly. 🙂
Mmm... on the internet there is an artist who makes covers of Disney songs, putting the heroines in the role of villains, and makes animated music videos.
Funny. I do that too.
Lydia the Bard
Don't know her. But I use my own audio in the videos.
If you want to see his song and drawing work
For instance. I removed the vocals and used the instruments only (but please, don't use my music and videos). 😉
What do you think of Lydia the Bard's work (youtube) and her songs about Disney heroines as villains?
Where did you get that? :/
Does the song go well with this image?
I don't know. I would put someone who looks younger. 🙂
And this audio and this image match?
That one is much better.
I'll make an Avatar Fx to see how it looks.
Same issue as before. The bear is doing the talking.
Try this image
Reversing it works? I'll try it tomorrow. I'm out of attempts. I generated an army of talking bears today.
Who knows... maybe he's looking for faces from left to right?
Anyway, flipping the image is the easiest manipulation... if it works?
I had videos where the animal or person on the right was working. I will still try it, just in case.
As a quick share of 'experience' or 'data', when I first did trials with AvatarFx, I did a video to mess with a friend with 2 characters in it. I don't have any special tips how it works🤷🏻♀️ , it just happen that the human is talking and not the cat, even though the cat seems bigger and upfront ? (👌🏻 The cat with the human hand🐾 )
I've stopped trying to understand how it chooses who speaks. The helium voice got me laughing.
😭 Flipping the image doesn't work either.
Arf. I thought so... 😦
I think you need a new image with the bears snout closed 😕
Maybe meridas mouth a bit more open.
(By the way, she looks so beautiful 😍)
Although I think the only thing that works is to move the bear so that its mouth is not visible.
I made these Fx avatars without modifying the description that I automatically generate, but they look very sad.
@hard umbra 🥰 Can you think of how to make them more interesting? 🙏🏼
Kid1
Kid2
Yes
Thank you!!
I found an app online that can change the voice of a song to that of another person.
So that it's not so obvious that I'm going to use songs from artists from the 80s 😆😆😆
Give me a minute please
PLEASE before you share ask 🙏
or before one uses my images!
Please ask first!
I put a lot of work and effort in my stuff!
I gladly share but please be respectful
This is my „outtake“ or „blooper“ video, just for fun
And because I was curious if I would manage to get it right
This is my djinn …
She gave me a lot of lags, but this was the best one.
AI made very, very nice animations but sadly most lagged, guess I promoted to much in other tries.
I’ll post two cards of my guide in the general 🙏
Please let us all leave this channel here to share our videos and talk about the „How“ in the avatar.general chat.
Would be a shame if videos getting lost between text walls
Thanks
Oooh, You really use dialogue generation... I found it very complicated and in the end I upload audios... It responds well to the voices of the audio
What do you think of the actor Timothy Dalton?... If you upload images of him and choose realistic... Are the characters in your Victorian style?
😭 I think from now on I'll use the reality mode for my own personal photos... they come out much prettier than as I really am 😭 .... catfish Ai 
@tight bone😭 I think from now on I'll use the reality mode for my own personal photos... they come out much prettier than as I really am 😭 .... catfish Ai 

Evolution beauty
inhales 🗣️ @tight bone
Sorry, but could you please make a small reminder what this channel is?
Thank you 🙏
Sorry, I was bored and started testing how Avatar FX generates images.
Match this voice and this picture ??
@ornate pebble How are your talking bears going? ... I think there's a high probability that by adding that audio to that image, it will end up being a femboy talking.
I haven't gotten around to it yet. 🙂
What do you think of this other bear?
or realistic mode.... ||( 🤷♂️ D cup 🤷♂️ )||
That might work. Thanks! I'll try it when my kids are asleep. 😄
Sorry.... ||( 🤷♂️ D cup 🤷♂️ )||
The problem with the last picture is that her belly button might start talking, ha ha ha
Haha. Besides, if you look at my bots, I don't use realistic style avatars. 🙂
Ew
This is sooooo annoying. I tried it with other variations by changing the bear on Canva. It keeps triggering the bear. GRRRRRRR
-# the snout is to big and prominent … meridas „too small“
If you’ll look at it, the light color of the bears snout draws eyes to it, also Ai will detect it. Meridas mouth is way smaller and harder to recognize
I have tried making it darker and the light. Nothing works. But now, I'm trying it again with food in its mouth. Just to keep this d*mn animal to stop talking. lol
Whaaaaa? You must be joking. 😭
I got two more takes. One, she kept talking instead of pausing. The other, she stopped talking when there was text. But at least, the bear was silent and just munching on his food.
Holy …
What did you prompted?
With my news anchors I promoted that the one who’s not talking is still and silent, not moving, nothing.
I wrote that the girl was talking to the viewer while the bear stood silent behind her, eating.
I do that, but it doesn't work. For instance, for my skull throne, I had prompted that the skulls were inanimate, still, silent, mute, immobile. They still talked instead of the character, despite making it clear that he was the ONLY one talking.
And editing some of her hair over his snout?
That will probably work. I'll try it tomorrow.
You still haven't managed to get the bear to shut his mouth?
Nope
desperate measures
It's fine, don't worry. I will use Canva to edit the new bear visual and add hair tomorrow. But thank you for the help, it was appreaciated. 🙂
I wonder if flipping the image could help at all, or changing something in the description, since sometimes it automatically identifies someone as simply "a character." 🤔 "A woman with red hair speaks in a museum. In the background is an exhibit of a stuffed bear."
I already had it to a woman with red hair. 😦
Okay. I kind of figured but I couldn't help but ask anyway 😁
When asked "man or bear"... The Ai chose bear
👨🏻vs🐻
Beautiful rug, just watch your step
Can you guys give me a hint of trending ai videos tutorial
🤔 It would be a good idea to change the description of the image and say that it is a bear rug... rugs don't talk, not as much as bears do. 
Lmao I tried for fun
Did you do it on your mobile or the web? Now you can do it on the app too.
I did it on the web. I don’t have it yet on the app
Appears in the latest app update
And they've also announced this... Although I don't know what it means... Will it be possible to have a conversation between two characters?... I have to try it because they've put back the group chats.
StreamsFX: Create animated snippets of conversations between 2 public Characters and let em cook! Get creative with three different formats:
Video
Text Only
Text & Audio
She is beautiful
, but she is talking without pauses.
Yeah that’s weird
You can put [|pause|] in between dialogues.
Thanks i didn’t know
EHHHH HOW THE AI DONT PUT MOUTH ON KÖNIG'S FACE
His other pic didn't get selected because no face detected, but this one was fine.
These two were also failed attempts, but 
In any case, if can saw his face, it would recognize him as a person and you would have to modify the image. It's very realistic.
That smile at the end is so PURE
ITS ALWAYS THE END THAT RUIN IT HAHAHAHA
@sterile tundra My last attempt at trying the AvatarFX with masked men
I tried the pause thing and it sounds better now
It's not Ghost anymore with the mouth
The ears 
20 seconds of fanservice with your crush 
Have you seen the tricks that @hard umbra has posted?
Sometimes you can put some simple actions in the image description.
@kagome 
@burnt gate knows she can always dm me if she needs something 
If I need something I will :D
I added more personality in the description and I think it worked...
In this one, at the end, I didn't modify the audio to enhance the voice, but I really like that it remains as background music while they joke around.
https://character.ai/share/v/d24b8a38-28b6-42cd-bd40-3aa608859c59
Older Sister's only warning
Avatarfx can't help but let the mask have it's own mouth, even after I removed the skull pattern 
@bronze yew
The description said he was wearing a black mask covering his mouth?
That is ghost of Call of dutty?
Yes. The original description said, 'a man in a mask looking to the side with a thoughtful expression.' And, yes, it was Ghost from Call of Duty, but it was for my AU bot.
Next goals:
- Stop the bears from talking.
- Stop the masked men from showing their faces.
BAHAHAHAH see! AvatarFX have beef with masked characters. They NEED to put a lips and it will turn into cursed image
-# Why his lips much plumper than me.. And the small chuckle, goddamn 
Trick... upload a picture of yourself to Avatar FX and select the realistic option... it will make a doppelganger of you,
but much more handsome.
No thanks. Im already cute in real life 

Tomorrow I will test how this image looks with this audio.
"Videos Unavaliable"Is it possible to remove those?
Anyone knows how to remove the upper video?
They have been created but the page has not updated correctly... close the browser and reopen it and you will see the videos.
The streams don't work for me... I have the option but nothing generates.
They announced it recently... I don't even know what it will be
😭 ias videos are becoming harder to distinguish from the real thing... well, we can always use the method...
if it's a fat and ugly person it's real, if it looks like a model it's ia
@ornate pebble Aren't you going to try avatar fx with realistic images?
Hmmm. No. I am going to keep with my style for now. 🙂
Talking bears 🐻 🎶 .... the group that made the song Burning down the house
Do you still see the video unavailable or has it already been updated?
still see it in my browser
Now you can also make videos in the app
ok thanks.But what about the browser "unavalible videos"?
It happened to me once, then the video was created.
They end up showing up after a while.
2 successful videos.2 wasted videos and one video left.I think an unsuccessful video shouldn't consume a video attempt.But that is just me.
Ha ha, those aren't fails... they'll appear.
The fails are like this... they help you know that in cartoons the mouth has to look good.
.
This must be some kind of... joke 😭
Will the paper bag camouflage work?
10 tries, 0 good ones.
Barbra Streisand Effect
Ai: Where did she say the bear is? I already found it.
That's the only one that is okay-ish. Although she is not synched with the sound, at least the bear doesn't talk.
Why does the bear have those eyes at the end?
He's just looking at her
Do you have any attempts left today?
Maybe expanding the description?
the bear stays still, like a silent guard, only watching.
I will try it tomorrow. 🙂
Don't let the bear defeat you
😭 Why does he do that at the end?
.
A person covers their forehead with a balaclava, evoking the feeling of being in a harsh and desolate environment. Their posture is static, adding an air of solemnity to the moment. As they sing, their expression reflects deep nostalgia and a serious mood, as if they are reliving past memories. The atmosphere of the scene is intense, capturing the internal struggle between vulnerability and the desire to remain hidden.
Good job! Blurring the bear seems to have worked! Thank you!!!
Have you seen the option in the app to choose two characters for a conversation and generate a video of them... spoiler, it gives an error.
It's not a spoiler anymore. 
Yes, I had an error too
Made it for Tiktok. I used C.ai for the video and talking. Canva for the rest.
Yes, it's great that c.ai lets you upload images of different sizes, so you can create them in different formats: panoramic, 3:4, mobile screen. 😍
I generate a square format and then add the banners. 🙂
Have you seen that the stickers are also available for the C.ai Discord channel?
Ah, yes.
😭 @hard umbra After I get Romeo to shut up... how do I change the image description so Juliet doesn't laugh?
The image depicts a woman with an intense and emotional expression, preparing to make a romantic declaration. Her face reflects a mix of vulnerability and determination, her eyes shining with emotion. The out-of-focus background allows all the attention to focus on her and the meaningful moment she is about to share. The atmosphere is intimate and charged with anticipation, making her declaration seem even more powerful and poignant.
😂 sorry this is amazing.
Hey you make progress 👏
Try promoting that her expressions are calm, soft, never overly exaggerated
Avez-vous vu les nouvelles fonctionnalités de Streams ? Vous pouvez créer deux personnages pour discuter d'un sujet et générer une vidéo de la conversation.
✌️ Now it seems that there is less laughter
The image depicts a woman with a calm and serene expression, preparing to make a romantic declaration. Her face reflects a mixture of vulnerability and reverie, with a look of love and devotion. The blurred background allows all the attention to focus on her and the significant moment she is about to share. The atmosphere is intimate and full of anticipation, making her declaration seem even more intimate and poignant.
Is it a selfie modified with the anime option? 🔍
Yes
It's funny ✌🏼
Tank you
Ughhh what’s wrong with the hand ?
I mean how can I fix that hand cuz it’s weird x)
Arf. Welcome to the club of the crazy hands.
Modifying the image so that the bandana doesn't come off at the end of the video🥰💕💕
The video I made of Napoleon and Josephine, the hands do something strange... but if you don't look closely... 🤫🤫
.
It’s a bit creepy but not that weird lmao 😂
An old one...
Do you still have the tamagotchi with the puke?
And that right after „clean tama!“ 👌🏻
(awh, my shiny name matching the video 🥰🤪)
What song?
🤔 Total Eclipse of heart
🤔 Gimme! Gimme! Gimme!
I need a hero
@ornate pebble In the end I had to use the blur effect so that Romeo wouldn't speak in Juliet's voice... 🙂 ...
Ah yes. Blurring is the key, it seems.
The bad thing is that with realistic images, you make a base and the editor applies the realistic retouching (modifying the image you upload)
😂👌🏻
🤔 So, the streams... will they have their own channels or will they be within Avatar fx since they are video creation? 🤔
Ask a mod, you’re + now
Streams are also text and audio 🤷♀️
Streams don't have as much hype as Avatar Fx anyway 🙂 ... Avatar Fx is more fun because you have more creative freedom
I have fun 😄
By the way @daring matrix Make a video with „let me entertain you“ 😂 would love to see that
❓
And this week the C.ai guys are going to the Anime Xpo... have they prepared themselves? 💪
Somewhere around there... 🔍
Next challenge for Avatar Fx
Hahaha. It already has a hard time with ONE pair of hands! XD
The feature gonna implode 👀
A video I made for Tiktok with one of the streams I did yesterday 🙂
You really are a genius when it comes to this 👏
Thank you. I'm glad you like it. I must admit, I was rather proud of myself. That's why I shared it everywhere. Haha. The music I generated especially for the video. 😅
You have every right to be proud
May I download this and send it to my friend who loves CoD?
It's the Hard Knock Life - Annie's Song
I mean your video 😂 it’s cool
I’ll download the video and send it to a friend
You inspired me, now I have to make my potato sing „my way“ 🤣
rain go or rainbow? 🤷 😭 don't ask him to remember the whole song...
I like these way more then your first ones 👏
but I don't use dialogues, I use audios
Making an audio with intonation and pauses is complicated
Not really. Think how you would talk, generate an audio and listen to it, too fast? Add [|slow|] or [|pause|]
I never go with the first audio, there’s always something you can refine
That smile at the end of the video ☺️😂! 👏
@daring matrix So... how did it go with the hands?
too much ands
Hah. That's too bad. It would have been cool if it had worked.
Opps I did it again
🎤 im not that innocent 🎼
So... Avatar Fx is for 20 seconds of fanservice... 🤔 nooooo... it's to encourage creativity 🥰
I should really try making my potato sing „my way“
Make your potato sing, and remember... add a smile
smile
He doesn’t smile, he’s a meanie 😂
Still cute though haha
I need this thing sing „I did it my way“ for the next Server outages
Muhahahaha 😈
I was lucky that he didn't detect it as children... I wanted that classic animation... 
I love ‚otter eats fish‘ 😂
If he ever gets his own YouTube, tell me please, I’ll follow in an instant 🫡😆
He’s my spirit animal 🤣
lol you got it
Vtubers ✌🏼
@hard umbra Where is the singing potato?
No time yet 😔
Maybe later
@zealous ridge Are you from Germany?
Nope. America.
And you're awake?
I live life on the edge.
@zealous ridge Have you seen moonny's work, she edit the videos with animated border effects and frames
No. If I'm here, I just post my stupid videos and think nothing more. 😋
I like to make music montages... it's fun to see it synchronized
I like making stupid stuff lol
Congrats on the spotlight by the way 
mmm... have you seen the Streams (making a video with two different characters in a conversation)
Thank you. It's pretty cool seeing my sleepy pig up on an official source.
🇦 🇷 🇹
Thanks for the advertising 😊
My last one (which I had posted under my bot collection, but not here to avoid posting too much stuff). 🙂
Hahaha
I can’t 😂😂😂
He’s such a cutie, like saying „oopsie, did I crashed the server?“
In the end, he’s so pleased with himself 😂
I love you potato
(He’s my other spirit animal, we Germans call ourselves potato’s 🥰)
Which one should I use for Reddit the next time the servers go down?
The attack of the pluged patatoe
As a potato who loves tech, I think this may also be my spirit animal 😔🙏
-# hold, make it sweet potato and you are sold LMAOO
😮 a sweet potato as his better half, his counterpart …
That is a great idea!
Omg, if you're going to create one. TAG ME PWEAEEE 👀🍵
This is so on point 👏
May I ask if you speak German?
I do not. I just have a German friend who I really, really like to tease with the worst accent impression I can muster (in otter form).
You nailed it!
The otter sounds like our German soccers 😅 Lothar Mattäus would be a good example!
I love your otter 🦦
@zealous ridge So... what about making two characters with PFP that complement each other to make them talk in a video?
I am awake now. 2 characters in a PFP? I'll see if it can be done.
It would be to be able to make videos of chats and make it look like they are in the same place.
lol, I tried to use it here.
The next step, .. make an otter character, a pig character and make a video conversation with them and make the changes look like camera changes within the same studio
clip 1 (so far). How many do you want?
I'm going to do it myself and see if I can do it...
It's in pre-beta... it works terribly, it always gives an error.
It's one way to do it... but I was referring to the featur (always error-prone
) of doing a video chat.
A stream?
You would need two bots!
As far as I can tell streams even pulls infos from definition!
So you need two well written bots!
anyway I'm trying to do it and it always fails ... I think that feature is not even in Beta... an experiment
This is one of mine.
The infos that the cat hisses at vacuum cleaners, shadow boxing and eats eight raw eggs for breakfast are from his definition.
I didn’t mentioned them in the stream info box.
(Leaving sans out tho he is a fandom character)
Hehe, I think they should check the c.ai watermark... if the background is white it can't be seen.
Well, if you look closely, the letters have a gray border.
Every time
the five videos united
Beautiful.
Sigh. I wanted to do a video clip and cut up songs. But I don't know what I did. I completely messed up the images, and put them on the wrong parts of the song. On that one, she is almost in synch. Almost.
😭 You don't always get what you want, just ask Captain Price, who couldn't get anyone to dance with him... 😭
😍 Give Captain Price a second chance 🙏
🤔
Picture not mine, it was developed by AI
The curse of animating people with masks that you don't want to have mouths continues.
Nice work though so far! That is a really interesting challenge to limit the animation to just what you want.
Two recent animations for Characters I have created. For Misty Spellocks, getting the right gestures to align with the dialog has been a little tricky. Lily Kindred was much easier.
This is the moment ‚Franky‘ and me waited for 🎤
hehe is live on Reddit 😈
Hey, I also made one for the Spanish channel.
Good idea! Let me entertain the Germans with franky
Why would it be so fun to use this... ha ha ha
Do you use German for your avatar FX or just English?
Only English
Take that to general chat this is for avatarfx
no media perms :P
@tropic dew could you be so nice? …
ofc Mademoiselle
The problem was that the image (it's a person) wasn't modified in the app so as not to use it as bullying or a deepfake of a person who could be harmed.
😭 Because the AI has put a wedding ring on him... He didn't have it in the image 😂🤣😂🤣
@hard umbra Have you seen Cap. Price's colorful video?
The good thing is that until I get it I discovered a lot of ways to make analog horror videos 🤣😂🤣😂
Looks nice 😄
The one before had nice gestures.
Did you prompted them?
Our pets aren't safe... just goofing around
In the description of the image I put that he will use his hands to express himself.
Thats nice,
But price is getting a bit old eh?
Can you do something with Obi Wan? I love him 😂 something crazy and silly would be fun
😮 or Darth Sidius
🤔 Darth Sidious It's not very human, I think you can use its images directly.
Worth a try 😂
Music really makes it better 
This video brought me three new followers on my socials ❤️🐈⬛
that sounded meaner then it was meant 😅
👍
Meet Krissy Jones- your southern girl
Wow, the accent is really good.
I wish i could take credit for the voice but its one that was in the system
Family friend's dog expressing some concerns before she heads off on a business trip to Mexico tomorrow...
It's a little far from great.
@daring matrix @hard umbra How are you incorporating music into your videos? Are you doing that during the initial creation or after downloading the file?
Both ways
Oh I see, so instead of uploading just a voice, you upload someone singing, along with the background music. Or, of course, edit afterward.
I'm too boomer for such things. 😌
I upload the audio with voice and do not add anything to the dialogue
If you upload an audio with music, the voice must be clearly distinguishable.
Its so easy, if you want I can give you a „how to do it“ via dm later 😅
I'm definitely going to give it a try. I'm extremely new to this platform. Been having a lot of fun with it. But not new to video and design though.
That would be appreciated 😋
Then... you can go back put on some Todd Rundgren to that drum video 🥁
https://character.ai/share/v/353c055f-d8be-4e3a-8654-1b98194c7ee3
wish the character looked at the "camera" more, and was a lil more animated but it is quite fun to tinker with
Try editing the image description to note that character is looking directly or intently at the camera.
ooooh oki I'll try that next time
Results can vary I think depending on the image and what the AI is able to recognize. I was hoping this one would move his arms more, but the AI was not really cooperating. I do like how the grill for his mouth works though. 🤖 [... off to sleep now!]
true true (have a good one lol)
How did you wrote the prompts?
Did you edited them?
I tried different verbs for moving or gesturing with his arms and hands and just no luck. It mostly ignored them. But on other ones, like this one, it works, so I don't know what the problem was.
Yeah, but what did you wrote?
Did you chose „she’s moving her hands“?
I didn't save the prompts that I used. But I'll try it again when I have a chance. I mostly asked it to move his arms, not his hands, so maybe that was the issue. With the wizardess, I think I just generally said to gesture, so maybe I just need to try some more.
I saved a few of mine, if you want to see, send me a dm
Done! And thanks again!
mmmmm.... Have you tried animating your pfp... or a realistic version of your pfp?
I did in the early stages of Avatar FX testing. Looked like it woke up and had a bad trip.
😆 🤣 😆 🤣
Just for you.
Does this group double for AvatarFX addiction? I stayed up too late last night with this... thing. Ugh, couldn't help myself and in a haze this morning as I hug my coffee cup for moral support.
Oh yes, don't worry. We've all gone through that phase. 🙂
I've been working on a video clip of one of my songs... but until now, I have not managed to get a good result, thus the lack of posting. It's driving me crazy. 😅
It’s so cool that you make your own songs for the videos 👏
Yes, I first made the bots, then I made the songs for the bots and I use it in the videos I make. This is why the song in the moderator bingo video goes so well with it. 🙂
I don't speak English... does it lose a lot of meaning between "But I must be moving on" and "But I must be moving"? The last word got cut off. 😭 I don't want to speed up the audio to fit in 21 seconds.
I think it's okay because he does say moving on at one point and then it seems to cut off at the end , but is sort of implied
The audio has a limit of 21 seconds, they recommend 20... but sometimes I leave it a little longer... because it can be accelerated, but I always take the risk. 🤯
Anyway... I did it again, ...
she is kilias of kilias dark land and hates blue woman. (oc)
@stark needle Can I ask you a question about deepfakes (I'm sure they have all this clearer in the US)
Sure, what's the question? Feel free to DM if you prefer.
It's a question... if I use images of an actor from a series and he's in character, it can be assumed that they're a parody because everyone knows where they come from... but if I use images of that actor outside of filming (interviews and without character)... would that be okay or not?
So this can be a complicated subject, however generally...
The production company of course owns a copyright and trademark to the series. And so you would have no right to use them unless you have permission from them or in the situation where exceptions are provided by copyright law, such as for clearly educational, editorial, comedic purposes, etc.
But even outside of that, likely somebody still owns those interview images, unless you took the images yourself while in a public place.
Buying such images online does not necessarily confer the right to replublish those images either.
It's a complicated subject, and even if you believe you have the rights to use them, you could still be legally challenged and have to defend the usage.
But, I routinely see people doing this online all the time anyway. Often worst case is a DMC takedown, unless some kind of libel situation is involved.
the problem with generative AI's for conversational chatbots is that there was no control over what would be done with the bot, but Avatar Fx makes a finished product that cannot be modified... it's easier to use the premise that it's a parody (in the USA I think parodies are allowed and Avatar Fx would be like a video editor)... but it's also a Deepfake tool (that's why it modifies images of people)... I usually use distorted images (apart from the C.ai watermark) and the voice I use has nothing to do with the actor... and I guess the person detection protection is more to avoid bullying (most C.ai users are students)
And of course... the companys always say that a chatbot's responses are generated by the company's Ai system and that it's his responsible...
AI isn't yet very regulated because it's very new and still in development.
But Avatar Fx is more of a video editor, I guess that makes it easier to interpret what is parody and what is impersonation.
That is a legitimate risk. Parody is a good approach, especially if you state in the character's description that it is a parody.
I'm not sure if this image is from a movie or from his daily life. That's what I mean.
.
Ah, I see. As a personal creative endeavor, in this closed group, risk is extremely low, but were you to publish fully public, worst case is usually just a DMC takedown on the platform + a strike against you on the platform (slap on the wrist).
Not that it necessarily means you did anything wrong. But those DMC take downs are just often times super aggressive.
Digital Millennium Copyright Act (DMCA)?
Note that I'm no lawyer, but I have personally dealt with these issues in my career.
That's what I get for trying to test how he handles a cigarette, ha ha
https://character.ai/share/v/5a08f17d-d976-4a3f-b52b-3acd64b167f3
Her eyes popping out of her face. I'm tempted to keep it for the video clip. 😂
Slaaaay 💅🏼
Awh I love him ☺️
My newest ghost story 👻
@hard umbra
When I try to upload this image to Avatar Fx I get a strange message... does the same thing happen @hard umbra ?
What does the message say?
„Seven days?“ 😂
Image not safe ... 🤷♂️
Did you took a screenshot?
Could be the logo from my TikTok account?! 🤔
Hehe, tsk tsk tsk
We told you to ask before taking 😄
If you promise to be respectful to Albert, I’ll give you the pfp from the bot without the watermarks
No, because I've uploaded several with the logo... although I could delete it to test...
Some of them were identified as real people and gave me the option to edit them, but that one... not safe.
🤷♂️
Hmm interesting 🤔
None of us know exactly how the censors work for this feature.
I had it only one time that an image was rejected, two realistic news anchors
But it usually tells you:
- There is a child in that image
- A real person has been detected in the image
- That image appears to have protected content, please choose another one
The mystery is... what's the difference between the first image and the second? ||In the first, it has one more layer; in the second, one less, and the features are more diffuse.||
@hard umbra If you want to test it by uploading one and the other to Avatar Fx
Thanks not really, I can edit those effects afterwards.
Here, be nice to him.
He’s a lonely soul 😢
I made a previous one because I wanted the effect I got in an animation that came out very analog horror, but it didn't have the same effect.
Okay wow …
You have the same „problem“ like I had back then 🤯
He’s only looking at the right side
I had to redo the door in the background. The image wasn’t balanced …
I made a new one of him today for my feed, let me screenshot my prompts, this might work better …
In that one I used the prompt you put a while ago. In the one that only looks forward I put a different prompt.
Awh I love his lil grin ☺️
He’s a nice ghost
I made this one for my feed on c.ai, the first one I made when the bot wasn’t finished back then
I added something to the prompt and edited the door.
Huge difference in result
Wow, it is a very noticeable difference
Absolutely
I noticed that it’s important that the image is „balanced“
If something is too prominent or outstanding, the ‚character‘ acts different
(Anyone understands the silly non-native? 🙈)
If an image doesn’t ‚work‘ I change it or make a whole new one.
That’s why most of my avatar.fx have other images then the bots Profils … 🥲 and so don’t show up on their Profils 😭
@stark needle What would you say differentiates these two images? It's not the TikTok logo.
The color and contrast are just a little different between the two. And that little bit, makes a difference to the AI?
The first one lets you upload it, the second one tells you that it is not safe.
Funny. I have run into some similar trouble uploading an image recently. Funnier... if I keep reloading it, sometimes the AI will accept it. Its answer changes! LOL
Maybe the second is just too blurry?
My wife runs an ecommerce business and we once had a product get banned because some AI mistook a soap bar (just a normal rectangular bar) as alcohol. I mean, I understand sometimes AI can see things, but really? It can be so puzzling sometimes what the AI sees.
I have sometimes gotten a message saying... the dialog is not appropriate, please change it... when in fact the problem was the size of the image, if I resized it there was no problem.
What the … how nice for a little experiment 
Fingers crossed I get a good longer one.
Just look at the mouth(s) 🤯
its great
Love it. That is awesome.
The problem is the unpredictability of AI; the person and the cat may suddenly separate...
@hard umbra Have you seen how to get videos that have long... Using the last frame to generate the next video.... (A bit of saltiness is appreciated but it is because I use images that are not detected as people And I have to edit it again and I made it blacker)
I have 5 tries left for today.
Got some stuff to do, but I’ll try again later with longer prompts. It was only an experiment to see what happens
@hard umbra Have you seen how to get videos that have long... Using the last frame to generate the next video.... (A bit of saltiness is appreciated but it is because I use images that are not detected as people And I have to edit it again and I made it blacker) But if the image is not of a real person you shouldn't have that jump
Neat. Sort of a manual technique to "expand" from the end of one video into the next to create a longer sequence. That's a cool idea.
It has a 20-second limitation... at first it is only to make a short presentation of a character

So how is "Jaxpurr" doing? Or whatever his name is?
His name is Kuro 😅
Or „Mr. Paws“ if I want to annoy him in his cat form
I’ll try again. I have 5 left, so I think about the prompts one more time
I'm sure some thought does have to go into that! He potentially has a conflicted personality, haha
😭 I already had plans for this photo but... it has a signature
Create a similar photo. HD doesn't appear to claim copyright. Just says not to steal which doesn't carry the same legal standing as properly copyrighting.
Not to be mean about it. But that do not steal reference just does zero to enforce legal ownership.
okay it wroks... but it kinda creepy 👀 ill change the image description
(Please don’t use him!)
Amazing. That is so cool.
I’ll take this, number 9 of today and the 4. serious try on the shape shifter.
It’s amazing that it works!
I once tried to animate a black cat and it didn’t worked.
Guess the human half of the mouth helped here.
-# human teeth on the cat half still bother me, but could be worse 👀
It’s sooo much better on feed 
The cat is so much more detailed 🥰 feed zooms in
🤷♂️ Capt. Price isn't convinced either... songs with a lot of vocal effects don't come out well.
.
Discovered my new ringtone! 😂
Looks like he lost a karaoke bet 😭
If you want a really crazy one...
.
😭 lost a bet in karaoke
.
After much ado, I think figured out my robot problem. Turns out the issue was his raised right hand in the original image (attached for reference). The raised arm position and the fist's similar size to the robot's head was actually confusing the AI. Took me a lot of fails to figure this out. Once I did and then changed the arm's position to below shoulder height (also simplifying the image as well), the prompts began working a whole lot better, especially after some great advice about prompting from @hard umbra (ty again!). Learned a lot from all this... more to learn too. Happy to make progress.
Yeahy 🙌 good job!
This one looks so good!
It’s so cool to see what others create and what the feature is able to generate.
This is so good!
I have only one Price's voice (other songs are Ghost and Konig mostly) Feel free to use it.
What I usually do is find songs on the internet and use a voice changer... (it's fun, even if the voice doesn't match the character)
It's great that it syncs lips and expressions with any audio
Minor things I wish I improve but really glad to at least have figured out the major issue 🙂
The smaller things are hard to get, this is really luck.
(Tho you can experiment with stuff like giggling, breathing and sighing for example! But put it into the dialogue box.
My original potato has a diabolic laughter and I love it 😂)
But you definitely made progress.
Thank you again! Couple late evenings working on that one...
Funny thing, the music is Don't Wait Up by Midnight Generation, haha. A band from Mexico City. One of them bought a beat box and did an ode to Daft Punk.
Tried to add a link to the song in the last comment but looks like the server blocks such things. ¯_(ツ)_/¯
you have c.ai+ you can upload files on any channel
Nope, links to outside will be removed
upload files not the link
Seriously sick 🤩
I finally managed to make my video clip
Beautiful. Fantastic. Exceptional.
Thank you. I'm really glad you like it. It took me ages (about a week) to make (I had to cut the audio in 11 and so I had to generate 11 good pictures and videos). My bingo movie and news anchor didn't take that long. 🥲
Wow, I can only imagine the work that went into that. Well done!!!
It's Peter's mini clips that inspired me. 😉
Small thing I noticed. She signs the contract upside down. But the contract is still just as legally binding. The intent to sign outweighs a clerical isuue, in legal precedent, generally. So the assassin should have no worries about being paid, haha. 🗡️
Yes, you're right. 😬
Still, really impressive work on that. That's your song, right?
Although a contract which purpose is illegal, signed by a party who clearly can't read (since it's upside down) would be void. 😬
Yes, it's my song. I make them for the videos.
Well, amazing song too!
Yeah, the assassin probably would not want to argue any complaints in court, lol
Thank you!!! ☺️
I can read upside down, so maybe she can too 
[If you have extra space in the Definition, you could add that 🤭]
@ornate pebble If you aren't already, you should consider publishing your songs too. Lots self publication platforms out there. Doesn't have to become some celebrity status, but can be a cool addition to your other content.
I use AI for the songs, although I own the rights. I give the system lyrics and tell it what type of song I want. :]
Even then something like a soundcloud account could be a really cool way to get that out there by itself. 🙂
Oh. I'll look into it! Thanks for the suggestion!
Adapting the bot to fit the clip would be funny. But my definitions are always tightly squeezed in these mere 3.200 signs.
I just created an account with Soundcloud. I'll upload my songs and I'll DM you with the link, if you want. 🙂
Yes please 🙏
Are you ready for My Talking Tom & Friends 2?, Becca Sparkles ranting about upcoming My Talking Tom & Friend2 game release in 31 August 2025.
This is the worst one yet, and I'm almost proud of how bad it is.
I might be going out on a limb here...
but I'm beginning to think you are a fan of 8️⃣ 0️⃣ 's music 😃
I'm an even bigger fan of showtunes 🤫
Ah, I'm a man of culture myself... very much enjoyed "Once More, with Feeling" from Buffy the Vampire Slayer. I didn't say cultured, just culture.
Beetlejuice the musical has some bangers, if you've not had a sip of that fine wine.
Really loved that movie. The original. Haven't actually seen the second one.
I don't know anything about the sequel, either. I'm referring the Broadway musical of Beetlejuice.
Oh right, doh! We do see an occasional show. Will add that to our list.
See, now look what you have done @zealous ridge ... you've caused an 80s thing in the channel
@stark needle I need you to check something... does it really say "Tiger"?
No, it doesn't. But perhaps an intentional misspelling in the prompt might force the pronunciation such as "tie grr"
but it doesn't say toilet 🤔
It does sound a little like that... haha... not quite... sort of like tunget or something... but a roaring toilet is quite the visual, I must say
Hey, how's the robot animation going? Does it beatbox? Have you thought about adding a snippet of Kraftwerk's song The Robots?
I did post an update a few days ago. I've not worked on it since, as just taking a little break from that project. Kraftwerk's The Robots could be great for it though. 🙂
I've been working on a few other interesting projects though.
One of those is for our local Comic Con coming up in a week and my family and I are going as a group cosplay featuring characters from Demon Slayer. My son is going as Tanjiro Kamado, wife as Shinibu Kocho, daughter as Nezuko Kamado, and me as Inosuke Hashibira. We're going all out on it.
Maybe Comic Con will inspire doing a c.ai bot on a character from that show...
Then there's some musical inspiration from @ornate pebble's latest video. I tried my hand at an AI generated song. On a lark, I transcribed a meeting at work, had AI create lyrics, and then generated a song based on a Teams meeting. It's in a classic rock format. My coworkers got a good laugh. I'm really impressed with what you can do to create your own music for these videos.
Have you seen the C.ai Parent Activity Summary option?
I wasn't aware of that, but thanks for letting me know. 🙂
Yes, it's associated with an email account and informs parents how much time they spend online (web, app) and which bots they talk to the most. I think the Ai song thing is currently in vogue... like making songs out of comments on posts... especially if they're haters. 😆 🤣 😆 🤣
And then there's this... a different character from the robot... a violet witch. I really like the emotional expressiveness of her hands. But still a work in progress. This is sort of a sci-fi / fantasy hybrid... she's a SUPER POWERFUL witch.
The artwork is inspired from the cover art of the song. Very mysterious and my curiosity just wants to explore who this person is. I created the basic image in Photoshop, inspired from the cover art.
@daring matrix I'm a pushover for anything sci-fi or fantasy. Played a lot of role playing games growing up. Most notably AD&D via the old books and dice and written character sheets. Our group would sometimes play to the break of dawn and crash in the mornings. Good times.
It has not been said yet, but I find that the voice acting in Character FX is REALLY good. Sometimes, I make Avatar FX videos just to be able to rip the audio. I tried to use the AI voices of Canva and I was surprised as to how robotic they were in comparison.
Hehe, also... generating dialogue audio doesn't consume one of the 10 videos... 🤷♂️ you just have to have an app that records desktop sound
I don't, I had to make the videos and to rip the sound.
When I finally saw that they had left open the possibility of using C.ai's free text-to-speech (it no longer depended on the bot's generation and if you modified an answer it was not played in audio) I was really excited... then I remembered that I am very lazy and I prefer to use already made songs to generate videos.
I have not seen another AI tool that makes such great voices. I wonder if there is. Because they sound freakingly good (when you pick the right voices, and the right prompts, which takes a long time though).
As hackers say... if you can see it (and in this case hear it) you can copy it.
It would be great if you could save the audio for when you run out of attempts. Or even if you could download the audio file without generating the video for future uses.
Do you have private messages blocked?
Hmmm. I think you need to be friends with me to message me. I have had a lot of people message me to flirt (🙄) or to sell me their cr*p. So I disabled the messaging out of the blue.
send me one so I can send you something privately
You asked! So here you go!
🤖 Kraftwerk's "The Robots" 🤖
.
💅 👠 🪩 🎶
https://character.ai/share/v/b23be3e4-2e49-4d11-ad33-35923147a960
@ornate pebble some day
Some day, just not today. 😂
But at least, it moved around the end.
almost all the special effects appear at the end
That's crazy to even think about.
Possibly could reference all arms moving, but...
Might need to refer to Arm1 thru Arm14 or maybe even Arm1_Shoulder, Arm1_Elbow, Arm1_Wrist, Arm1_Hand thru Arm14_Shoulder, Arm14_Elbow, Arm14_Wrist, Arm14_Hand depending on the animation desired.
All within 500 characters. 😬 Probably need at least twice this much space for intricate instructions.
That would give the Golden Voyage of Sinbad a serious run for the money!
Almost! Can't quite seem to get him to dance. Towards the end you see a little groove start up there.
all the time could be spent on something better like learning Blender or some shit lol
Blender seems nice and true it is probably better for 3D animation, but ultimately if we're here for the characters and stories on c.ai, then Avatar is the only way to go.
But learning Blender is not fun.
yeah to you maybe
scatman song is copyright protected btw but of course you wouldn't care about that
It could be... the same as Nintendo's use of the Crazyredd brand and character
it's a fucking username as if there isn't a million BuffaloBillsfan123 usernames in the world
also it's stylized Crazyredd one word and not Crazy Redd like the character.
🤷♂️
also the character in my profile isn't even the Animal Crossing character in question, and not from a franchise at all.
This thread is for sharing Avatar FX creations. So those who have nothing to share and nothing constructive to say, have nothing to do here.
By the way, I tried to do a Streams video last night with a couple of characters discussing the Fair Use Doctrine of US Copyright Law, i.e. non-commercial, artistic/transformative, and learning oriented works shared within the confines of a private group. But trying to get these characters to discuss that specifically is beyond my current skill level here.
That would have been funny. With a big sign "NEVER TRUST A BOT ON LEGAL MATTERS"
Most definitely. As an aside, I work with a team lawyers pretty regularly... and though I'm not a lawyer myself, I will get asked questions sometimes regarding trademarks and copyrights. Once I actually met in one on one conversation with one of the key authors of the US Copyright Law. It was a fascinating conversation.
What we all do here generally falls well within the Fair Use Provision. Of course, anyone can claim otherwise, as arguing is a core tenant in legal.
To be honest, no one is going to sue an individual for having made a 20 second clip with their character in it unless the use of the character is offensive and causes them a damage or unless it is being monetized and they are not getting royalties for the use of the character.
As there is not issue with taking a character's name as a username if there is no ill intent behind it. 🤷♀️
And the "ill intent" can potentially fall under Parody. Comedy in general, as a form of Free Speech, is strongly protected. Of course outright defamation is never a good move.
Which reminds me, I need to have a word with my two characters... the disparagement! How dare they!
@ornate pebble BTW, I've continued to doodle on the music. Excited I might get to finaly live out my fantasy as a rock star, lol. Still shopping for the right polyster outfit and trying to decide if I should grow out my hair or just buy a wig.
And that's a really good option too for anyone who wants to add music but maybe feels squeamish about copyrights. Attached is a little snippet I made... and of course one can do any genre, lyrics, length, whatever.
The easiest way to add music is to add it directly on a Tiktok video and to download it from there.
I'll give it try. But let's say someone is trying to make it big shuffle dancing. Would this hurt their algorithm? Or maybe just delete after downloading?
You can publish the video in "private" and then remove it.
Or you just leave it. Who knows, maybe some people will go after your robot. 🦾
The same old problem: they are designed to speak and they don't recognize sustained vowels 😭
Ran into a similar issue trying to upload a voice based on non verbal singing... system did not like it. Similar problem with a robot voice. Seems it has to be very close to human speech.
Maybe I'll put in a fragment of Psycho Killer by Talking Heads.
Great band. I can hardly contain myself when Burning Down the House starts pkaying.
and I guess it works better with beatboxing than with sustained vocals in songs
Great point!
The robot's anodized metal effect looks fantastic, and the artificial intelligence has taken it as that effect and not as paint; it changes color with movement and light.
🤷♂️ I don't understand the lyrics
right, makes sense. sorry about my attitude earlier.
there is something to be wary of though, in terms of "fair use" and if avatarfx starts allowing longer clips.
Yes, that is true. The longer the clip, and especially if most or all of a song, or its core part, is involved, that raises the stakes.
I think sooner or later they will eliminate Avatar Fx as it is now... just because it is popular or fun does not mean they will keep it.They will surely integrate it with the bot generation and only generate the bot greeting or to do a video conference in a chat. 🤷♂️ It's supposed to be an experiment and character.ai is just for creating bots and chatting with them, not for making funny videos. 😭
Additionally, character.ai users are obsessed with the bots' memory, the romance, the fact that they're franchise characters, and the availability without server crashes... and of course, that everything is free😂🤣😂🤣
Free romance or paid romance? 🤔 They say you get what you pay for, but for some things, free is priceless. 🌹
I'm not really sure which one I like more. 😅 I like the drumming, but the expressions and gestures at the end of the second one are pretty nice. Next time will have to choose a different pose for the lifted paw.
I actually only use C.ai for Avatar Fx... I stopped using C.ai chats for a while and started using deepseek, it's weirdly a lot of fun as a model. Although it does not have the possibility of creating bots, if you are an experimented user you can create prompts to emulate them as well as a well-defined bot... It does not cut off your answers, It always surprises you with interesting facts, creates NPCs about the plot, you can make it access the internet to chat with a bot about current affairs. And although it is from China, its censorship is less noticeable than that of C.ai.
In August I will test the chats again intensively, when I returned I only used the dinamic model But the only reflective model that is really capable of interpreting bot definitions well is Nyan (it makes sense that it is the specialized one)Still all bots have limited response tokens and although Nyan offers longer responses they always get cut off and you have to send him an empty message to get the full response with the drawback that if you take too long to do so, he will lose the thread of the original answer and invent another one.
I also think that to avoid problems they adjusted the bot response temperature to such a low level that the typical response from a bot now is:
Oh, that sounds interesting, I really like it, tell me more...
Instead of developing something imaginative
They are both good, but I feel like the one with him standing up is a little better. Unless drums are somehow important to the character?
@daring matrix Fascinating stuff! I'd be interested in hearing about your findings in August!
↑ I removed it in this video XD, sorry, Thanks AvatarFX, you make my dream come true ♥, Becca is sooo cute~~~~ and the voice is matching with the real official one, Oh! and the newest game (My Talking Tom & Friends 2) will be release today (± 3 hours since I type this message).
I love that now in character.ai the AI offers a tangible product
BRB I cooking something.
🤔 Really.... Hooooooow?
We need more information about multiple speakers and turns....
These videos are from the Avatar Fx presentation.... will they release those options? 
A game? I've been a little curious about that, especially with the unpredictability of AI. Will it consistently deliver the game as intended? Or do you just accept that there will be variation in the game's outcome? I'll check it out when I get a chance.
I'm kind of similar... I mostly interact within the AvatarFX discussions. I love the serious design focus.
I use ChatGPT, Co-Pilot (Enterprise edition), and Gemini mostly. And a few lesser known ones.
But writing is really my main bag, more than anything else. Have done and am currently doing some significant writing outside of c.ai. Both fiction and non-fiction. But as much as I love words, visual story telling is incredible and growing. So that's a big reason why I'm here. Plus, it's just friggin fun!
Also use graphics AI too. Mainly the built in versions within Canva and Adobe Creative Suite. Plus a collection of several AI voice and video editors. That includes Riverside for podcasting. So much going on out there.
A nice feature c.ai could add is captions for language translation 🙂
It's Romanian, it became famous and there is an English version... it practically says...
Si te rog iubirea mea primeste fericirea```
```Hello, hello, I'm an outlaw
And please my love receive happiness```
I'm not sure you want the translation of most songs in foreign languages.
Agreed. Just thinking out loud generally
The song is "Dragostea din tei"
I have a selection of non-English songs on my Playlists. Most recent addition is Nemonemo by Yena. Very cute song!
Anyway, the fun part is watching the video clip in Romanian, 😆 🤣😆 🤣😆 🤣
Platonic Dating, Or How to Date a Robot
Funny she says "trip" .. because I think she was supposed to be literally tripping 😂
Its just a silly kids mobile game (I not into mobile game), I rate it 1/5, watch fishes in aquarium swimming around left and right is more enjoyable than play the game, I just had a crush to Becca Sparkles, that it, me being silly.
The last game I play is Black Desert Online, its a Korean MMORPG, I just face a financial problem lead to the point I cant pay my unlimited internet and need to use my phone as modem (tethering) with really limited data (6GB/3Days).
Ah, ok, so it's a pretty simple game. Btw, familiar with Black Dessert.. played a whole series of games like that one. 🙂
Stealth archer here, haha... trouble maker 😂
Guess with who I had crush in BDO? ||(I basically have a crush to all bunny anthro)||
And this is what my character look like in BDO ♥
why ?
Got those reflective surfaces going 🙂
Created a couple of AvatarFX intros for two new characters I just added. The first four characters have really all been mostly about just learning how to create characters, more than anything else.
Note: The Violet Witch is inspired from the cover art character from SLAVA FUNK! by MVSTERIOUS, hxmr, yngastrobeatz, & EVO. The character's plight, bound in anti-magic chains and facing a dangerous foe in an arena fight, is just something I made up for her backstory.
The help bot is just a customer service and tech support assistant. About as useful as most of them tend to be.
I had to censor the audio... 😭
So mean the AI wouldn't let you include it?
It's for teenagers... it has censorship
If it detects an inappropriate word, it tells you that the script is not allowed and that you should modify it.
This is not all that great but I was just curious what I could get away with... couple of rando pics from the Internet.
the second is great
Welp, no more baths for me! 
Then I better not upload the one with the cockroach singing Frank Sinatra's I've Got You Under My Skin.
Oh now you have to!
They don't want to cooperate... even President Washington has gotten angry with them.
Seems George was just as surprised about the hand as I was 😂
Hey @hard umbra It's been a while since you last logged in, a mini-vacation, or a project with tight deadlines.
Just taking a break from C.ai discord 😅
Am still cooking
Here takes this as example 😄
I love him so much 🥰
the little grin at the end 😍
Have you named him? "Kitty"?
Kit Kat would be a great pet name
Someone seems to be distracting him during the filming. 😬 Where is set security when you need them? Off somewhere having a sandwich?!
😆 🤣 😆 🤣 😆 🤣
Is it my imagination or does he briefly switch languages in both near the end?
I have a couple I've done where the accent changes. From an American English to British English for example. But that's always been with the AI speech of course.
Wait how do you get the ai to sing in your video?
Uploading an audio with voice
Uploading an audio with voice
Amazing. Head is turned but it recognizes the face.
The good thing is that you can use practically anything as a base for animation.
And similarly the upload can make the AI yell. Or whisper. Perhaps beatbox.
A nice song, but it lengthens the vowels a lot and doesn't detect them as words.
Funny thing in c.ai... many choices for characters who are singers, but if you ask any of them to actually sing, all you get is a description about them singing.
I tried uploading someone singing as the voice, but the AI couldn't handle it. So I guess no musicals! At least not yet!
Just an experiment with cubism (mostly). The first few (more traditional anatomy) work pretty well; the others (less traditional anatomy) struggle. The vocal track features both some easier and harder vocal segments.
Updated the avatar for my Violet Witch. This substantially deviates from the original design. But decided this would be more practical to work with. Getting the Definition mechanics for this character to work was a struggle. Thank you @hard umbra for the advice. Still ironing out a few things, but was fun to do.
@daring matrix Nice, that one worked quite a bit better. 🙂
Comic Con here today... Getting ready ⚔️ 😁
I finally decided to hack the original image.
I love adding sound effects now ☺️
The little „what?“ even matches the moment the noise stopped and the gesture at the end = 😂
I love this feature
Real images always animate better
I like pin up style.
She gives me this Christina Aguilera vibes from her „candyman“
What would also be a great song
Ahm… you know what’s he’s singing? 😅
Mmmmm...... No 😏
There's a video I found on YouTube explaining the lyrics... 🫢
A classic "love ballad" about a guy looking for a girl and describing the characteristic most important to him... (which I won't repeat) 😮
Day 2 for me today at Comic Con... I did about 9 photo opp requests yesterday... people seem to love Inosuke!
By the way, some users use Avatar FX to take a selfie, convert the image to anime or a 3D model, and add a recording. Things to know about using the app to create an Avatar FX
Inosuke? Where? How? 👀
... about music and love... romance is clearly one of the more popular themes on Character.ai. I'm increasingly leaning into that... not solely all on its own, but as an important element to weave into character backstories and videos.
For one of my characters, I created a sort of "searching for love" song last night... it was an interesting experience being able to do that via AI, which is enabling me with a level of artistic expression I simply could not have otherwise achieved. My thanks again @ornate pebble for the music tips.
I am cosplaying him at our local comic con here. I had both adults and kids getting pictures of me posing with them.
Used it as a bargaining chip at a vendor booth haha... my wife got a deal on a illustration of Frieren in exchange for me posing in the vendor's booth looking at the wares. 😂
still trying to figure ts out
The music is a nice touch. One thing you can also do (not sure if you did or not) is edit the image description to put in things like "she gestures with her left hand to emphasize her words" or "her left arm rises to gesture at the stars" or even both arms if you like.
Which can for example allow you do things like what @daring matrix demonstrates in the earlier above three videos. 🙂
i did not know you could do this, thank you so much!
You're welcome. Just paying it forward here, as someone helped me with the same thing not so long ago. 🙂
SHE'S GORGEOUS AHHHHH
How difficult was that to do @daring matrix? Getting all their arms and heads to move?
Good luck, the animation is automatic... it depends on the image and the modification you make to the description... but after that it's a matter of luck.
The only thing is that realistic images work better than drawings for movements.
Anyway I was bored the 10 videos expire at day So make many videos with image variants and description until that
Typo seems to be tinkering with something, clinks and clanks emanating from a his little workshop.
Poor Harry Du Bois
Franky and me are in Potato-heaven now! The new stickers are 🥔🔥 🥹
Hey, how are you doing with Ai's music and videos? The problem comes with songs with a lot of autotune and resources like word lengthening.
Too young and strong to be Harrys Du Bois?
Downgrade to look like Harrys Du Bois...
I need to find a trick to make it do something while the vocals are held in the song... 😈
-# and check that he says the lyrics of the song and does not make up words
I've been working on something myself. Hoping to publish and share this weekend. 1000% better than anything else I've done here. Required stitching together a dozen AvatarFX videos. Which can be tricky to do, because of the black c.ai AD they insert at the end of each video.
Yes, the 20-second thing is only good for memes.
I stay with background noises and sound effects for a while ☺️
I like it, feels like making trailer for my stories.
Bad animation today, I don’t know, it refused to acknowledge my prompts
Haha lol, just noticed on my feed, that the food in his hand changes and then vanishes 😂
The zoom in feed is sometimes a curse
Still, looks like a scene from a show. 🙂
Thank you. Yes, I like this. It’s like making little trailer, telling part of the story.
I also liked the SETI with the noise in the beginning.
I wonder what I should try next.
Maybe one of my Karen’s in a mal?
Haggling with a shopkeeper 😂
Karen doesn’t haggle, Karen demands 😏
But made so much with her now.
Maybe I take the Karen neighbor.
She feels blinded by the light coming through the crack under your door
Could be very funny. Well, maybe not for the recipient of her attention. But I think I might be safe here on Discord, haha
Nice! All three moving!
@hard umbra Link to the potato?
I think this is it: https://character.ai/chat/qCA3kHk5uoJGXs6tKw-H0Mr4sgr0DFHjYFtx4ntos4o
Chat with Evil Potato-Server: Character.ai’s evil Potato-Server | OC Villain. Powered by an industry leading, AI large language model (LLM).
I will add Potato server support to the Help-Bot 3000's support script for handling customer service requests, LOL
Typo already gave you the link
(Thanks typo
)
Here I have a bingo-card for the knobbly root of all evil.
@ornate pebble and me started to make chat-bingo cards. It’s fun.
No pressure, just leaving it here
I can't get Harrier to be the one who takes the lead... but at least he spends all his time laughing and making faces, it fits the character.
Perhaps tell the AI the guy in glasses is super shy and afraid to sing. 😂
Any idea what the maximum file size is that we can upload in this channel? I've got a new animation that I'm hoping to share in the very near future. But I think potentially I won't be able to share it here due to the file size. And I don't think video links are accepted either. Thoughts?
If you don't have nitro, there is a file limit... I don't remember the limit, I would compress the videos again.
I think the maximum was about 16 megabytes... but it was quite a size. This video have 9 megas
Hey... how are you doing with the new models, the renewed Nyan and the new DeepSqueak?
I've not tried DeepSqueak. But have tried Nyon and it was actually pretty cool. But most of my testing has been on Roar, as I've been trying to ensure the characters will work well under that more limited model.
Nyon definitely seems more intelligent and able to handle nuances more readily than Roar.
Yes, but I end up keeping my characters unlisted... and Nyan works very well. I only have a few audiences, some with little plot. But the ones I like, since maybe only I like them, I leave them unlisted and make them with Nyan, which is a model that executes the definition well and is very dynamic and inventive.
Personally, I do use Nyon, for my own conversations. Makes for a richer experience. 🙂
I am pretty excited about the new video I created. It's a three minute video built out of 14 segments. Much more of an ambitious project than anything else i've done here. And it ties into a character I created.
I may be limited to sharing a low res version in this channel. But the full version is 4K with a Wav file audio. Going to publish it as my first foray into being a content creator. When I do, I'm planning a parallel post to provide background, including thank yous to you and other people in this group. 🙂
I do really appreciate all of the helpful advice from everyone here!
@hard umbra My character Help-Bot 3000's customer support protocols have now been updated to include assistance with issues involving the Evil Potato-Server, should any users happen to inquire or complain about said server. Note that the Help-Bot personally believes the server is just misunderstood, but is committed to providing exceptional service and support on par with Character AI's own standards of support. 😂
New AvatarFX video for my character Violet.
I think between compression and sharing only as a snippet preview, with a link shown at the end of where to find the full length version, that's probably the only way I'll be able to share it here. But that should work.
This one looks amazing 👏
Thanks! Yeah, she's a sad one in need of cheering up.
Now you inspired me.
What might happen when this dialogue gets a sound added with like ‚sad noises‘? Sobbing and so? 🤔
Would that effect the facial expressions?
Maybe I try something like this later
I did actually make a song with it. Using Suno. I'll be sharing that this week.
Would cool to see! It's a good idea. Some well placed sniffles!
Who knows? Maybe it works.
I was surprised how well my sabertooth got animated with the simple roaring sound.
Maybe a good cut audio where the sniffles fit in a pause of dialogue…
Question on this. Was your thought to just upload an audio track that includes the sniffles/sobs, or were you just suggesting it as an image prompt? I had not tried that before, but sounds intriguing.
My plan ist to create an audio file.
But for the animation I would still edit the image description.
It’s a bit complicated.
In my restaurant yesterday, the chef had a little sigh in his dialogue, but when I mixed the noises into it, … it’s not audible anymore 🤷♀️
It’s a lot of trying and experimenting
I use audios of songs, but... What I would do if you want the music or sounds to have an effect (the music is not going to influence but the sounds might because what it does is transcribe the audio to text) It is generating a dialogue, recording it when it is generated and mixing it with the audios and uploading it as sound to Avatar Fx
That's how I did the Harry Du Bois one, but what Avatar does is transcribe the audio to text.
I do something similar.
use C.ai text-to-speech, record it and add effects to it and then pass it as audio to Avatar fx?
Not completely, but close 😅
I won’t spill all my lil secrets, at least not now 😬
Or could simply strip out the audio track from the mp4. Then edit to layer in the sound effects for a new audio file.
Only have 10 videos and you can generate audios all the time
If you have a great mic or a direct connection, it might work well. 🙂
I mean generating audio in the dialogue section, playing it, recording it and using it to add effects (it will only distinguish the words, but... can a sneeze, or any recognizable sound affect the animation? 🤷♂️ )
Hey @stark needle How is your macro film project going?
It's finished. 🙂 But I've just been prepping some related things around it. Will post in a matter of days. 🙂
Love the background effect! Like a painting!
Yes, as if it were a watercolor of a sunny city
@daring matrix Here we go! File size upload limit is 100MB for non-Nitro users and 500MB for Nitro users.
Thomas
Nice to see you! 🙂
The accent is nice!
Dunno why she made that face at the end but otherwise this ain’t too bad. The audio is golden ngl. Sounds like she’s seething when she says “domestic.” 😂👌✨
Might animate the audio myself ngl
It was a reference to this old post of mine tho
This was the original image. It was struggling because one of her arms is already up. I didn’t want to how it down, however, let the ai give her black paw beans.
its so peak.
I love it so much 😆
And the little greeting at the end!
uncanny realistic saness whispering some enchanting table language
Cool. The world needs more pirates! 🦜 🏴☠️ ⚓
Indeed 🏴☠️😅
Next I try to recreate that infamous scene with Maximus by using my own gladiator ☺️
I just need a nice background sound, maybe a crowd cheering or the clashes of swords
Ocean waves and seagulls too maybe for the pirates!
Tried an arena recently, but had some challenges with the AI staying on script for combat. This is an example LOL
So no calling her "kitty" then? 😹
This is the example I was trying to share. Being able to tell your opponent to go stand in the corner was not quite what I had in mind, LOL. Well laid plans (or perhaps not), as they say! #WantingInDefinitionSkills 😂
NO-! I made two videos about this! 😭
I need a responsible person who is willing to work
Harry Du Bois? He's a character... who doesn't have a very defined appearance...
Looks pretty close to me!
What's this for?
Let's say it's a channel where you're allowed to upload media whether you have C.ai+ or not.
You can upload your Avatar FX creations.
Ohh I see now but people barely send anything here
People prefer to chat with bots, even if they are broken.
@daring matrix my dear, you inspired me to use odd songs 😆
I hate this bot so much and now I found the fitting song for him 😂
Oh yeah YOU COOKED!!!!! 🙌🙌🙌🙌
This one is way better then mine!!!
I love it
@daring matrix may I download this and post it on my TikTok? I’ll of course mention you, give me your c.ai if you want that too
It doesn't matter, put it however you want.
It's super good!!!!! 😮
I'm sorry. I know it doesn't make any sense. 😂 I just put videos together and then added yodeling. 🥲
Whatever do you mean? We're all crazy people here, so we understood it perfectly. 😵💫
finally got through a first avatarfx video, played it safe the first time
Here's that video project I've been working on @daring matrix @hard umbra @ornate pebble. All character animations done in AvatarFX and combines 14 video segments into one 3 minute video. Can't share the full video here as don't have Nitro, but the full high res version is on YouTube (link shown at the end). Song created on Suno.
Based on my character Violet @ https://character.ai/chat/kazA8mHz9s379uF9JtvrpIMy9fAQsiZdcb1O5IJx3PE
https://share.character.ai/9j2b/k4aq8mmo
The mask... 
AI chat, reimagined.
Character.AI isn’t just another AI chat app—it’s an infinite playground for your imagination, creativity, and exploration.
Dive into immersive conversations, create your own AI chatbots, and join a vibrant community of creators, just like you.
• Text or talk—your next adven…
I don't know if it's just me, but here's a screen shot of what I get when I click on that link. I'm on my laptop and the link is trying to direct me to the iPhone app.
Ah... It's probably because the feeds are only available in the app
Ok, it works from my phone. Just strange. Because previously on my laptop, it would open the web page view.
Nice! Just followed you too!
Hmmm. You're right, the video used to appear for me directly here when it was posted and now it doesn't anymore.
That's sweet, thank you. I followed you back. 🙂
Makes me think of the one I did this morning 😆
Gahahahah
Funny the power of a accents. To Americans, he's intriguing. But then give him a Middle America accent then suddenly he's just ho hum. 😂
Russian accents included! Americans notice that, but no accent, she's just some woman. 😂
Finally I was able to create something
So.. Sharpness IV?
Lore/Analog Horror videos of my OC i made with the AI voice thing for a Zombie Apocalypse Minecraft Server :3
Everything apart for the AI voice is original and or made by a real person
AI Voice is just a placeholder, will get the voice actor i want to voiceover these.
Is it obvious that my OC is literally an inane person
!! WATCH IN ORDER !!
-# TW for #4, gagging and zombie sounds.
Ok, that is very nice. But here, the purpose of the thread is to share content relating to AvatarFX. Not just any video made by users. 😅
oh im sorry are you a moderator? Ive shared videos like these before (#🎬︱avatarfx-streams-share message) and had no problems, this WAS made with the avatar FX just the audio part of it.
(It came out a little creepy)
I see an influence from the holotapes in Fallout, notably the various research notes you find into the Forced Evolutionary Virus (FEV) that ultimately gave rise to ghouls, super mutants, and certain other creatures within that game universe. That and the Xenomorphs from Alien which gestate within human hosts.
Everything has its influences of course! I'm currently working on something that in no small way IS influenced by my long time fascination into renaissance and medieval history and epic Tolkienesque fantasy depictions of heroes.
((are you talking to me i dont wanna seem stupid this message is just really close to mines))
Yeah, more or less. Just my reaction. I immediately thought of the holotapes!


