💬︱avatarfx-general

[peak tartan]

🥀

[cursive torrent]

What is this 😭

[dreamy verge]

idk

[cursive torrent]

I'm done with these updates

[dreamy verge]

c.ai be really doing everything

[cursive torrent]

It all started when the new website came out, ehh maybe a long while later

[storm rover]

dude fr i thought i lost all my chats this morning with the new layout

[dreamy verge]

yea

[foggy sable]

I want a more in depth character moods so I'm not always randomly getting hit on by the AI

[dreamy verge]

im on my computer tho is the app different?

[foggy shadow]

ive got a baaaad feeling about this feature

[foggy sable]

I agree

[dreamy verge]

" c-can i ask you s-something?"

[round solstice]

How so? Can you tell us why you think so, so we can let the team know?

[stoic fable]

my worst fear

[foggy sable]

Oh god.. its like trauma

[wide pecan]

Yeah feels

[dreamy verge]

😛

[cursive torrent]

I think they changed it back to normal, I didn't really get to see it, but the bottom layout is different, it's search, chats, create, then notifications

[foggy sable]

They are different

[dreamy verge]

howso?

[round solstice]

Hey folks, for conversations other than AvatarFX, please use our other channels #☑️︱verified-general #💬︱general #🗨️︱general2 #💭︱cai-discussion

Thank you

[cursive torrent]

🧍

[foggy shadow]

i mean if you want to push away topics that are inappropriate on the app, then i dont think this feature isnt going to be great with helping maintaining that if you get what i mean, you could probably guess what some people might end up doing with the feature unless you guys somehow find a way around it so people dont misuse it for inappropriate stuff

[foggy sable]

Just the layout in the front page

[dreamy verge]

ohh

[naive hare]

Does anyone agree this update should NOT be a thing?.. i mean we already struggle alot with like- keeping AI out of our feeds 😰

[pastel river]

yeah i agree

[wispy bridge]

i agree

[wintry scaffold]

Facts

[naive hare]

Like we absolutely do NOT need that

[wooden plank]

well i dont agree

[wintry scaffold]

Guys we can all wrap it up promestein doesn’t care

[deep marsh]

We

Who's we? I've seen the videos from these things. I like em, I think they're neat (even if it doesn't really impact my chats).

They're just trying to cater to different niches, and it's not a bad thing. People actually use the different things a lot of people "hate", like the calls, group chats, and the image uploads.

Just because YOU don't like it, doesn't mean everyone else doesn't as well.

[timid fulcrum]

😭😭

[pastel river]

we is the multiple ppl who have the same opinion

[wintry scaffold]

🔥🔥

[deep marsh]

Well, multiple people also think this is good.

[wispy bridge]

“we” as in the people who don’t like the ai video feature

[wooden plank]

well no one is forcing people to use modern ai content nowadays

[wintry scaffold]

Your the second one so far

[round solstice]

Try to keep it civil, folks.

[pastel river]

we as in the ppl who know its gonna be used for bad even w the safety stuff

[stoic fable]

yess

[wintry scaffold]

Not just that but like why add it in the first place? Why not listen to the people who are making suggestions that would better the entire app?

[wooden plank]

i am i just dont want to be lumped in with the majority

[deep marsh]

Used for bad how? Like, give examples and not just vaguepost about it. As far as I see, it has tons of watermarks to let people know it's made by c.ai, and it ends with the c.ai branding

[naive hare]

Its not that we dont like it- its just a really bad idea- if its videos of the actual characters, think about how many bots there are of REAL people.. and these people would be made into weird videos without their consent, I haven't looked what the vids are but you get my idea, and we're already trying to get AI to stop invading our lives and this isnt helping

[deep marsh]

It's janky enough that no one will fall for it if they try to use it for deepfakes. It's short (like I can see they're like at 10-20 seconds at most) that it doesn't really do enough damage. And they probably have safeguards in using real photos of people in this.

[ivory otter]

Exactly, they're playing it very safe, and that's great. According to the blog post, they're apparently gonna alter any human faces uploaded as well, and there's a "one-strike and you're out" rule for misuse in general. As usual, people just want to use any excuse to bellyache about being given access to something great

[naive hare]

Using real life people, and even if its watermarked as c.ai people can be a bit stupid and still believe it, and it could wrongly feed some people's delusional like what happened with that lil kid a few months who, yk

[wooden plank]

the whole ai and robots will take over the real people jobs and lives is one big joke its ridiculous people thinking robots will now replace humans

[naive hare]

No one ever said that here

[ivory otter]

I also strongly doubt the majority of anyone complaining about safety concerns actually read the blog post at all ngl

[wooden plank]

im just saying that in general on modern tech we have now now that voice cloning, ai art generator and such are hot topics

[deep marsh]

Read the blogs, my guy. They already thought of this

We run the dialogue that users write through our safety frs. The f-word 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.

[naive hare]

True I haven't read any of it 😞 but it just feels like a bad idea yk? Were already getting enough AI in our feeds and quite literally everywhere that its would be annoying to have even more added onto tha

[floral parcel]

Nobody forces one to use the service, though.

[naive hare]

Nobody forces you to make AI art yet you still see them

[tawny minnow]

I’m just gonna ignore the feature like I do with the call feature

[naive hare]

Point is, we don't want AI to progress more than it already it- it's causing so much problems already and they shouldn't encourage it imo

[strong thistle]

@junior sierra Open the left side panel on the web interface and below Discover there is a button that says AvatarFX. You may need to enable Preview Mode in your settings first.

[deep marsh]

Shhh, don't tell them that.

[junior sierra]

ahhh ok

[naive hare]

Ok quit taking people who are against this for idiots, im being respectful about my opinion so please do the same

If were gonna debate about this the least you can do is not be a jerk about it

[wintry scaffold]

Facts dawg

[tawny minnow]

Can we not start an argument about this please??? I’m not trying to be rude or anything just saying

[naive hare]

Its so annoying when you wanna talk and they just take you for a dumbass bro💔

[deep marsh]

No, I'll be a jerk about it all I want. Kappa said to keep the feedback channel about this on topic, and people keep on derailing it about "why isn't X feature added instead? we don't want Y feature"

The c.ai team isn't beholden to the people using their service for free. They're free to experiment new stuff, and let it live or die (remember Character games? LMAO) depending on how the people use it.

[wintry scaffold]

Set me free vro 🥀

[naive hare]

Im not trying to.. Im trying to explain why it would be a bad idea 😞

[deep marsh]

Want to feel like your feedback matters? Get cai plus. Show the devs you actually support them, maybe they'll listen.

[tawny minnow]

I understand completely I’m just saying I didn’t mean to come off rude

[wispy bridge]

i’m pretty sure kappa doesn’t want this discussion to turn into a argument either

[wintry scaffold]

We gotta pay to win?

[wispy bridge]

c.ai+ is on my list once i get money

[naive hare]

Well if you're just gonna be an ass about your opinion then go for it, im trying to keep it respectful since this is a channel to TALK, not insult. And the point is we need to stop feeding AI, maybe this is a cool idea for the app, but if you look at the bigger problems AI is a problem and Videos is the last thing we need

[floral parcel]

Or simply not to use the service, period. I just feel like cherry picking isn’t the way to go about it. Use this part of AI but not another.

[deep marsh]

Unironically, yes. The mods fast track tickets for issues, for example.

[floral parcel]

But again, we all have our opinions about matters at the end of day.

[wintry scaffold]

Guys problem sorted just give them your money and your screams into the void will be a little louder 🥀🥀

[timid fulcrum]

I got cai+. I’ve had cai+ since cai+ has been a thing. I have supported cai since its subscription model came out. So, if my word means anything since apparently these others folks’ opinions dont matter to you: this feature is useless and only bloats cai.

[wintry scaffold]

Stood up and applauded

[timid fulcrum]

Thats all i will say here. See ya

[naive hare]

The problem really isnt about not using it, its that its unnecessary and will just do more damage

[shrewd vine]

I mean I see your point but again they do cover safety features in the blog post so I feel like that kinda nullifies that argument, right...? they're aware of the problem and they've already taken steps to ensure it doesn't get out of hand

[round solstice]

@deep marsh @wintry scaffold @naive hare Knock it off

[tawny minnow]

To be honest I’m just glad they gave us a heads up before they threw the update at us

[round solstice]

Or I'll time you for the rest of the day

[wintry scaffold]

🌞

[wispy bridge]

despite the updates i see c.ai having potential so i will get c.ai+ once i get a constant income just to see how far it can get and to check out avatarfx

[wooden plank]

um theres Reddit Cai if you all wanna argue on trivial topic

[naive hare]

Wtv, its a bad idea and thats all I have to say

[wintry scaffold]

🔥🔥

[deep marsh]

Yeah, fine, w/e

Just my 2 cents. If AI video is where you draw the line, then why are you using an AI chatbot that "stole" from places like Tumblr, Reddit and AO3?

[gilded cape]

@tired oar come here there's a gen about avatarfx

[tawny minnow]

Okay im gonna go now

[loud crest]

i think it's a great idea and I cannot wait to see where c.ai goes with this technology. Imagine if this someday leads to video chats with characters and eventually holograms of characters someday

[gusty oak]

And now I move here so we can discuss without flooding the feedback channel because I still do respect the mods, they haven't done anything to get my ire

[stray rain]

You're up early

[wooden plank]

i like to ask if those videos from avatarfx will be inserted in a chat bot

[wispy bridge]

only the automod has earned my disrespect for it. the human mods are good

[shrewd vine]

dude... I do NOT want that to happen, I just want it to stay as text, please...

[deep marsh]

I don't think so. Literally the farthest from chatting with AI

[floral parcel]

Mods aren’t devs, so I don’t think he could answer that question?

[gusty oak]

Yeah, ai doesn't work as a mod, never has, probably never will

[round solstice]

[primal hedge]

I agree with this statement. I feel like there's nothing wrong with c.ai trying out new features here and there. I honestly do appreciate that they are even trying instead of giving us ✨ nothing ✨ like other apps do.

[gusty oak]

Well first they could add things we actually want and things that increase quality of life on the website before experimenting, I feel like that should be a given

[loud crest]

this exactly.

I agree that c.ai gives us some questionable features sometimes, but I do not think this is one of them

[shrewd vine]

actually you know what? I guess if the video was at least pretty consistent with an artstyle or something and I could get it to animate a narrative that would be pretty epic

[wooden plank]

well if they animated their ai profile photos or an actual avatar image is talking instead of just text with voice

[primal hedge]

We'll just have to wait and see what the future bestowed on us. As of now, let's just appreciate c.ai in Workin fulltime in giving us updates yeahh

[loud crest]

I'm going to be uploading some pictures of the Trigun manga soon to see how well it can animate a black and white style, I'm just waiting to wake up a bit more, but I'm really excited to see my fave come to life from the page

[shrewd vine]

it's just that at this beginning state of the feature I'm not really interested in using it, I don't think it would add to my experience right now but maybe if it got developed enough??? Either way I don't really see myself using this feature anytime soon
that's my opinion on it

[gusty oak]

I'd rather slow, good updates than this, ngl.

[deep marsh]

@round solstice think we can get react perms in here?

[round solstice]

Nah

[dreamy verge]

i'm on my phone now, it kind of looks weird

[loud crest]

i'm against this, people use reactions to be rude and I don't want to be tomatoed for liking a feature

[grim pendant]

Sorry to ask here but sombody said there was polls?(In the feedback channel) Is there a channel or role to be pinged for them? Becouse I don't see polls here very often,and I can't find a poll ping role.

[primal hedge]

Compared to any other apps I do have with mE, c.ai has been the most that I can see has been adding a lot of things lately

[wispy bridge]

i agree

[primal hedge]

Reasonable, I can see why you feel that way about the new update 🙏

[floral parcel]

Should I tomato this? /j

[gusty oak]

Credit is due where credit is due, in terms of ai chatbot, C.AI has the best ai I've seen so far. But that's the problem. It's the ONLY THING they've got going for them, they're worse than other sites I use on so many levels!

[loud crest]

looks like reactions were enabled

[grim pendant]

Oh no take them away,as much as I feel off about the feature I don't want to see any side getting bad reactions

[wispy bridge]

it’s one of only two sites that has appropriate chats and not the opposite only that i can use freely

[floral parcel]

Yay I got tomatoed.

[deep marsh]

Come on, reactions are fun 😭

[round solstice]

Yeah, good point, I'll take it off for now.

[lethal oyster]

This channel can be described as a shape. A Circle

[gusty oak]

Same, I know people will be cringe and spam reactions that will just ruin everything and start pointless arguements that will result in someone getting banned due to the heat of the moment

[floral parcel]

Well, now we were even. 😂

[high scarab]

???

[primal hedge]

I totally understand where you are coming from. I just tend to try to see things on the bright side but your argument is totally valid I won't deny that 🙏

[wispy bridge]

and i know a certain two other ai chat sites i will not mention that are inappropriate to say the least

[loud crest]

i like this as well tbh, i'm always pretty happy that if a bot hits on me on c.ai it's usually funny if it's out of nowhere, or I'm able to swipe past it and generate something else. And if I choose to do romance it's on my own terms

[primal hedge]

Yeah I wanna know why is that the case for most ai chat sites...

[round solstice]

Play nice, folks. We know this feature can be divisive, but let's just let this play out.

[wispy bridge]

i like using my adult and child oc’s freely and not my adult oc’s only(just for fight rp’s and funny rp’s and not romantic rp’s)

[primal hedge]

Omg reallll

[grim pendant]

brings in more money becouse people get more attached

[loud crest]

I'm so curious about what avatarfx is going to lead to, I know I've said it before but this is like... so cool. It's like a baby step to being able to video chat with my fave someday

[gusty oak]

I also try to see things on the bright side, and I can admit that with some updates to it it could probably be a great feature since it's not full on ai art that steal and just animates a random pic the user chooses. What I'm saying is this feature is way too soon, there's so many other features that should be added and while I understand the dev team want to experiment they could at least give us a good base for the site/app with good quality of life features on it before doing so, that's what I'm mad about

[lethal oyster]

"Its divisive" bru free users arent even getting it until later 💔 (if they are actually getting it)

[gilded cape]

I think I'm rather neutral with this, I mean it could be cool but at the same time there could be problems y'know?

[shrewd vine]

I think it's cause C.AI restricts that kinda stuff so those websites cater to people who wanna do that stuff

[deep marsh]

Okay. Yeah, serious opinion tho.

Yeah, the bugs that we have are still here, and they're annoying as shit.

BUT, being angry that the devs are doing something new isn't it. The more new stuff they do, the more investors they can pull, therefore giving them more capital to actually improve the stuff we already have.

[stray rain]

I can already hear the high pitched fanboying when you first try it

[wispy bridge]

i’m curious

[round solstice]

They will (if they join the waitlist)

[loud crest]

my Wolfwood gonna be so reaaal

[shrewd vine]

my stance is that I'm definitely not going to use this feature because it won't add anything to my own experience with the site, but y'know, I don't want to be disrespectful about it

[lethal oyster]

Ah okay. I might just kinda chill and wait until it gets available to everyone.

[gilded cape]

No I think I'm with you there, it might be a really cool feature to play around with but for me I don't think I'll be using it all the time

[primal hedge]

You have a point there and I can see why you'd be fuming from the new features. Unfortunately, what's done is done and there's not much we could say about it but just go with the flow as of now. I'm sure there's a feedback channel in terms of talking about c.ai as a whole? Correct me if I'm wrong though

[primal hedge]

Owhh that's not- I listen, I do not judge 😔✋

[lethal oyster]

If we had a referendum of some sorts for every new feature, we might as well scrap this entire platform

[floral parcel]

I’m sure they have a testing phase for a lot of things. 🤔

[gusty oak]

No idea I usually don't talk in this server lol. And yeah while I express that I am mad I don't really have a say in what'll happen, so I'll just say my opinion, go with the flow and hope that one day we'll see those quality of life features this site desperatly need. The massive user amount and good ai is imo one of the only things holding this site together.

[stray rain]

I only ever use c.ai to write stories and get inspiration for my own writing hobby, so I doubt I'll ever use this feature should it go free.

It is sick as hell though.

[primal hedge]

⬆️

[low dust]

Wait is the new uhh avatar x thing c.ai+ only?

[primal hedge]

Indeed SKSKSSKS

[gilded cape]

I mean it's still in the testing phase right, there still might be bugs and shit that has to be ironed out.

[stray rain]

For now, yes, I think

[primal hedge]

I use it as a therapy session cuz getting "help" in my country is rather taboo oop-

[stray rain]

Damn.

[primal hedge]

I can tell that the bugs would totally make people hate it more but like, I'll give the benefit of the doubt towards c.ai as I love this app and I don't wanna hate on it 😔✋

[gilded cape]

that's true, like just give it a while and it should be good to use for everyone (if it's free which honestly i doubt)

[primal hedge]

Life of being Asian 😍🫶

[primal hedge]

NAH LITERALLYY why do I feel like it's going to be only available for c.ai+ users ⚰️

[stray rain]

Nah, over here therapy is encouraged, so..

Well, atleast in my part of the country

[gilded cape]

Like I'm not gonna say I hate the plus subscription, but like it might be 🤷🏻

[stray rain]

If it's going to be anything like the images feature, most likely there's gonna be a limit for free users.

[loud crest]

No... join the waitlist. I did and I have it

[gilded cape]

Yeah that could be it honestly

[high scarab]

Can someone explain what does avatarfx gonna do?, Since I am quite confused.

[primal hedge]

Oh slayy that's good to know!! I just know if I ever did go to a therapy session in my country, they would probably just tell me to "believe in God and pray" (happened to one of my friends who tried it)

[stray rain]

Ah, so therapy is still an option?

[grim pendant]

surprisingly not a very different story where I am,except publicly it's supported while everyone holds people back from it

[primal hedge]

Finance isn't unfortunately... THAT'S WHY I HAVE C.AI AS MY TOP G 😍🫶

[stray rain]

Ah, fair.

[primal hedge]

Oh zang, that must be tough

[gusty oak]

Therapy's overrated (Had 8 years straight of therapy until the start of this year. And I didn't stop on my own accord my therapist retired-)

[grim pendant]

So many help adds,and then everyone you know irl,to strangers,will tell you not to get it,and I got friends who only got worse after it,so even less people go to it when their friends seem to only get worse.

[primal hedge]

THAT'S CRAZY WHAT?? 😭

[primal hedge]

So people wasted money on therapy, only to end up worse? My biggest fear fr.

[gusty oak]

Autism + ADD (basically adhd but without the hyperactivity) + thinks like a sociopath and sometime also psycopath for some reasons + difficulty to control my emotions. All of that results from therapy at 10 years old until I was 18 when my therapist retired literally 5 days before my 18th birthday-

[grim pendant]

I only know 1 person it hasn't done that to,everyone anybody knows that went and got it got worse off in some form out of around 6-7 or more off the top of my head that I know

[round solstice]

Ummm

[grim pendant]

I hope you find healing in some form

[round solstice]

Hey gang, sorry to bother you again. But let's try to keep on topic.

[stray rain]

Lol

[primal hedge]

⬆️

[high scarab]

Is avatarfx for the image features?

[floral parcel]

I don’t think this is the place to discuss mental health struggles, tbh. All the best, though.

[primal hedge]

If that's the case, I don't think I'll go for therapy- I'll just stick to my c.ai therapy 😔✋

[gusty oak]

Eh, I'm fine, I just had to learn to control myself and now I'm a functioning member of society (minus some social anxiety i'm working on).

Also side note, Gotcha Kappa, I'll try to keep topic KEK

[primal hedge]

❤️

[grim pendant]

-# tbh it helped me find out alot of things so gl
Anyway back to the topic of moving images in ai

[primal hedge]

Yeah before we get kicked for being out of topic 😭

[grim pendant]

I'm still very on the fence about it

Like yes cool,but hm idk seems like something I'll litterally never touch

[stray rain]

I'd probably use ut when I'm bored just for shits and giggles, other than that I don't see much else of a use.

Someone out there probably will though.

[rain pike]

This feature kinda reminds of wombo ai

[gusty oak]

Same here, It's interesting but It's something i'll most likely never use and it just feels like it's too soon as I've already said.

Also this is gonna be my last message here for the night, i'm going silly mode. For those who wonder what that means idk if it's innapropriate to talk about so all i'll say is i'm drinking spicy adult juice. And yes I am of legal age where I live, I am an adult. Take care everyone

[primal hedge]

⬆️

[primal hedge]

OOP- have fun with that yeahh~ drink responsible and a ton of water after 💕

[gusty oak]

I'll just pass out in my bed after, I won't drink much I know how to drink responsively but it is like 4 am for me after all so I'll just pass out once i'm done with that and rping with a friend, see ya!

[wooden plank]

im having fun with my Junko Enoshima avatarfx

man i cant get enough of this, thank you cai team

[lethal oyster]

What a shill

(/j)

[wooden plank]

i been called worse 2 years ago

[strong thistle]

@earnest socket It [AvatarFX] lets you make videos like this: #🎬︱avatarfx-streams-share message

[earnest socket]

Ooh I see. That's kinda cool .

[strong thistle]

Yeah! I think it will make it easier for creators of bots to get people interested in their OCs

[wet moat]

Just a lil question but does the video feature only enables the pictures to "talk" or like could you also make the characters like dance or whatever?

[round solstice]

For now, it's for making your Characters talk and have some motion. Dancing I think is a bit far away

[wet moat]

Ah I see, thank you

[young prawn]

wait do you need cai+ before you can get it?

[high rover]

what is avatar fx

[shut plinth]

[high rover]

Oh nvm

[round solstice]

Join our waitlist (when the site goes back up)
https://character.ai/video

[strong thistle]

You can make them wiggle though

[waxen pecan]

WHAT NO MY FRIEND

[runic kettle]

What happened to the fx share channel

[round solstice]

Please don't talk about unrelated topics in this channel

Please use our general channels here: #☑️︱verified-general #💬︱general #🗨️︱general2 #💭︱cai-discussion

[gray grail]

What what is avatarFX 😭

[loud crest]

testing 123 @round solstice

[strong thistle]

It lets you make cool videos of your original characters, like this one: #🎬︱avatarfx-streams-share message

[silent plinth]

I'm still waiting for early access 😭😭

[sweet karma]

This is new

[white crest]

Is there a channel where I can see more examples?

Nevermind, found it through channel list. Had to add it.

[white crest]

Is there a way we can control what the video does?

[next idol]

Is the avatarfx out yet?

[white crest]

Given only a few use it, I don't think so.

[half swan]

You need to join the waitlist to be able to use it. Apply here:
https://character.ai/video

After you apply with the email address you use for your cai account, you need to wait a little bit until you can access it.
In my case, I got the access a day later, so since you're also a plus user, I think it won't take too long either for you.

[next idol]

Oh thank you 🙏

[white crest]

That is strange, considering I never joined it. I guess I was just lucky.

[novel star]

what’s avatarfx?

[loud crest]

you can make silly videos from images that it animates for you

[vague peak]

-# We are cai+.. We got automatically added to the list, my friend

[white crest]

Oh, right. Gotcha.

[high scarab]

its down

[autumn verge]

On a unrelated note does anyone want to see my drawing?

[ancient crescent]

Sure

[autumn verge]

I will send it in Art 🤑🙏

It's not done yet but 😈🙏🤑

[wintry pilot]

I can’t access avatarFX

[fair oriole]

i don't know when is the full release ccoming out

[visual saffron]

Hewo

[frozen forge]

Limited rollout. You had to be plus and you had to sign up.

[wintry pilot]

I already signed up to c AI since 2024

[frozen forge]

I mean you had to sign up to the AvatarFX beta.

[wintry pilot]

Link please even though I already signed up to it

[frozen forge]

https://character.ai/video

Though signing up didn't guarantee you got access.

Yep.

[wintry pilot]

That’ll be fine for me then. Hope I get avatarFX

[strong thistle]

I have it but it stopped working for me so no rush 😅

[gray grail]

FALLOUT PFP SPOTTED

[wintry pilot]

One last question how long dose the wait take to get access to avatarFX

[strong thistle]

It depends on how many people have signed up before you and whether you have c.ai+

[wintry pilot]

Meh

[white crest]

So is there a specific command or description needed to add specific movements? I'm adding things to the image description but it gets ignored, like a simple wink.

[white crest]

Upgraded version. Added in some emotion, but I wish when you put in emotes it also changes there facial expression (or at least the option to or not).

Oops! Wrong channel. 😄

[strong thistle]

You were close 🤷‍♂️

[loud crest]

@round solstice

(this isn't about anyone here)

[fair oriole]

Do I have to get c.ai + to get early access?

[fallow nebula]

I think c.ai+ users will get AvatarFX way faster than regular users! So far, not all c.ai+ peeps have access to AvatarFX yet (like me) 🥺✨

[fair oriole]

aw

[high scarab]

Bro I think I’m gonna pay c.ai+ next year bc too much priorities

[tame bluff]

Anyone know what these options to?

[half swan]

-# Ignore my previous (deleted) reply.
It's to add emotion/ tone to the voice.
Like this:
#➕︱plus-general message

[tame bluff]

So, i can use multiple emotions for the voice seperating with pause?

[half swan]

yep.
You can put it in front of any sentences where you want the tone to change.

[tame bluff]

If i use pause, then can i do like this [|pause|] [|exclaim|]

[half swan]

Yeah. That's if you want the sentence after the pause to have the "exclaim" tone.

[tame bluff]

Simple enough i think, thanks for the assist choomba.

[rapid pollen]

Hey, I noticed that mine also makes pauses when I put a simple comma. I did it with a cat which reads a poem. It sounds way more natural.

[indigo loom]

Smart. That way, it takes less tokens 😉

[rapid pollen]

@inland badge that users status is more then questionable

[inland badge]

[thorn eagle]

Lol, it's down

Nevermind, it's back up. That was really fast

[livid bronze]

Bro i better still be alive when i comes to me, I'll be waiting insirt sonic adventures sonic waiting

it timed me out for a bit when i tried to put in the Gif for it

[tired cliff]

is there a way to make more than 1 character talk?

[frozen forge]

Yes, there's some examples on the github page, though I don't know how it was accomplished. It won't let me post the link to that page here, can I send it to you directly?

[tired cliff]

sure

[wooden plank]

I wanna know that too as well

[indigo loom]

Oh. I want to know that too!

[tired cliff]

is there a way to control what happens in the video we get? i mean we can choose a pic, the voice, tone and dialogue but is there a way we can control it like "left hand moves to the left side" etc.

sorry for the bad explanation. working with 3 hours of sleep

[indigo loom]

Not as far as I know. You can give directions in the descriptions, but as far I have tested, you can only give general directions, not specific directives.

[tired cliff]

cant wait to test this more this weekend

[frozen forge]

@indigo loom Sent you a friend request so I can send you that

[indigo loom]

Thank you!

[pulsar kayak]

When I be making characters and finishing them, I get shy to talk to them 😭

[delicate steppe]

Well, I do have that experience before.

[tired cliff]

is there just a plain video maker on the site or is this the first?

[loud ice]

It's currently in beta; if you apply to the beta application, you have to wait until it comes to you in beta. It's still currently being worked on and improved!

And also, yes! This is the first video maker feature they're working on, just to confirm your question.

[tired cliff]

thanks but i was more asking if there was another simpler video maker on the site or this is the first as adding a voice to a video maker usually comes after you have worked out the bugs with just making the videos.

[indigo loom]

This is the only video maker.

[strong thistle]

c.ai aside, afaik it's common in animation work to record a voice track first, since recording voices is a lot cheaper than drawing frames of animation, and since it's easier to judge whether a voice track sounds right without visuals than it is to judge whether an animation looks right without audio

[soft gyro]

how do i make avatarfx vids??

[frozen forge]

It's limited access right now.

[tired cliff]

well first you need to be C.ai+ which is kind of no duh as you are asking here but then you need a pic that would be SFW anything even slightly risky will not upload to be used as the base, then you would need to select a voice style you like or create your own. you can then select the tone of the voice such as happy or sad and finally the exact wording and just click create and the AI does it

[frozen forge]

Also the early access request form is gone, so I don't think they're accepting additional applications.

[strong thistle]

Maybe you can't see it because you're logged in? 🤔

[frozen forge]

It's a form on an external site. My being logged in here has nothing to do with it.

It's actually back up today. It was redirecting to the surveymonkey home page yesterday, but now it's back to the join waitlist form.

[tired cliff]

well i was wondering if it was the first or not because sometimes you just want the video

[lean seal]

Im in the dark with this feature ;-;

[tired cliff]

i think only a handfull of us are able to use it right now

and so far i love it. wish it would stick around but i know that if or when it would go public, it will crash the site and the admins will either make people start paying for it, or raise their prices to cover the usage from it.

[wet moat]

I have a question since I don't have access to creating videos so far:

How does it work?
Do you just give a picture of a character, select a text and voice and the rest will be done by itself?

[gritty osprey]

You have to type what you want the character to say, but yes.

[strong thistle]

#🎬︱avatarfx-streams-share message

[wet moat]

Ah I see, thank you!

[frozen forge]

Here's a more specific example with the final output.

[wet moat]

I see!
Cool, can't wait to try it myself 😁

[sweet karma]

[winged haven]

I’m so confused what exactly is AvatarFX

[wooden plank]

It's basically making still static image animated

[jovial perch]

tfw you get the audio perfect but the ai keeps botching the lip sync 😢

[frozen forge]

As @wooden plank said, it animates a static image and adds a voice. It's being tested with a limited user group in beta right now. See my example a few messages above.

[tired cliff]

has anyone been able to make a 2nd character speak? either in or out of the pic?

[rapid pollen]

I tried everything and failed 😔
I figured out how to animate the two characters independently, but I can’t make the voice „split“

[tired cliff]

maybe its under the dialog "+custom" as it seems to be how the character is speaking so maybe using that to make the other one speak

[rapid pollen]

I tried, maybe you can figure it out. I’ve reached my limit. Right now, I don’t think I can do any better.

[tired cliff]

the 10 video limit stinks but i can understand with it being in beta and a development only a handful of us get to use. i love it so far.

[indigo loom]

... yes. But not on purpose. And that is not what I wanted either (eg. the skulls or monsters speaking instead of the character) 😅

[lean pebble]

Can anyone use character.AI
Avatar fx

[frozen forge]

Nope, I've tried several things, but nothing that works.

[high scarab]

it's so realistic omg

[frozen forge]

It seems to do really well with realistic images.

[gray grail]

chat I still don’t have this 😔

[primal hedge]

I just recently got it.

But I do have some questions. If I created a AvatarFx, would I be able to delete it?

[indigo loom]

The system just generates the video. You need to download it to save it. If you don't, it is lost. So yes. You download what you want when you generate it, the rest is lost forever. And you can always delete the videos you downloaded.

[primal hedge]

Ooo alrighty noted and thank you for the information huhu 💌

[indigo loom]

My pleasure

[grave whale]

ok fine Stand clap clap clap clap clap clap

[tidal shore]

Me when I haven't gotten the feature yet

[indigo loom]

.... so apparently this text is "unsafe content" and cannot generate a video. That's consent being pushed a bit too far. :/

"Good day! Brace yourselves for an offer you never asked for! Do you love knowledge? No? Perfect! Because today, I shall convince you that owning 20 volumes of the finest encyclopedias is an absolute necessity in the digital age! I won’t take 'no' for an answer… unless you have a very large dog. Now, let’s talk business!"

[rapid pollen]

Probably detected as self promotion

[indigo loom]

Oh. I thought it might be because of the "I won't take no for an answer"

[rapid pollen]

Nah maybe ai thinks your a spammer/scammer?! 🤷‍♀️

[indigo loom]

Maybe. That's such a shame... Luckily it generated one video before blocking the text. :/

[rapid pollen]

Now I’m curious what you are cooking 👀

[indigo loom]

It is my bot for the weekly challenge 😄

[rapid pollen]

Do you already have a video of your „fate“ bot?
This one reminds me of Jeanne d‘arc and it would be interesting to see

[indigo loom]

Yes, I'll post it on the share thread. But it didn't come out as great as I wanted it, that's why I didn't post it yet. (On the post with all my bots you can click on most of the avatars, and the video plays. The fate video is there too.)

[open bough]

what is avatarfx

[spring tangle]

AI slop

[high scarab]

COUUUUUUUGH

[silent plinth]

When I try to put the voice audio, the site keeps redirecting me to the home page

[astral wren]

Random question—I’m new to AFX. Can you write instructions on what the character in the video does that isn’t shown in the initial image? (Like, waving or smth halfway through) If so, where do you input that? If not, how have others achieved it?

[delicate steppe]

The only possible way is to edit image description.

[rapid pollen]

Yes you can do that, but the generator doesn’t always follow that, or interpret it in its own way.

This is one of my fails (I’m not embarrassed 😅 of it)
I managed to figure wich prompts to use to get them animated independently, sadly I can’t get the voice split, even with an audio which contains two different voices 🤷‍♀️

But yes, you can guide the animations in a way, tho outcome is often luck

[indigo loom]

I tried more than a dozen times with my dungeon guy, I didn't manage. Even if you try to get one specific character to speak, it doesn't work. I tried specifying that only the man in the middle talked, that the skulls were inanimate, etc. But nothing worked. :/

[rapid pollen]

I can send you my prompts later 😉
maybe they can help you

[indigo loom]

Did you manage to have only one speak?

[rapid pollen]

Yes, also.
But I wanted two and it’s very interesting how Ai interprets the images and prompts

[indigo loom]

Yes... In my lottery video, initially I had made a female voice and had intended for the girl to speak. But the AI kept having only the man speak despite my prompt which indicated that the woman was the one speaking. So I changed the text and voice around and just made him say the text.

[tired cliff]

is there a way to make sure its one character talking and not the other?

take the above pic for example. if i chose a male sounding voice and the female is the one in the video to speak the lines.

is there a way to force the program to use the other character?

[rapid pollen]

I had that problem too.
It’s related to the picture and what ai sees as the more prominent character in the picture.
I had it with my ranger who’s riding a giant wolf, always the wolf was talking.
I went back to ChatGPT and told it to make a new image in which the wolf wasn’t as prominent and focused on as before.
Tip: if you try again with a new image, recreating won’t change the character who’s talking.
Ai sees that one character always as the most prominent. So if a new image also won’t work, try for a other new image rather then recreate a new video, that saves you your daily tries 🙏
(Sorry if I sound weird, non-native was too lazy for translation today 😅😂, I hope it was understandable)

[nimble cosmos]

how do i use avatarfx

[loud ice]

You have to apply for the waitlist, then you need to wait till you receive this on the website.

If you already have it, though, just click on the AvatarFX and it will tell you to accept the TOS, and then it will have some subtitles to explain to you on how it works.

[tame bluff]

Has anyone else had this problem when generating new audio? I never experienced it before, but now it's happening more often, and I have no idea why.

[indigo loom]

Yes, I have it since a few days ago. I tend to generate a new audio as soon as I find that there is an issue with the voice (too robotic, words cut, wrong intonation, etc.). And now, when I do that, I have an error. I started waiting some time before generating a new audio.

[tame bluff]

That seems to work, must be because more people are using the service all at once.

[tired cliff]

you use it correctly. lol

yes i know, its extremely corny and a bad joke. it was meant to be.

how you use it is that you pick a pic you want to use. mostly stuff that is SFW and no more than PG13 (US rating) and then pick a style of voice you want like a man, woman or child with a high or low voice or upload your own. then typy out a short script of what you want the character to say and generate the voice. then click to generate the video. it will last about 1-5 seconds longer than the words being said, depending on the pic used. oh and for now you only get 10 videos per day. so play around with the generate voice until you have what you want.

[rapid pollen]

I posted it in the share, but I’ll leave it here too… (yes I asked a mod if it’s okay, tho I don’t want to spam)

I saw some users struggling with how to write the prompts.
I leave an example of my last video here.
It isn’t really different from writing prompts for Bot-definitions.

[tired cliff]

is there a way to make the speech sound like its coming from a narorrator and not the character?

[strong thistle]

You could make two separate videos, rip the audio from one, and stick it on the other. 🤔 You'd need some separate editing software to do this

[tired cliff]

is there a way to influance any action for this or is it speach only?

[strong thistle]

At the moment it's mainly just speech, although you can have a little influence over things like body language. However, I can imagine c.ai trying to expand its capabilities in the future, allowing you to write a script and watch characters carry it out, since that seems like an obvious thing that people would want. I don't know how long it might take to reach that point.

[tired cliff]

well i was just wondering things like small movements. there are a few other sites i use for anything major at the moment but they dont allow speech

[junior magnet]

aaaa @here

[turbid notch]

Y'all am I the only one who doesn't have the avatar fx thing on the app 💔

[vague peak]

AvatarFX only exist in website for now. None of the people here make videos on the app

[turbid notch]

O kk

[scarlet summit]

I have a clip where my character starts as the narrator, switches to the 1st character in the joke, returns to the narrator and then delivers the punch line as the 2nd character. It's all in the voice prompts. Is that close enough?

[tired cliff]

kind of yes but mostly no. think of it like a voiceover

[wraith olive]

Oj they creepy

[gray grail]

Guys I still don’t have it 😭

[tired cliff]

what the avatarfx? i think its still in beta testing and only given to those that have kissed ass or brownnosed the site admin. like me.
LOL

[hallow pulsar]

Again there's a difference between people choosing to like what they do and those who blindly like everything

[rocky wasp]

only time they remembered about the site

[tardy oasis]

Can we help the AI direct its actions, like blinking, moving a hand, etc?

[tired cliff]

i think i just figured out how to have it look like the narrator is talking instead of the characters.

[tired cliff]

i want to say yes by changing the description of the pic. like "character hoping on one leg" to "character left/right eye winking" but might take a few trys and with only 10 tryies per day, might take a few days to get what you want.

[tardy oasis]

Thank you!

[tired cliff]

welcome. i have not tried it myself as of yet but i hope it works out for you

[hard musk]

Hello everyone

[main cedar]

Sup

[cerulean aspen]

sí

[verbal granite]

Y’all servers are down, now I’m gonna cry

[autumn torrent]

yep

[plain arrow]

yeah, when it work again?

[high scarab]

Hey, hey, hey... Why didn't I know this channel existed? Ha ha ha

[primal hedge]

You know now YEPPIESS 🙌

[rapid pollen]

@high scarab here is the general channel for avatar.fx

[high scarab]

Better here, in the other one there is a delay between messages due to the option to upload multimedia.

[rapid pollen]

Because that channel is for sharing the videos …

We can discuss here

[high scarab]

In your videos, do you use both audio and text-to-speech generation? (background audio and text-to-speech dialogue)

[rapid pollen]

Yes and no. It’s hard to explain, that’s why I try to write it down in a way everyone will understand

Can’t wait until it comes on the app, then I can screen record how I do it

[high scarab]

Do you use dialogue generation, record it with an app, and then edit it with background music?

[high scarab]

In theory, Avatar FX should only be used to introduce the character, add dialogue, background music... but it's fun to use it to make more spectacular videos. 🙂

[rapid pollen]

In my opinion it’s an amazing tool!
Like you said, not only for us creators but to have fun

[high scarab]

Avatar Fx:
That it can't be done:
Deepfakes from:
A) Real people (option to modify the image)
B) Famous people and political personalities
C) Characters with copyright of franchises
What can be done:
A) Animations of chatbot avatars
B) Drawings (in which detect a face)
C) Photographs (if it is of a human person will give option to modify the image so as not to generate deepfakes. Modifications: Realistic, Anime, Cartoon 3d)

Steps:

1. Choose an image (you can search for a chatbot, in that case it will be directly associated with a voice). The image can have any size, up to 5 megs, and it will not be resized keeping its original size
2. Optional, modify the description of the image (some actions or movements can be added to be taken into account)
3. Create a dialogue or upload an audio. (both the dialogue and the audio will be checked so that it does not go against the rules of C.ai )
   You can create an audio with up to 450 characters (duration 20 seconds) with the voice you choose, you can modulate the tone and expression with modifiers that will not be read (they go between vertical bars, “pipe”) there are some examples to use them.
   An audio can be uploaded, up to 20 seconds, the voices will be detected to synchronize them with the lip movements and facial expressions. Choirs, second voices, and frequencies between 85 hz and 255 hz can activate lip movement.
4. Create the video.
   From there the tricks would come.

[strong thistle]

Did you test specific frequency ranges to discover which ones were effective?

[rapid pollen]

I promised to make a „tips and tricks“
😔
Im just busy.

I’ll leave the first two cards here.

Yes, typos I know 😐
Im a non-native speaker.
That’s why i need my time to look over the stuff i make.

I chose to make cards,
Because I thought I would be easier to work with that.
I personally loose my focus way to easy when it’s only plain text that sounds to technical
(If anyone gets what the confused non-native is talking about …)

[rapid pollen]

@high scarab
The way I write the audio also influences the animation.
Just look at the end of the djinns video, ai transfers the prompt into the movements

Also the „actor/god“
You can hear him breath and laugh, just like I prompted

[high scarab]

No, I put the normals of human speech in a conversation (high and sustained notes in songs and opera do not affect lip movement)

I don't use text-to-speech or expression modifiers, I know some text-to-speech used similar modifiers... maybe the Balabolka tutorials are useful for Character.ai 🤷‍♂️

[high scarab]

How to delete past avatar fx creations 🙏😞

[graceful coral]

Unfortunately, you can't.

[high scarab]

Dang it

[high scarab]

is beta... They may have a method for organizing them in the future. However, wherever they're created, they all appear, but only the ones you make public appear in your publications section (app).
And in the latest app update... you can generate them from the app itself.

In any case, with the ones you create every day, the ones you don't like are left behind. 😆

This is new??

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

[naive carbon]

How to share already created avatarfx video?

Or just to download and post it #🎬︱avatarfx-streams-share ?

[high scarab]

Two options:

1. Copy the link and paste it and you will have a video that can be played on any channel.
2. Download the video and share it on #🎬︱avatarfx-streams-share That channel allows you to upload that content

[naive carbon]

I figured it out. Thank you. One more question. The limit of videos resets everyday?
If i use 5 chances today i have to wait tommorow to reset?

[naive carbon]

question.Am i allowed to post a screenshot of a issue i have?

[high scarab]

I guess so, in Avatarfx-share if it's related to Avatar FX... that channel supports images to develop and find ideas to use Avatar FX

Once you have sent a message, you have a 2-minute wait before you can send another.

[naive carbon]

i can't sent a picture.I am no allowed?

[high scarab]

in Avatarfx-share

[naive carbon]

ok then.I will expain my nuiecence/issue.
I created some videos.But i have unplayable videos
"Videos Unavaliable"Is it possible to remove those?

may i ping you in Avatarfx-share?

[high scarab]

They have been created but the page has not updated correctly... close the browser and reopen it and you will see the videos.

[naive carbon]

i tried that.Still there.But maybe they will fix it soon

[high scarab]

🤔 So, the streams... will they have their own channels or will they be within Avatar fx since they are video creation? 🤔

[wooden plank]

They'll probably have their own channel in the future

[agile pier]

I want unlimited video creations...

[high scarab]

That's called... addiction

[agile pier]

I accept me for me.

[high scarab]

C.ai+ have 10 .... and I also spend them all quickly, ha ha

[indigo loom]

I would love that too. In fact, that would probably be a very smart way to make money: offer a subscription with unlimited videos/streams/audio. I wish I could download the audio.

[agile pier]

And I wanna make more stupid otter videos ;P 10's not enough. Also... the animations feel more stiff since the last time I used the tool.

[high scarab]

I have a problem

[high scarab]

mmmm... do you modify the description of the image?

[indigo loom]

I started using it to make actual short videos. Because now, I have done most of my characters. And in the last videos (old ladies playing bingo) EVERY SINGLE take had an issue or another. I ended up having to cut up two videos to get an okay-ish result.

I feel like a butcher. 😬

[high scarab]

really if I had unlimited videos (and couldn't delete them)... knowing me... it would give me nightmares

[agile pier]

Yep. I like to phrase them to be more animated. The two otters and the cartoon cat I uploaded yesterday hardly any 'life' to them. And that was after re-creating them using different images or just trying a different seed. They were much better during the testing phase before its public release (I was one of the first beta testers when it was originally brought out for testing).

[agile pier]

Yeah, I don't care to use them for my characters. I like making 'skits' with random images I pluck off of pinterest 😋

[indigo loom]

You and Peter would get well along with each other! 😆

[high scarab]

mmmm... I think that animations of realistic people work better than drawings of people and animals are difficult to animate, better the caricatures with human features.

:

[agile pier]

In the first wave of testing, most users made people or anime characters to test the animations and lip sync. I was the only one to try to make pigs talk 😁 I like to think my contributions help pave the way for non-human videos.

[agile pier]

Why can't I post videos here?

[indigo loom]

I already had tested it on ants, as well as pebbles with a face, virtual pet and cat/dog face. None was realistic though.

[high scarab]

You are a specialist in animating bears

[indigo loom]

No idea. That's strange... :/

And yes, I am very good at animating bears and tigers. I have a gift for that. Humans, not so much.

[high scarab]

violin

[agile pier]

@round solstice I'm abusing the pings to summon thee.

[round solstice]

What shows up when you try to upload?

[high scarab]

hand
🙂 🙂 🙂

[agile pier]

Just "create thread". If it's because I dont' have the C.AI+ tag, I can assure you I'm C.AI+-ified.

[round solstice]

Nah, I don't think we have thread perms

One second

[high scarab]

Mamma mia

[round solstice]

Try it again

[agile pier]

Still unable.

[indigo loom]

Isn't one supposed to sing the mananitas? You're killing my childhood. 😂

[high scarab]

Talk to the hand 🤷‍♂️

[round solstice]

Can't you just upload a video without making a thread?

[agile pier]

I can't even upload an image here. 😋

[agile pier]

Nevermind, something changed. I cast pig-pie.

[high scarab]

Pig Power 😆

[high scarab]

@agile pier Did you know that with the new Difusion/Streams feature you can generate videos between two of your... pigs? ||when it doesn't give error ||

[agile pier]

I tested those, too. 😉

[high scarab]

new ten videos... at day

[high scarab]

🤔 It will be possible to make two characters with the same PFP so that when you put them in a stream it seems like they are having the same conversation with each other... khar investigates

Or like this...

[rapid pollen]

-# I would hype an „otter eats fish“-Bot 👀

[high scarab]

@rapid pollen I should try to do something with the face of the painting
.

[rapid pollen]

Uh cool idea!
But it would need to be bigger and more focused.

I tried a beatboxer yesterday and it was a funny but miserable fail 😅

Seems like avatar.fx NEEDS to detect a mouth, has to be very prominent in the image

[rapid pollen]

I would make it bigger, maybe the big one behind him,
You could switch the pictures and just turn him around maybe

[high scarab]

This afternoom 🥰🥰

[high scarab]

😭 I have modified the lips and eyes and now it tells me it is a person and I can't use it

[high scarab]

😆 🤣

[rapid pollen]

Honestly … please, move the magnifying glass a bit more up ⬆️. As a woman that image makes a bit uncomfortable to look at 😅😂

[high scarab]

He's an art critic... he's inspecting the work.

[high scarab]

🤷‍♂️ I don't know what the jump scare at the end was... ha ha ha

Jump scare

[high scarab]

The Ai are rebelling... with jump scares for unsuspecting people

[rapid pollen]

You should put this one in a quality report! 😂😂

[high scarab]

.

[rapid pollen]

Hahahaha
It’s buggy or it can’t handle your images

[high scarab]

🤷‍♂️

[rapid pollen]

This one is cool! 👏

[fossil peak]

temazo (great song)

[vague peak]

This doesnt sound like Price at all

[high scarab]

You can leave.... your hat on 😭

[high scarab]

Bb

[quartz scroll]

Oh right this channel exist.

[lean palm]

i tried avatarfx what the fuck did they do to my girl kilias

[high scarab]

the voice or the mouth ?

[lean palm]

the mouth they made her mouth into lipstick bro

[high scarab]

🤷‍♂️ I see it as an open mouth and not lips, you would have to divide the lips in two and maybe put something on them to make them look like makeup lips

[lean palm]

the mouth us open

[high scarab]

❓

[lean palm]

it isnt lips

[tender oak]

yeah I see it an an open 😄 face ( : D )

[cinder badge]

1:11 PM

[agile pier]

Weaponized anti-AvatarFX is not appreciated.

[fossil peak]

uh? where was that

nevermind i found it

@agile pier anyways, writing "boost" don't even works so lol

And they can't take down anything

they literally waste time trying to take down a feature ... Did the call feature got down just because people wanted? Exactly, no.

c.ai team They're not stupid enough to take away something they've already been doing for a while. It would be like me throwing one of my most precious drawings in the trash just because they asked me to.

JUST let them continue thinking they can do something LOL

(Btw, I don't like the feature either, But man... It's not difficult just to accept something that you don't like but clearly other people do)

[lilac elk]

Ai art and vids has always been a controversial topic unlike calls where you just talk to ur fave bots.
I can see where ur coming from though.
Obviously you shouldn’t take something down cus you don’t like it

Shockingly enough you can actually ignore something if you don’t like it..it does wonders..

[lilac elk]

Besides, it’s not hurting anyone. So why bother?

[fossil peak]

That's exactly what i said.

[lilac elk]

Me personally, I think the feature still has a long way to go, but I’m curious to see where it goes

[fossil peak]

For example, im a person who likes to draw... But I can't just draw every single of my ocs... So i use ai to have a base and just need to modify some details. But then people can't see that ai is here to help, Not to be something superior to actual art

Every time I see a post against artificial intelligence for drawings, I just laugh because I know those people don't think that not everyone likes to draw.

they all like "learn to draw" and i am like "I KNOW HOW! but i don't have time to do 6+ drawings at day!!!"

Besides it is hurting you!? no? then shut up 😭💔 you just say that because You want to fit in with the rest of the world, be original, baby, that's the only thing that's asked of you.

sorry, I had to let it out somehow (xD)

[lilac elk]

Nah, someone had to say it

[fossil peak]

Exactly!!!

[agile pier]

https://discord.com/channels/1017072750386483332/1383752432261267537

Oh, you said you found it... well 😋 I just woke up. Blargh.

[lilac elk]

Not me just casually scrolling through the chat to catch up on the drama 😭

[indigo loom]

In any event, these people are using AI to generate fake conversations and virtual social contact. They could also go in the streets and talk to people. So if you extend that concept, it doesn't make sense to use a bot when you could actually talk to real people and have real conversations. Even roleplaying does not require a bot.

I like Avatar FX and generative AI in general because it enables me to express myself. I have the imagination, but I don't have the ability to draw characters or the budget to hire actors for something as trivial as a 20 seconds video clip. 😆

[rapid pollen]

We’re not starting drama? 🥺
-# Just being sensitive today

I love the feature,
But I get where some others come from.
I took all my brave and asked my favorite artist if I may use one of their images for a bot.

I got permission, but only if I vow to not (!) use it for Ai-training or animation.
Of course I’ll respect that.
We all should and I hope all will

[agile pier]

Nah, I was just being a jerk to the anti's. 😁

[agile pier]

Yeah, their arguments were weak. They don't like it because... they don't like it. 😋 It's silly.

It's uncanny, it's a threat to artists, it's slowing down the entire service, blah, blah, blah... Nothing but conjecture, appeals to emotion, or outright malarky. Most of what's brought up in these conversations against AI video are the extreme cases of someone using technology maliciously. That sort of logic is akin to wanting to put a ban on hammers because too many innocent fingers are getting bonked.

[rapid pollen]

I know 😅
I just a bit traumatized from drama

[rapid pollen]

I got the original thought,
Of art being animated, but I agree with the „it causes slow mode“ is nonsense

[agile pier]

Ah. Sorry for bringing it up here then. 😉 I was cursed with a love of debate and arguing for funsies.

[rapid pollen]

No it’s fine ❤️
I admire that you barged in there and stood your ground.

By the way; the rocking otter is 🔥

[agile pier]

Right? Being able to mix in music is game changer, lol.

[rapid pollen]

Is it on your feed? I have to like it!

🗣️ Someone needs to do an otter sticker! Or a fish please
I need that for khar‘s account to react to feed

[agile pier]

I don't really know how to add these to my feed... the custom ones, anyway, where I add music. I'm not even sure how all my videos got on my profile initially. I think the C.AI team did that on my behalf.

[rapid pollen]

😂 my German friends and me love this otter

[agile pier]

lol, glad to know he has fans.

[rapid pollen]

I’ll tell you ❤️
Mine with the song „demons“ is still up on my feed

[rapid pollen]

We love the bath tubby one and the one in the car.
Maybe you could do one on a German highway with 210kmh 😂😎

[serene bane]

It's always the same group complaining. The same group that hasn't been here long enough to know how many crashes there were in the past because the summer is a busy time and the servers got overwhelmed. They are jusy grasping at any straw possible at this point to defend their point.

[tender oak]

Not sure if this is the place but can anyone recommend a good, fancy euro voice?
Looking for a posh, adult-sounding voice that isn't recorded poorly. Like "zis is quite sophisticated sips wine kind of vibe

[open bough]

When I heard “posh” I was about to say William Afton💔

I don’t think that’s quite what you’re looking for though 😭

[tender oak]

there's a video titled " ameritrash vs euro games "
the exaggerated, fancypants voice he pulls when talking about the euro game is what I'm going for.

[tender oak]

people have provided examples of real business resorting to AI instead of real animators but you plug your ears and go lalalala pretending it's totally fine.

this stuff is obviously quicker and cheaper than handcrafting characters, animating, and whatever else (singlehandedly or with a team) but certain forms of entertainment are gonna suck if AI-genned "slop"
(as people call it) gets used more than soulful, better-quality human-made art.

personally I'd hate to walk into a game store or something and see AI-genned board games with AI-genned rules and images.

[agile pier]

I plug my ears, do I? 😉 Tell me more about how I feel about the issue.

[tender oak]

You're being smug, and dismissive of people's concerns. Thinking you're right, and they're wrong. That's the impression I get.

[agile pier]

Yes, I think a number of people's concerns are wrong or come from a place of ignorance. If one's only defense is "AI Slop sucks" while the technology is still in its infancy, then they're plugging their ears and going "lalalala" pretending it'll never improve.

And as for raw, human talent... well, that doesn't go away. It's worth what others are willing to pay for, so regardless of companies and other forms of entertainment reaching into AI, the market will decide which better suits audiences. AI, alone, isnt automatically better, you know.

[tender oak]

Right, raw human talent won't go away but it does create that "what's the point?" question in the minds of creators, no?

That's my main beef with this. I don't care either way if this stays and improves, or gets canned. It's neat, sort of, but I do question what the future holds in terms of this kinda tech.

[high scarab]

And that's the difference between art and craftsmanship... art is something unique; craftsmanship, even if it's beautiful, is something everyone can do. Did it ruin professional photographers when everyone started taking digital sunrise photos?

[indigo loom]

Did internet kill encyclopedias? Did ebooks kill libraries? Did MP3s kill CDs? Perhaps, but the fact remains. We use internet, we read ebooks and we listen to MP3s. With new technologies comes change. And changes come with shifts in industries.

Artists have always inspired themselves from one another, even in Antiquity. This is why there are movements such as cubism or impressionism. They didn't all suddenly come up with the same idea at the same time. They copied each other. But despite drawing inspiration from each other, they each had their own style in doing so.

Art is not just the ability to use a technique or to mimic a style, it is the ability to create something new. And in that regard, AI is nothing more than a tool. This is why even artists and their editors use it (see for instance, the recent scandals in video games and mangas). It will generate almost anything. But it will only make something great if YOU have the creativity and if you know how to get it to translate your vision.

You can see it as an art killer. I see it as an incredible booster for creativity. And I think that this is a tool that artists may come to use on the long term too.

[tender oak]

I don't feel accomplished or creative when a generator matches the vision I have though.
I may feel mildly amused, but that's about it. Even if I had to fine-tune a prompt or whatever, and sift through 100 "meh" creations to find something to my liking it still isn't the same as putting in the time/passion into doing it by hand.

Sometimes it can be useful, like snagging a reference image/pose to draw on my own.
And stuff like story helpers can come in handy with things like pacing, proper word usage, and suggestions towards what could happen next.

Other times, it can feel like a creativity vampire - an all-too-convenient shortcut.

[agile pier]

I don't think one's sense of accomplishment matters in this instance. AI, as a tool, can do many things. How someone uses it can have different emotional outcomes, and even that differs from person to person. That doesn't make it bad, either. It's just not for you.

[indigo loom]

The fact that you do not put time and passion into what you do with AI doesn't mean it is the same for others. You are one individual, your own experience explains how you feel about AI. But that is your use, not that of others.

Personally, I use AI. And I spend a lot of time on it. I combine several AI-generated elements together. And I do creative work (in my own biased opinion). I don't just enter a prompt, generate it 100 times and call it a day. It does not hinder my creative process, it gives me the means to create things that I would not have the technical ability to achieve.

[tender oak]

Yeah, I just wanted to comment about my own personal experience I'm not saying my experience is universal.

[indigo loom]

Good. I'm glad we agree. Different people, different uses. 👍

[covert patio]

I manage a creative/content team for a global company. Among others, I have two full-time designers/artists on the team. They are exceptionally talented and I pay them well. While they could design everything from scratch, which they will do part of the time, we also use AI quite a bit too. AI helps us expedite the production of graphics, videos and other content against a highly demanding amount of project work. Does it lessen the quality? Well no more so than basic PhotoShop lessens the quality vs RL canvas. No matter the medium or method, ultimately we are still responsible for ensuring the end result we produce is always high quality and on brand. We take that very seriously. If AI didn't allow us to achieve that end, we would not use it. It's a tool, no more, no less. And it's a great tool in the hands of a person who really knows how to use it.

[covert patio]

I will just add, I'm a big believer in transparency and honesty. If a piece catches my eye at an art show, I'll often speak to the artist. I love to hear about their inspiration and how they created it. So hopefully they are being truthful. But if not, that's on them. That's their ethical failure. Not the fault of any particular oils or technology they may or may not have used.

[high scarab]

About Avatar Fx complaints... Avatar fx is not to blame for the slow mode, the slow mode is because people don't have C.ai+

[indigo loom]

On a completely different subject... Has anyone managed to make streams? I manage one a day, when I'm lucky. I'm really disappointed, it's a hilarious tool. They slay the roasting battles. :/

[high scarab]

Streams are something very beta... until a technician realized that no one did anything because it always failed and tried to fix it, it didn't work for any (and if it didn't work, it didn't consume resources).

Streams are fun... but they depend a lot on bots, I prefer to use Avatar Fx more... you have more control over everything (except the final result 😆 🤣 )

[indigo loom]

It did work the first day or so. It's after that it constantly failed.

I guess that to have a good content in the stream, you need to have good bots with definitions as the bot uses them as the basis of the discussion.

[high scarab]

Anyway, you know that even though the link is in the app, it can be used on a computer... it's a website.

[indigo loom]

It doesn't matter where you can use it if it keeps failing. XD

[high scarab]

I made this one with two characters that complemented each other (they have no definition, it was simply to see the effect in transitions)

[covert patio]

I'm going to give it a try. I happen to have two characters I created. One is an artist and the other a wizardess. Perhaps they debate on whether hand created works are better than spell conjured works. 😉

[indigo loom]

It's a great tool, but it doesn't work. Hopefully you still get one across.

My last one.

[high scarab]

In Avatar FX and the songs there is a problem, when vocals are lengthened and maintained (when singing) it is not detected as lip movement...

[covert patio]

I will give it try.... I just asked both characters if they would be up for a debate. The topic will be hand painted vs. spell conjured works and whether there's any artistic or ethical superiority of expression of one or the other. Seems like they are both up for the challenge.

[agile pier]

That was cute.

[indigo loom]

Hah. I never thought of asking the characters for their opinions.

[covert patio]

Success! Sort of... did have a couple of problems...

First, Streams substitutes a random male's voice for Misty. I tried a second attempt, after clearing my browser cache, but same result... a man's voice where Misty is supposed to be female.

Second, when I tried sharing the Stream here using Character.ai's offical share link from the video, the Discord channel moderated it. Hmmm, perhaps the admins here should fix that?

Anyway, it's interesting debate... but just look past that male voice on Misty 😂

[indigo loom]

The man's voice is hilarious and so unexpected. 😆

[high scarab]

It's funny when that happens.

[indigo loom]

Fresita sounds like a sports commentator.

[covert patio]

Good news and bad news.

So I solved that voice problem with Misty. I created my own voice for her and that seems to have solved the problem I had before.

But a warning here to other creators considering using Streams... after introducing my two characters, Misty and Lily... well now they're talking smack about me. 😡

[covert patio]

Struggling a little with some second thoughts on having introduced them. I mean it's great they're getting along and all, I guess.

[covert patio]

Tried out creating a Group Chat of the two of them too. But looks like that feature is not fully baked. No visible option to share or post it. Can only just create and engage in conversation on the fly, but not a permanent fixture.

[covert patio]

Any other experimental Character.ai tools out there? Besides AvatarFX, Streams, Group Chat?

[high scarab]

All I know is that all the animation experiments are aimed at achieving real-time conversations with animated characters in fluid conversation in the future... without delays.

But you can suggest things... they have new CEO and a channel for feature requests.
#1151576592280203314

Because a Tom the Cat with predefined movements and animations is not the same as any image you upload being animated in real time in a conversation, and also having animations that don't create an uncanny valley and show emotions.

[covert patio]

100% agree that Real Time AI-driven is much more interesting than pre-defined. Those who want pre-defined can watch YouTube. Not that there's anything wrong with YouTube, haha.

But the unpredictability of the AI makes it so much more interesting. We can still cultivate a particular happening, even if the exact circumstances might vary. I did that with one of my characters where a funny predicament always happens. Getting that to work in the Definition was fun to work out. But exactly how it transpires will be unique to each user and each conversation. 🙂

[lament sleet]

https://share.character.ai/Wv9R/ay9ioxra

[graceful coral]

Wrong channel. Share it on #1139065685935394876

[carmine widget]

This is absolute, no soul, boring thing ive seen all day

[tender oak]

👍 somehow can't do reacts to posts but 💯 🥱

a Window World commercial has more entertainment value...
(in the most boring tone possible) "at window world we do not only stand behind our windows stands on window we can stand on them. Can you stand on theirs? Maybe. But it'll cost you"

[covert patio]

Keep in mind the exact dialog is created by the AI. The creator really has no control over that... only the ability to generally specify the topic. In this case, probably something like "two characters are having a podcast together." It may not be Joe Rogan level debating here, but I don't think that was the goal.

**What the topic is about is actually missing the point anyway **because what matters is the transitions being demonstrated where there is a common background between two halves, so that it looks like the two characters are together in the same space. The current system doesn't inherently allow doing that.

If the devs could build that capability into the system to make it easier for any creator to do so without jumping thru hoops, then all kinds of possibilities open up for visual story telling. It could be two characters on an adventure together in a dungeon, or maybe they are going out on date with the two of them sitting in the movie theatre eating popcorn, it's just up to the imagination really.

[high scarab]

Anyway, I did it with an empty definition; I wasn't looking for anything... just exploring the changes in focus. Besides, I don't even know what they're saying because I don't speak English.

[covert patio]

Stumbled upon another experimental feature... in the Chat on mobile for any Character, there is a beta feature in the lower left corner. The Character Actions icon offers the ability to "Create chat story." It records a video as a narrated version of the selected chats - limited to 200 words max. No animation. Just words.

[scarlet summit]

I've been using AvatarFX since late beta with a single character and loving the ongoing enhancements. However, I would like to add some characters and spent a bunch of time creating 5 new ones on Midjourney. They're obviously AI-generated with the usual finger issues, etc but I can't get past the 'looks like a real person' security wall. As these characters are easily identified as AI generated, this is within the realm of 'intended use'. Over 48 hours without a reply from Support... Anyone else experiencing this?

[silent hull]

I need a responsible person who is willing to work

[trail thunder]

First time.seeing this general

[trail thunder]

How come nobody talks here?

[fading birch]

Dunno

[rapid pollen]

Because only a few seem to make avatar.fx

[strong thistle]

They're too busy developing the perfect avatarfx prompt

[covert patio]

Hmmm, maybe we should do something about that?

How I Created 3-Minute AvatarFX Video

Amazing what one can do despite the 20 second limitation on AvatarFX. Here's a video project I just posted earlier in the share channel for one of my characters (Violet) in c.ai. This is a preview version of what is actually a 3-minute video (link to full high-res version at the end) that combines 14 video segments. I produced the song on Suno using my own lyrics based on Violet's back story from her Definition. The forest scenes were created in RunwayML. I stitched it all together in Clipchamp. I do have Adobe Premiere but was curious and pleasantly surprised about Clipchamp. I don't know that I'll do a video like this for every character, but I did have a lot of fun doing it. Happy to answer any questions!

#🎬︱avatarfx-streams-share message

[young rain]

https://share.character.ai/9j2b/pvo4nu8d this smile…

[elder wren]

Ugh I can’t share a video on the app it says failed to post to Character.AI I don’t wanna lose it

[indigo loom]

It happens often, try to do it a few times. Sometimes it works on the second or third try. Not always.

You can always download it. And you can also share it by copying the link. 🙂

[elder wren]

Thanks for the advice

[indigo loom]

Of course, with pleasure, that's what the thread is here for 🙂

[elder wren]

Fortunately !

[jolly sage]

Is avatarfx down?

[covert patio]

Not that I've seen. But sometimes the platform can get busy.

[tardy oasis]

Tfw you realize the limit resets every 24 hours and not when midnight in CA passes.

[covert patio]

That's a good point! There is also a 10 credit limit per day!

[wary shoal]

I’m working on a project to build a real-time AI avatar chatbot that interacts via text and voice, with a human-like avatar speaking responses live. Do you have any ideas?

[tardy oasis]

You should apply to work for c.ai. I hear they're hiring.

[tardy oasis]

Stream generation fails due to moderation errors are the bane of my existence.

[strong thistle]

If it reset for everyone at the same time, the servers would get swamped, so it may be just as well that it doesn't 😁

[indigo loom]

Wb is right. It resets at 2 AM my time. Every time.

[strong thistle]

Hmm, I wonder if it resets for everyone at 2 AM your time. 🤔 I don't actually know what system CAI is using to schedule it. I just assumed from @tardy oasis's message that CAI must be calculating it based on when you last started making them, or something like that 😁

[high scarab]

They can tell what time zone your account is in; in fact, they know what state it's in (there have been feature restrictions by state). They just need to set a reset time when you access.

[high scarab]

I mean that the app or website can identify that information, even if it's locally, without having to go through the server.

[tardy oasis]

Yeah idk it's a mystery.

Also, can anyone make vlogs? Thought it wouldn't work for me due to banned words, but they work when I try to do regular video. It doesn't seem like I can do vlogs at all

[high scarab]

Sorry it took me a while to reply. I got banned for using the word "fistro" (even though you use it in the context of editing an image).
Opening labs on PC is as easy as opening the page the app redirects you to.
url: labs+character+ai
change + by .

[tardy oasis]

Oooh! Thanks! I hope you have fun with it!!!!

[obtuse cobalt]

e

[dim raptor]

does anyone actually use this feature

[strong thistle]

Yes

[high scarab]

Everyday, eveytime

[indigo loom]

Yes, A LOT

[glacial sedge]

Hii

What time is where everyone is ? ^.^

[indigo loom]

It's the same time for everyone: time to make videos!

[tardy oasis]

I'm reading the descriptions for the new UI, and I'm concerned they might be changing vlogs into POV. 😰 Vlogs still doesn't work for me, so I assume it's still on vacation.

[tardy oasis]

😂 I'm loling because others are using, what I assume to be video clip, saying voices have been added, but it's still video with the wrong style, and I'm over here trying it and my video clip outputs have changed to be POVs. 😂😂😂

[green herald]

Hihi

[tawny thicket]

Where is the comic strip stream? I wanna check that one out

[strong thistle]

Seems not to be available yet

[tawny thicket]

When is it gonna be available?

[strong thistle]

I do not know

[stark grail]

Yo guys is it true that on the 27th C.ai is reading all our chats?

Cus I saw it on my fyp earlier today.

[graceful coral]

Ah, no wonder they never fixed my 0 interaction issue, they’re too busy reading my chats

[pine oxide]

Yes, all cai devs will be too busy reading billions of chats instead of worrying about the bugs in the app and improving the app

Sure buddy

[lime dune]

Sobbing

I’m going to fucking cry if they do.

[strong thistle]

No. However, the devs have always had access if they needed it. That's how the system is able to remember your chats when you reload c.ai

[stark grail]

Ohh oki

[obtuse cobalt]

does anyone chat here anymore

[fallen field]

no

[brisk breach]

no

[tardy gale]

no

[keen token]

Way

[livid grove]

The

[vague peak]

Yesn't

[strong thistle]

I tried to make him dream that he was on a candy rollercoaster, so the AI thought, "Great, I'll make him climb it." It's like this the whole video. He never gets to ride it properly even once. Poor guy.

[strong thistle]

Is the Video Call option broken for anyone else?

I'm trying to make a chocoholic monk debate what's an appropriate level of sugar intake with an evil chocolate rabbit. It's very important!

[Update: It's fixed.]

[indigo loom]

Yes, I get it a lot.

I also have a lot of remixes which are "against the policies" but the previous versions were not :/

[strong thistle]

When I download one of my character's avatars from CAI, it is 400x400 resolution, even if I uploaded a larger image. I'm wondering: is there any reason to upload a larger image for the sake of video generation, or will the video be generated based on a downscaled 400x400 image anyway?

[strong thistle]

Getting these a lot more frequently

[strong thistle]

Now it produced a 10-second song 😅

[strong thistle]

Anyone else?

I got 4 of these in a row, so... at least it's reproducible.

[high scarab]

Ngl think it's just you

[strong thistle]

Hmm. It only seems to happen when I make Clips. Have you tried making Clips?

[high scarab]

Nein.

[past eagle]

I've been having issues for a couple of weeks now 😭 I've had a several of these too

[wintry olive]

Please add a FaceTime feature

[strong thistle]

I think that's their goal. Google "Character.AI’s Real-Time Video Breakthrough"

[past eagle]

I guess this is the sign I need to touch grass 😭😭😭 POV streams are rude to me

[strong thistle]

@formal rose Hi! I have a question, if you don't mind. When I download one of my character's avatars from CAI, it's 400x400 resolution, even if I uploaded a larger image. I'm wondering: is there any reason to upload a larger image for the sake of video generation, or will the video be generated based on a downscaled 400x400 image anyway? 🤔

[dense sinew]

ded chat

[charred adder]

[maiden saddle]

I don't understand why it keeps giving me Fallback Mode in Character.ai Streams... -_-

[strong thistle]

Sometimes it fails to render—probably because the server is overloaded. As for why it shows that message, there's probably some part of the service that reasons that if it hasn't heard back from another part within a certain amount of time, there must have been an error, so it should give up.

[floral parcel]

I keep getting the fallback Mode in every video myself. Even at a time where there barely should be any overloading in a server. It wasn’t like this several days ago.

[elder wren]

I put an audio just to try it

[karmic vale]

@floral parcel try now.
I fixed a few issues yesterday

[karmic vale]

@maiden saddle

[floral parcel]

Not working still, I’m getting now errors with “no video content available.”

Text chats work, but the voices are also unusually robotic lately.

[karmic vale]

Are you cai+ ?

Also...
Define lately.
Last couple of days? Couple of weeks?

[karmic vale]

@floral parcel ^

[floral parcel]

I am cai+ and I would say over the past week.

[karmic vale]

thank you

[karmic vale]

what's your cai username?

[floral parcel]

Nessie (also, the last time I had a good stream with much more natural voicing was on the 21st of September).

[karmic vale]

[floral parcel]

Went back on my logs to check.

[karmic vale]

@floral parcel I've excluded you from the new model for now (cai+ only model) can you try and see if new streams are better audio quality ?
also - if you can specify characters you're using... maybe it's something specific to the these voices,
as in general the newer model is A LOT better than the old one

[rain meteor]

I need help

[karmic vale]

Talk to me @rain meteor .
How can we help ?

[floral parcel]

I’ll do so after work.

[karmic vale]

thank you

[floral parcel]

I’ve just tried it with about three different streams and I don’t see any difference than when it began to sound robotic again. About a week ago it was great, then it just tanked recently.

[karmic vale]

Can you share the characters ?

[floral parcel]

Sure. Can I DM them?

[floral parcel]

If that’s okay.

[karmic vale]

@subtle hinge

[karmic vale]

Nothing personal.
Just asking Marie or a mod to take the info on a DM 😀

[floral parcel]

Huh. My old streams that used to work now immediately go into fallback mode, weird

[rain meteor]

I can't open chat history of one character i

[karmic vale]

@subtle hinge can we help @rain meteor

[rain meteor]

Help

[graceful coral]

The app lags when there are too many chats in history. Use the website.

[rain meteor]

I can't do anything 😔 else

[graceful coral]

[narrow salmon]

Is anyone else having issues with stream sounds? The only sound that truly works, that I have used so far, is the music and selfie one. However, texts chat and roast, video roasts and call, are not working for me on sounds. What is the meaning of this?

[narrow salmon]

This makes no sense either. I have made multiples of videos and it keeps saying an error on the voice when it is on the character. Though, minutes later coming back, it makes the video anyways and it cannot be played with either the sound or the video itself.
The prompt is there, I just refuse to share that since it is a bunch of nonesense.

[indigo loom]

I've been having the issue of roasts being really short.

They each speak once. And that's it. There is no back and forth. I don't know if it's intended and if the

Examples: https://labs.character.ai/public/gc-vid/zlXkRZnC
https://labs.character.ai/public/gc-vid/NiR0Xzjb

[grave thorn]

­

[karmic vale]

I deployed a fix for the missing voice IDs.

Let me know here if it's better

[round solstice]

Hi, just following up on this. Are you on the app or using the website?

[strong thistle]

@karmic vale When I have made multiple videos with the same title, one right after another, it would be very helpful to see a full timestamp here. That way, when I am reviewing the videos I have made, I can more easily recognize which one I was just looking at when I return to the list of all my streams.

P.S. I can't post images in #📝︱avatarfx-feedback.

[strong thistle]

Alternatively, if I could navigate to the previous/next video in my list, that would be a fine substitute—perhaps even better!

[rain meteor]

I am using app

[narrow salmon]

Do mods in this server even handle these issues? Or do I need to reach out to the company of the app instead?

[graceful coral]

Did you read this message? #💬︱avatarfx-general message They’re the one who maintains the Stream feature. Mods are here to maintain the server and bump your tickets, not to fix specific features.

[narrow salmon]

They were talking to me? I was unsure.

[graceful coral]

I understand the confusion since it wasn’t a ping, but yes, they did address your issue. So if you ever face issues with streams, just let them know here.

[narrow salmon]

It is still causing issues. Thank you for getting back at me though.

[karmic vale]

let me dig deeper - the system that provides the voice IDs is a bit annoying

[round solstice]

Have you tried seeing if you can access the history in the website?

[rain meteor]

I can access in website not in app and all characters history are working in app

[round solstice]

When you change your chat history for the website, does it also change it for the Character that is having issues on the app?

[rain meteor]

Yes it's does

[elder wren]

Where’s the button download when we want to download our streams ? Because I created videos :(

[strong thistle]

@karmic vale Whoops. The text field asks for a string up to size 100, but in the database its size is limited to 50

[strong thistle]

I download them from the website by right-clicking on the video and choosing save. Does that solve your problem?

[elder wren]

Yes. It works on pc ! The problem is that I mostly use my phone for this

Maybe I should do on my pc instead..

[strong thistle]

Can you hold your thumb on the video to get a context menu to pop up? 🤔

[elder wren]

No

[strong thistle]

Try a different finger /j

[elder wren]

Lol 😂

But thanks for trying to help. I’m gonna use my pc for this now !

[karmic vale]

ack - will fix

[rapid pollen]

You can download them in your ‚streams accaount‘.
When you open streams, uppper right corner is your avatar image. ➡️ click ➡️ click the stream you want to download (it opens again with the download button)

[elder wren]

I’m on my phone and it shows this

[strong thistle]

@karmic vale Its ability to maintain continuity is impressive. How do I turn it off when I need to cut to a different scene?

[karmic vale]

[cut]

atleast that's what I'm trying to get it to consistenly do 🙂

[strong thistle]

@karmic vale we got two fives here in vP_0BLyE

refreshing doesn't fix it

I wonder what's in the db

[karmic vale]

I just did a big update
It may have messed with existing links

[karmic vale]

💡 Neon

[strong thistle]

I posted one to the feed. I hope that's okay. It was nigh on perfect, so I thought it would be

[karmic vale]

You're good.
Can I see ?

[strong thistle]

I don't know if it's what you intended it to be used for, but it worked 😄 #🎬︱avatarfx-streams-share message

[high scarab]

What happened to all the Star wars characters Yoda Darth Vader

[strong thistle]

The Empire got them. The Disney Empire

[high scarab]

What

[strong thistle]

google character ai disney

[twin matrix]

Ok really irrelevant question am I the only one facing slow responses

[strong thistle]

Generating a video always takes a while. If your question is not related to generating a video, I suggest you head to #💭︱cai-discussion to ask

[karmic vale]

NEON is coming to all

[indigo loom]

I know what feature to use if I ever need to promote a cheese sandwich shop 😂

[covert patio]

Do you accept uber eats? Got the app open.

Though the transit might be a really long one across the atlantic 😂

[indigo loom]

I was going to say, it makes me feel hungry

[strong thistle]

@karmic vale Just a heads up in case you are unaware. Previously, when I clicked here, it would display a list of characters. It doesn't do that anymore. Now when I click, nothing happens. I can type in a name and find a character that way, but only if it shows up in the search, because no list of recently used characters is included in the popup anymore.

[karmic vale]

yeah - a bug with recent chats we're working on..
I'm looking into it

[maiden saddle]

I can't download the video. It same happens

[elder wren]

Thank god someone has the same problem as me 🥲

[strong thistle]

Did something in song generation change recently? 🤔

[tired cliff]

is this working? seems to be very picky now

[karmic vale]

Nope

[covert patio]

Would c.ai be able to add a character voice creator/editor to Streams? Sure this already exists within the character settings, but lately that old system has been unreliable and error prone, especially when making character videos in Streams. So maybe the answer is to migrate and update the old system via a voice generator in Streams. Perhaps that's hard to do because of all the legacy character voices on the system. Still, I was just wondering about it.

[karmic vale]

Some of the reasons is that were using a different (and a lot better) model for Streams.
Still the same voice library, but different models will give different results.

When you pick a character on streams you see a little play button for an audio sample of that character with the streams voice model.

** Doesn't apply to A Star is Born yet **

[covert patio]

Where is that button exactly? I don't see the little play button. I tried Neon, Selfie and Video Call and I don't see any little play button. Someone else told me they do see it. But it's not showing up for me. I'm on Android.

[strong thistle]

Bug: The PFPs are not updated when I search for a character. As a result, if I want to render a video with a different PFP, I'll have to wait until... I don't know when 🤷‍♂️

[covert patio]

Button is missing for me. Someone else showed me the little orange button. I don't have that on mine.

[strong thistle]

We used to see a button to sample the voice, but we don't anymore. It's been missing for a while

[covert patio]

Some do still have it, apparently. As someone showed me. I'm like you though... I don't have it on my app.

[indigo loom]

I have the button on my app... but on the other hand, on the app, I can no longer download the videos.

In general, I think that the ability to download videos should depend on whether the person who makes the video owns the bot featured: you should have the ability to download the video if you are the creator of the character, and you should not be able to if you used someone else's character (to avoid copyright infringements).

[indigo loom]

On the app, with the new setting, I really like the fact that:

• I get notified when the stream is ready
• I get the whole list of recent characters I've chatted with. Because as a result, it makes finding the characters so much easier than just the last 5 chats.
• I love the fact we can change the title of the video, rather than being limited to the title we set before the video was generated (1. we can change the title if the video does match, 2. we can comment on the video when we post it, 3. that also solves the issue of videos being marked as unsafe because of the title, then one can just change it and repost)
• It's nice to have new styles such as the mystical one (which I've been trying with my spaghetti monster)

[strong thistle]

I can't generate videos of a new character until the search index updates

[indigo loom]

I have the same issue with the web version. I used to only use the web version. Now I use mostly the app (because it's easier and more intuitive), and I download the result on the web.

[strong thistle]

Would it help to reuse the character selector used for scenes? It's already operational. This one:

[elder wren]

Uh oh, I can’t download videos on my pc now

[strong thistle]

I know you guys at CAI are working to remove the Disney stuff. I hope you can remove it from the genie department. I'm tired of seeing this guy 😅

Why do the songs often start with someone who isn't my character anyway?

[strong thistle]

I tried making a Selfie of Bill. I don't know what happened, but he isn't himself.

[indigo loom]

Hah. Looks like an alien. It fits perfectly the "Bill".

[covert patio]

Perhaps Bill is trying to tell you something?

[strong thistle]

He was supposed to be telling me about hybrids 😅

[high scarab]

Guys, a doubt. A clip of mine was removed from the feed an hour after I posted it. How should I know if it was reported or a simply a bug if I didn't receive any notification?

[strong thistle]

Idk. 🤔 You could wait another hour and see if it comes back. If it doesn't, you could try reposting it and see if it disappears again. Do you think there may have been a reason to report it?

[high scarab]

Mmmh I fear there was certain element of the image that could have been misinterpreted. That's why I would like to know if I get some kind of warning.

Because if I make the mistake of reposting it, I could be banned. Don't wanna risk it.

[strong thistle]

Uh oh

@karmic vale

Seems to need a manual re-up

[karmic vale]

Ack

[karmic vale]

Fixed

[loud ice]

WHY DIDNT NOBODY TELL ME THERE WAS A LIMIT OF VOICES YOU COULD USE FOR STREAMS!?! 😭😭😭

[strong thistle]

What do you mean?

Are you referring to the screenshots I posted above?

[loud ice]

When I was trying to generate a stream using the streams? It said “Audio cannot be generated due to you using your limited voices.” And then it tells me something about upgrading like what?!? 😭

[strong thistle]

Can you post a screenshot of it?

[loud ice]

That’s what I get when trying to generate a stream!

[strong thistle]

I think this is a CAI issue, not a you issue

[loud ice]

I hope you’re right, because for some reason it does this with the text videos… 😔 but let me try if it does this using regular voice videos

Nope still says the same thing for face to face 🤨😒

[strong thistle]

Looks like another limit has been reached 🤔 @karmic vale

[karmic vale]

Well
That makes no sense

[strong thistle]

It didn't make sense to me either, but I'd hoped it would make sense to somebody 😄

[loud ice]

I finally found the problem, the two person videos both for text and video chat are currently erroring and needs to be fixed so you were right earlier, it’s a Character AI problem, because the single character video works… I just don’t know why the multiple characters aren’t working though. 🤨

[strong thistle]

Are you trying to use a voice that you haven't used before? If so, I wonder what would happen if you only used voices that you have used before

[loud ice]

Noo, I was trying to generate a stream, you know

[strong thistle]

I understand

The reason I asked is because it sounds like what is happening is that CAI is relying on ElevenLabs to produce audio for the streams, but every time they use it to synthesize a voice they haven't asked it to synthesize before, they need to create a new voice on there

That might not be the cause of the error though. It's hard to know what to make of it when the error says they're using 33371 out of 60 voices

[long blaze]

I think this is a CAI issue, not a you issue

[long blaze]

Ack

[strong thistle]

Should I change my prompt, or should I keep retrying? 🤔

[strong thistle]

It turns out that it was generating beads, albeit not ones I could use, but it wasn't navigating to the next page in the workflow. After I trimmed the input to only the first sentence, I finally managed to reach the next screen, where I saw this:

[karmic vale]

now you're just teasing people with a feature they may not have 🙂

[strong thistle]

Should I report the bugs elsewhere? 😅 I thought maybe this was the place to do it since you announced the feature here: #🎬︱avatarfx-streams-share message

[karmic vale]

You're good

[covert patio]

Just ran into an interesting error on Sizzle. Passing along!

[strong thistle]

For anyone pining for these new features: it still needs some work. Right now, it feels frustrating to use.
🤔
What if, instead of the Link workflow, things were broken down into more individual features?

• The ability to put multiple characters into a Neon
• The ability to generate (or regenerate) a reference image for each such character
• The ability to create a Neon that seamlessly continues from where an existing Neon ended. (The last frame of the existing Neon is the first frame of the new Neon.)
• The ability to concatenate two Neons (not other people's, but your own)

If there were also the ability to trim off some of the start and/or end of a video (since some actions don't need a full 10 seconds), then people could make all kinds of things!

[karmic vale]

FWIW
I did Neon AFTER Link - from my learnings on making Link - If that makes sense.
so - some styles are experimental and would never graduate to full ones, they are a tool for me to learn how to do better styles

does that makes sense. ?

[strong thistle]

Ah, yes, it does!

[strong thistle]

Can anyone help? I really want to set up the bot so that if someone wants to make a video of it, he can write the video prompt using the same terminology that he would use when talking to the bot. I've made quite a number of changes to the bot already to try to make it absolutely clear what is meant. Roar and PipSqueak understand, but it hasn't helped with video generation. Is there any hope of achieving this goal? 🥺

[tardy oasis]

Have you tried defining catfolk in the greeting and/or description? I noticed in an old vlog video that started with a blonde character as the pfp, the vlog would transform the hair color to match the description of pink hair, which was mentioned in the greeting and description. I have no idea if it was in the long description as it's not my bot.

Or perhaps you could use the term nekomimi in the RP instead, as that might be easier for the AI to pick up?

[indigo loom]

My cat and dog emperors tend to turn into cat/dog midway in the videos. And yet, it's very clear in the definition of the bot what they are (the bot doesn't confuse them with regular cats and dogs either when I talk to it).

[karmic vale]

Short answer.
A solid Shmaybe but most likely not.

Long answer will wait till I get my coffee

Long answer (as I take a sip)
Most video models (from all vendors) are trained to know what are certain things
What is a house
What is a man
What is a woman
What is a unicorn
(what is a Big Foot vs A Bigfoot 🙂 - sorry @strong thistle )
What is a Cat
What is a Dog
etc...

so when you write Cat Emperor - it in most cases will be very literal - A Cat (which it knows) and dress him up like an Emperor (which it knows)
when you do I2V (Image To Video - like most of our styles) it will look at the image and in these cases will categorize what it sees as a man (not a cat man / dog man)

More than that
Video models tend to have a much smaller context window (usually 2K) so very very long descriptions are usually lost on them.
and would most likely produce adverse results as you confuse the model .

Hope that clarifies things a bit.
Sorry for getting a bit technical, but you guys know more about prompting than most people. 🙂

[strong thistle]

Bug?: I wanted her to appear with a Halloween setup in a Selfie, but she insisted on babbling about her outfit and the props at the start of the video. I couldn't figure out how to get her to stop doing that. So, I had an idea: make the video once with a description of the setup, then remix it without the description, instead reusing the previously generated reference image. However, then a new problem appears: each time you remix, the reference image gets a little blurrier. Here are the last two as an example:

[covert patio]

Maybe just listen to her. Take a real interest in her concerns. But mainly just listen.

Oh, wait... I might have misunderstand the nature of your question. 😉

[round solstice]

May I know for what reason?

go ahead

[strong thistle]

Can't load my streams

[strong thistle]

I gave it a try. I think it has helped, even it it hasn't been 100% effective. Thanks for the suggestion!

[indigo loom]

I have the same :/

[strong thistle]

If you google big internet outage—that might be why

[indigo loom]

Ah yes, thanks 🙂

[tardy oasis]

You're welcome! Glad it helped a little.

[proud iron]

Hey there

[strong thistle]

This fellow is spamming greetings where they don't belong (here is fine but not in #📝︱avatarfx-feedback), sending random friend requests, and asking people for their number @simple cosmos

[simple cosmos]

poof

[tiny shell]

Hallo i need some help with the images being just black and the video's not loading with this error - ExoPlaybackException: ERROR_CODE_IO_NETWORK_CONNECTION_FAILED
Error Code: 22001

Im using the mobile app

[strong thistle]

@karmic vale New error message

[karmic vale]

Well... They changed it this morning apperantly and forgot to tell me.

Let me check

[tiny shell]

Re-downloading didn't work

[tender oak]

figured out the voice mismatch.
I assigned voices in the section where you can type to the character, but not in the actual "edit character" menu.
It was using a "suggested" voice.

[tender oak]

has anyone had luck telling the magical stream maker what NOT to do?

[karmic vale]

We pay that guy by the hour...
Just so you know.

[strong thistle]

Sometimes, in some ways.

[tender oak]

haha

[covert patio]

I have had success with negative prompts on Streams, including recently in some Neon videos. But I'm not sure it works on all forms of Streams, and potentially including it could in fact trigger the very effect you are trying to avoid. Still, I'd say just try it and see what happens! If it doesn't work, then just try translating what you don't want to happen into what you do want to happen... so a positive prompt that inherently negates the unwanted issue.

[tender oak]

"oh and don't worry I'm not turning into a bunny or anything"

• the wallaby character says while already morphed into a bunny

[strong thistle]

You might try "(Make sure the wallaby remains a wallaby!)" or "The setting is the Australian outback" or something but idk if it will work

[tender oak]

hmmm worth a try. There's plenty of space to try to make the instructions as clear as possible, including stating the obvious like the character appearance (and begging for it to not go crazy)

[indigo loom]

I did manage to make videos with three of my characters by specifying what they look like. Even if as a result it keeps commenting on their appearance, as it is part of the prompt.

[urban tangle]

i was bored so i decided to go make a avatarfx and i put a ai photo as the avatar and it said "this looks like a real person" 🙏 ig the ai generator i use is too realistic lmao
well all of my ai photos r being rejected lol

[tender oak]

my animal characters are still constantly losing their pants even when describing their outfits and begging the script to stay consistent...my little squirrel guy is always doing what I call the "no pants dance" in musical reels.

a kind of amusing fix is that he wears pajamas just fine, so it has been "pajama night" at the club sometimes lol

[karmic vale]

ok - something you've all (well - some of you at least) is coming to Streams (only web for start) soon....
wait for it...

[strong thistle]

Oh, what could it be! Perhaps... reliable pants?!

[serene bane]

Hmm...always the person of mystery.

[strong thistle]

🤔 Perhaps a fix for the character selector? That would be fantastic

[covert patio]

Perhaps it is a recently requested new feature 😂 🧚‍♀️

[karmic vale]

I voted for pants BTW... .
But had to disagree and commit for something else

[serene bane]

I ain't no party like a no pant party

[tender oak]

can we post gifs related to avatarfx/streams somewhere?

[covert patio]

Streams POV and Selfie videos seem to be having issues. Same error coming up again and again.

[tender oak]

ruh-roh suddenly it's saying all the labs stuff is for premium users 😅

[crystal cedar]

That's the same error I've been getting 😭

[tender oak]

oh false alarm, that was odd

[indigo loom]

This thread here relates to AvatarFX and Streams. Not to the chat with characters. You may want to post this in one of the other general chats instead.

[tender oak]

dammit the animal characters keep getting human lips with lipstick on them uuuugh I think I'm just done tinkering with this

[strong thistle]

If you think that's bad, trying making a video with centaurs

[tender oak]

I'm scared lol

[covert patio]

I know you're a very savvy prompter, so if anyone can figure it out, you can! Not sure if you've tried, but could reference "Greek mythology centaur," "Fantasy art," or "Dungeons and Dragons." The AI is likely trained on mythological/fantasy creatures and those references might help with context.

[karmic vale]

@strong thistle see my Bigfoot 🦶 reference.

@covert patio is spot on.

[stone gate]

@covert patio check ur dm's

[tender oak]

what are the lyrics?
the lyrics: no lyrics!

[strong thistle]

Worked like a charm. Take a look /j

[covert patio]

It's tough challenge to do what you're trying to do. 🙂

[strong thistle]

Another problem is the AI doesn't seem to understand the subtext that the user has basically carved out a job for himself as a professional troll by showing up with such an unsuitable ride. I'm afraid to tinker with the bot's settings though 😅

[covert patio]

On the positive side, you did get a unicorn. 🦄
(Just saying)

[strong thistle]

@ripe blade What was happening previously was your input topic would be elaborated on by an AI, which would then pass it on to the video generator. That is still happening, but now we get to see the elaborated version. [At least, that is my interpretation.] This should make it easier for us to tweak our videos to perfection—hopefully. Granted, it would be nice if we could also see our original input.

[ripe blade]

No, but I prefer to leave it blank so the app can give me a surprise! It shouldn't mess up with my topic like that! I create about over 50 snippets daily to practice listening so the topic should be consistent. I want the app to give back the yesterday's streams! No ruining my mood when I choose baby and toddler characters
Also, the app will never stop adjusting the topic once I hit remix repeatedly

[strong thistle]

For this bot I would like to explain to the video generator that the two figures in the Character's avatar are the same person—in one pane she is merely in disguise—so that hopefully the generator won't try to turn her into two separate characters. How should I refer to the avatar? Avatar? Input image? Reference image? 🤔

[tardy oasis]

I usually just say picture, ref pic, image, pfp. Not sure how it'll work for you in your case though.

Actually, if you don't mind, would you (and anyone who has done similar things) mind sharing more about your experiences with multiple characters in a stream? I was considering making a new character that had multiple characters in the definition and in the pfp, and I was wondering how that's been working out for people.

[indigo loom]

Badly. I have characters that refer other characters and the stream ALWAYS takes the secondary character as the main character. I have to specify who my character is, his gender, etc.

[tardy oasis]

Okay, thanks! That's super helpful to know. I'll make sure to specify the best I can.

[karmic vale]

would writing a guide on best practices help ?

[indigo loom]

That's a great idea

[ripe blade]

The Streams is good in overall, but some characters don't behave like it did before the automatic prompt change that happened yesterday. Especially baby and toddler characters: I've seen a decline in their quality of speech if I don't specify in the prompt. In the past, the kid characters still speak properly when I change tone from serious to other 4 remaining one. Now if I don't write in the prompt, they can only make sounds but not words or some single words, muted words in parentheses

Children AIs make me happier because they seem smarter.
Also, today they use more puns than sarcasm in funny streams too

And I must admit I don't like the AI change my prompt so I can reuse it when remix

[tardy oasis]

Thanks! I don't have the edit button or the modified button yet, and when I remix streams, the prompt is always the same still. I agree with you - I don't like the idea of the AI changing my prompt when I use remix. I only use the remix button to see what my old prompt is and manually change the prompt if I didn't like what the video did. Having the prompt change for me when hitting the remix button would be undesirable for my purposes. I also liked the remix button because I could see what other people did as directions. It's helped me figure out how to better my prompts or bots.

[tender oak]

Yeah I also like the prompt staying the same. I try to keep it simple, keeping in mind the limitations of the magical video maker. Meanwhile, the prompt changes to a huge paragraph which it will never successfully generate. It's like deliberately sabotaging itself.

[tardy oasis]

Ahhh I just got the feeling that perhaps the longer description actually might be useful for some of my purposes because then I might know how detailed I have to get and new insights on what to change. 😅 idk. Will have to check it out when it comes to me.

Does the AI actually go by this longer remixed prompt for the original stream? Or is it just for remixed vids?

[ripe blade]

As soon as I made the first stream with my short prompt, my original prompt isn't there anymore

[tardy oasis]

Ohhhh okay. I can see how that could be troublesome, especially since it's changing the intent and outcomes of your videos.

[ripe blade]

I usually leave the topic blank to open streams "blind boxes" which is exciting. I don't want the prompt to remind me of anthropomorphic animals and objects. I already knew that

[karmic vale]

the voice issues should be better now

[tender oak]

how come you cant right click and save the clips anymore?

oh, nevermind ya just click edit first

[crystal cedar]

Did you ever find a fix for this?

[covert patio]

No, but you might try again. A bunch of updates rolled out yesterday. 🙂

[crystal cedar]

Ahh nope, I'm still getting the error 😭

[karmic vale]

@tulip arch you can ask here - or we can ask @kind heron to connect us on a private channel

[tulip arch]

This is fine! What part of the character definition does the Streams feature pull from? For the most I've seen it pulls off the Description of the bot and not the full/long definition.

The image generation appears to include more info (because it has to take the context of the chat in consideration) and that seems to create more consistent results

I'm trying to find a way to get more consistent results without reinventing the wheel with every prompt

OR

Inadvertently conflicting with prompt information it's already pulled.

[karmic vale]

fair enough -
it depends on the Style and the phase - but most of them pull the full definition.
So for example, all the battles / conversation - take a lot more from definition.
BTW - I'm working on a prompt guide for Streams.
When to use each Style, and how to optimize for a better output.

for the images we generate as first frames on Streams - we ALWAYS take the character avatar image as baseline and expand on to it.

[tulip arch]

Thanks for getting back to me! I'll try messing with a bit more and see if I can get the things in the long definition to actually pop up. Do I have to summon up those aspects by name?

[karmic vale]

Depends on the characters
for example, for IP characters
You can call them by name, the model knows who Mario is or Batman etc.
for OC ...
Man with cape and cowel would probably work better than random-hero-4

[karmic vale]

To be more clear - difficulties come in as there are many models involved and each has their own world knowledge.
so try to be descriptive of traits more than names.

[tulip arch]

I'm... Not really seeing where Streams would pull on the long definition on reference to the characters. I can see aspects in the videos that pull on the description (short definition), for example almost every video I've genned of Joan even on random include a dragon. The only place that I mention a dragon is the description. Likewise with the cafe environment itself. It's very well described in the greeting and long definition. Yet when I mention "in the cafe" I get different versions of a kitchen. I don't mind doing the extra work to establish setting, I just need to know what's already there.

For my Walter character, he has a split personality that is defined in the long definition, but not the description.

[karmic vale]

So - I use Description and Definition.
not greeting.

it's not OR
it's Description + Definition

[tulip arch]

No I don't expect the AI to read everything. I just need to know what it does read so I can build.

[karmic vale]

I'm telling you - it reads both 🙂

[tulip arch]

Because I'm not seeing the long (3200) definition details popping up.

Even when I call them out directly which would usually trigger something

[karmic vale]

can you share a character, an example prompt and what you would expect?
you can DM them / ask Mod to send them over if you're not comfortable sharing them here

[tulip arch]

https://character.ai/chat/IDfYklP6l1VzF4tqS3jge23BlUpxuHtSCTT--AA0wo8

This is Walter. I use him for testing because he has a few surprises in his definition that aren't in the description.

One of those is Mr Keys. If he goes jeckle and hyde without prompting, the AI is reading the definition. If he stays as Walter, then I know it's not, or there is a misunderstanding in there.

I get that Streams isn't going to be perfect. I can see that it is reading some of the definition/description and that is manifesting in the video.

I don't mind putting the work in, I'm just trying to figure out where it stops and where I need to start.

The simple prompt I used for him was "Walter turns into Mr Keys" that's it. I think I picked serious or grimdark. What came out... Was not Mr Keys.

And if there is a more effective way to pull that info out, or a technique or tab or something, I'm more than willing to learn

[tardy oasis]

Whoops. I was making a bunch of streams with one character, but changing out the pfp before a previous stream finished rendering. Is that what this error is about?

[elder wren]

What is fallback mode ?

[strong thistle]

It means that something went wrong or isn't ready yet, so you're being shown something else in the meantime, even though what you're being shown isn't 100% of what you should be shown

[elder wren]

Oh well something went definitely wrong I can’t play it 🥲

[strong thistle]

I always notice it when I make a Selfie, but if you wait a few minutes and refresh it, then it will probably be finished [at least if you're making a Selfie]

[elder wren]

Ok then I’ll wait ! Thank you

[tardy oasis]

Been trying a new angle where I give a small prompt instead of a long, drawn out one in Neon. Writing, "Movie trailer", and then write in some minimal details is actually kind of fun for original characters.

[crystal cedar]

I feel like a boomer for asking but how do you embed the stream videos here?

[indigo loom]

You download it and post it 😛

[wintry olive]

Maybe you should add a FaceTime feature

[hard musk]

[serene bane]

Anyone else getting this?

[indigo loom]

Yes, a lot. More than half of them. Well... that was when I could still generate videos. But now, it just doesn't let me generate them (which makes me wonder if there is maybe a limited number of credits per user again)

[serene bane]

I've made two...

the others didn't come out

[indigo loom]

Poor Shartplate must have a lot on his plate at the moment. 😂

(I don't know why, it seems that today, I'm in pun mode)

But in any case, yes, we all have that. All you can do is try, and pray.

[serene bane]

These are the days of our lives...developer edition

[coarse epoch]

I keep getting error status code 403/404, even when I check streams that were successfully made in the past

[coarse epoch]

For me, I can still generate text videos.
Idk about you

[karmic vale]

This is super helpful.
I know where to look now

[indigo loom]

Sizzle doesn't work either

[loud ice]

Oh my god, I thought I was the only one! Sizzle doesn’t work for me either! That is so strange because it was working two days ago and now it doesn’t want to work anymore! 😭

[lucid epoch]

[strong thistle]

I think this happened because I wrote "No voiceover."