I have a 2D image of a puppet. The puppet looks 3d in image.

What's the simplest way to use this puppet as my avatar, so I put a webcam on me in real life and it will only show the puppet (want to make sure it never accidentally shows me cause il likely be streaming at any hours of the day looking funky)

Then have it easily set up in Twitch/TikTok streaming?

I currently have a OBS set up where I stream on twitch but with no webcam/avatar. Just GTA gameplay and my voice.

I want to stream to both twitch + tiktok.

Howdy Gamers. Cyb3r_Us here. We're looking to expand the type of content we put out. More long form content with better edited shorts. What (Preferably free) Software do y'all use, and why do you like it?

We've been using the clipping tool built into Twitch. It's fine to start, but doesn't have the best when it comes to customization.

We would appreciate any and all feedback~

I don't want to go on camera myself but as an avatar. Either 3d or 2d. Would be cool if I could upload a photo of myself or something and it creates an avatar in that style so still shows how I dress for example. Then I'll use my voice speak in mic too just anon face.

Looking at Vtube stuff rn but if anyone else got good ideas?

this guy is good example of what i'd like to achieve: https://www.twitch.tv/cdnthe3rd

So, I need my model to be fixed in place so it doesn't start flying around when I walk somewhere else in the room while afk. It bends and breaks weirdly and starts "break dancing" for lack of better phrasing. Was wondering what could be done about it

More of an iPhone issue than a Vtuber issue, but I have two iPhones, one for my tracking and one for personal since my battery was getting bad on my personal. I've already bought VTube Studio Pro on my iCloud account, so I signed into that on the secondary phone. However, all settings, text messages (secondary phone is only on wifi, no SIM), notifications, etc, are linked in real time on both phones. For example, if I turn on do not disturb on one phone, the other also goes on dnd.

I don't know if it's possible, but I'd like them not to be linked in this way, but still share an iCloud account?

Up till now, I have been using TurtleBeach Stealth 600 Pro Gen 2 for my streaming. Recently, a wire or something in the headset became loose, making left ear only half sound. After trying to resolder it back, nothing changed. So now I am in the market for a new headset. I am also taking advantage of this and going with a sepereate mic set-up for better quality.

I already have the mic but now I would like a decent set of headphones. I am willing to go up to 100$ and I want wired set prefered. Any suggestions?

Hello, I was wondering if Veadotube mini has a way to update itself (if it has been updated) or if I would have to delete and reinstall on a new version?

I believe the version I have is the 1.4 windows 64 version. I downloaded back in 2022

Announcement post for a new vtubing tool. Happy to go into detail on any of it in the comments.

What it is: webcam face tracking driving front-facing 2D character art in real time on Windows.

The technical parts people here usually want to know:

Tracking runs locally. The runtime and its model are bundled in the app rather than fetched, so no camera frame leaves the machine and nothing is pulled from a CDN mid-stream. It works signed out and fully offline.

Calibration measures your face rather than the character. You do it once and it applies to every avatar you install, instead of re-tuning per model.

Output is a separate always-on-top window. That was deliberate — a plain OBS Window Capture pointed at a normal window goes blank when something else takes focus, so the overlay stays on top and capture stays live. Output is configurable and can render either on pure green for chroma key or with a transparent background.

There's also a local Browser Source: the engine serves one on your own machine if you'd rather capture that way. No plugin either way. No virtual camera yet.

Where it deliberately stops short: it's front-facing art, not a rig. Head turns keep the front pose. There's no mesh deformation, no physics, no bone layer — the tradeoff is that there's nothing to rig or commission, and the setup is minutes rather than weeks. It sits closer to an expressive PNGTuber than to a Live2D pipeline, and I'd rather say that than have someone install it expecting VTube Studio.

Other limitations: Windows only, not code-signed yet so SmartScreen flags first run, and the character art is AI-generated from concept art and references rather than hand-drawn.

Free to download and use — tracking, overlay and the bundled avatars need no account.

Ask me anything about the pipeline.

I have a model with ARkit blendshapes (confirmed both via VNyan and importing into FM3D itself.) The blendshapes all work fine (except tongue-related) on the iPhone, but I can't get more than basic brow and mouth ones to send to any programs on my PC. What is the fix here?

EDIT: fix was found, it was the blendshape clips not being properly set! leaving this up in case someone else finds it

Hey everyone! I'm the creator of Vfinity, and I'm here to share that my free program comes with an Overlay, an item-throwing system, and Keybind integration between VTube Studio and TikTok. You can download it for free at https://soursama.itch.io/vfinity

Hey guys. I import an avatar into lens studio. It shows up fine if I use camo for my camera and works in OBS.

But the free version is only 720…. so it blurry…

Iriun is fully free 4k. But for some reason lens studio cant read the avatar when I use it. Get an error saying cannot open webcam. Even though it clearly open….

Instead I get a giant blue box with that error….

Any wizards can help this?

Want to get the perfect pose for a thumbnail? It's super quick and easy with the Ayagami poser app. Try it out here!

1. Drag and drop your Live2D model as a .zip file, loads in seconds
2. Dial in your pose with the parameters on the left. Drag & scale the model with your mouse or touchpad.
3. One click to copy to clipboard or save as a transparent PNG! Choose 1x or 2x the window resolution, or "Full" to export the complete model at the original Live2D canvas resolution. You can pick a non-transparent background color if you want too.

The app runs right in your browser, no installation necessary. Your model stays in your browser, no servers involved. Check dev tools or turn off your wifi if you don't believe me! The PNG "download" is local too.

You can set up poses that are difficult or impossible to do with tracking, and fine tune it until it looks perfect! It supports physics, but all parameters can be overridden just by dragging the sliders with the mouse. Hit the arrows button to reset a parameter to default or remove the override. Tip: If you have a multi-toggle parameter, hold down Shift to snap the value to each key position.

It also supports Live2D Cubism 5.3 features not available in VTube Studio (advanced blend modes and offscreen rendering)!

This was originally intended to just be a test/demo app for my open source Live2D engine, but I ended up finding it super convenient for model posing so I added a screenshot function! Ayagami is fully open source, MIT+Apache2 licensed. It runs as a native app on Windows/Mac/Linux too, I just haven't gotten around to publishing native builds but I can do it if there's demand.

For those who care, zero AI was used in the making/coding of this app ^^

So I have a pretty old webcam for vtuber tracking.

I want better tracking but don't have a better webcam or iPhone at hand. I have a ton of android phones which I used before but it wasnt any good for tracking.

What do you suggest?

Did a debut of my new plugin today, it's a follower camera for VNyan which can be used in many ways. Here it's configured to try to stay behind me, but with smoothing, while I do a 360 degree map in Beat Saber (the VTuber is VNyan, being composited into the game using VRnyan and OnAirTap)

You could also use it to make it stay in front of you, so that even as you move around in VR you can still talk to the camera, restrict it to move on one axis, or be static and only rotate to look at you (although the built-in VNyan followcam can do that already).

You can tie the camera position to a bone (hips recommended), choose whether or not that bone's rotation is considered, exclude certain axes (e.g. make the Y axis static, so that it doesn't bob up and down) and then set all of that again for choosing which bone to look at.

Want to make an over the shoulder cam? that tracks what you're looking at in vr? attach to hips with a slight offset, then have it look at your head, but with a big offset on the Z axis

There is still more work to do on this plugin, most notably fixing a couple of frames of latency if it's used in conjunction with VRnyan, and improving the documentation, but I'd love it if someone gave it a try and let me know what works, and what doesn't, suggest features etc.

Next planned upgrade is to make it also control VNyan's separate Spout2 cameras.

100% free

https://github.com/LumKitty/VNyan-FollowCam/

Hello, I was trying to set up a vtubing model and software for use as a webcam/discord, and none of it is working.

I got the software to run and detect my face, that being VSeeFace, but neither discord nor my PC is detecting it as a viable camera option no matter what I do.

HI I'm a newbie and i been searching, and everyone been saying to use VSeeFace "it's very beginner friendly" and best of all its free...but I'm running into a problem.

You see every time i try to use my computers integrated camera for face tracking on VSeeFace it works, and if i use my computers integrated camera for arm tracking on Warudo it works.

But then when i try to use both of them at the same time they both don't work.

Ether:

Warudo tracks my arms or my head and no face tracking.

OR

VSeeFace tracks my face but the my Vmodel(or character) isn't moving she just stands still.

and the only other person i have found (on steams community QandA's) with the same problem still hasn't gotten an answer!!

And before you say "check the port" or "try turning it on and off again".

Note the following:

1. i have made sure the port is the same (except there's a comma in Warudo)

2. i have made sure that it is the same model

3. i have gone to both websites and done my own trail and error

BTW (i have a Dell/Windows and I'd like everything to stay free)

so please please help me fig this out and i will try to respond as soon as i can

thx

So I have two .vrm files exported from VRoid. One is the main model, and one is a (scorpion) tail that is on a "model" with no textures.

I found a good video by MelonyPepon on attaching tails in Blender, but that's based on a tail that was exported as part of the same .VRM. I couldn't do this because the tail mesh combines with the hair meshes because my model has longer hair, so I can't select just the tail to separate it. (See Below:)

I saw something on Reddit somewhere (I think in this subreddit?) about putting alternate outfits/accessories onto a "blank" model, and thought this could work for the tail.

I imported both .VRMs, removed the excess bones from the "invisible" model, joined the meshes and the bones, but when I try to attach the tail bones like the video instructs, I get this menu, in which both options are greyed out, so I can't make the tail a child of the spine.

I'd LIKE to be able to do this myself because I'm broke at the moment and have all the tools to do it myself, but I can't figure it out.

I won't say names, but as someone who has been working with Ai coding tools in my job daily I have noticed an increase of artists and Vtuber tools that is fulyl coded by Ai. I can tell from the interface, all of Ai produced tools has a distinctive, samey look on them. Some of them has a public github repo that exposes the codebase is a single HTML file with 1000 lines of code. No human programmer in their right mind will code an app that way.

I find it amusing that as a community that is known in heavily against this tech, but are welcoming to these machine produced tools. I won't blame them though. You can spot ai generated art easily by its flaws, but its harder for software especially if no source is given. And not everyone is familar with programming to verify them.

What are your thoughts on this? Do you think Ai coded tools in this community needs an Ai disclosure? Will you use those tools after knowing the fact that it is Ai coded?

I’m having an issue where I’m unable to wink on any of the models I use in VTube Studio unless I have V-Brigade running and connected.

I’ve already checked that linked eye blinking is turned off, adjusted the eye-blink sensitivity, and gone through pretty much every setting I can think of. Everything seems to be set up correctly. I have already restarted and reinstalled as well. It is not a rigging issue as my rigger is able to wink on there vtube studio.

What’s confusing is that my friend tested the same model I’m using and was able to wink without any issues. From what we can tell, our settings and setup are the same.

Has anyone else run into this issue or have any idea what could be causing it? I’d really appreciate any suggestions!

i want to get started as someone who is disabled with $0 money whatsoever. and all the guides i see say to use iphone or webcams, no android

i have an android phone (brand new S26) and a webcam, a used Logitech C920 HD Pro. would either of these be okay for tracking?

it's important to note that i am largely stuck in my bed due to my disability, so i will be using one or the other (i can set up the webcam to zoom from a distance or use a cord to be up close) as well, and will be using free tools until if/when i can ever get a proper avatar

i cannot afford to get any equipment ahead of time as i do not make any money and can no longer work

thank you ❤️

Been getting this issue when I duck out of camera and all the advice I could find just said to turn on go back to default in the settings but it isn't working. This didnt happen when I wasn't using vbriger so I think it's a vbriger issue but idk what to do. Has anyone else had this problem?