Live 3D VTubing uses a webcam to track your face and upper-body movement, then applies that movement to a 3D avatar. In Live3D, the relevant application is VTuber Maker: you choose an avatar, check its tracking, and send its output to streaming or video chat software.
The practical chain is webcam → tracking → avatar → output. This is different from VTube Studio, which its maintainer documents as a Live2D-focused application rather than a tool for tracking 3D VRoid or VRM models. Start with VTuber Maker for this workflow, and check its current menus and compatibility because product interfaces can change.
What Live 3D means for VTubing
“Live 3D” here refers to Live3D’s creator software and, specifically, its VTuber Maker application. It is not another name for Live2D Cubism. With a compatible avatar selected, VTuber Maker uses webcam input to drive movements such as face, head and upper-body motion. Its manual also describes optional voice-based lip-sync, which makes the character’s mouth move in response to audio; it does not mean the app transcribes or understands what you say.
A useful way to think about it is that the avatar is the visible performer, while VTuber Maker is the tracking and presentation step between you and the software that carries your stream or call. That receiving software does not need to be the same programme you use to track the avatar. The usual arrangement is to run VTuber Maker, enable its virtual camera output, and select that camera as the source in the streaming or chat application.
Keep the 2D and 3D setup instructions separate. VTube Studio’s maintainer describes the app as supporting Live2D Cubism models, not 3D VRoid or VRM models. Its calibration guidance may look relevant because both kinds of avatar respond to a person’s movement, but its model workflow is not a substitute for VTuber Maker’s. See the VTube Studio model requirements before following a guide that names it, and confirm that the guide matches your model format.
This distinction matters most when something does not appear in the software. A 3D model that cannot be imported into a 2D-oriented application is not necessarily broken; the application may simply be the wrong tool for that format. Likewise, a virtual camera appearing in your chat software does not prove tracking is working. Check the model and its movement inside VTuber Maker first, then check the output separately.
Prepare a webcam and 3D avatar
VTuber Maker’s manual describes a webcam as the tracking input. Use a camera that gives the software a clear view of your face, set it at a comfortable height, and arrange enough steady light for your facial movement to remain visible. You do not need to begin by buying a premium camera: the research for this guide does not establish a tested webcam recommendation or a minimum VTuber Maker resolution. A webcam for live streaming is one part of a setup, but choose it around your room, computer and other uses rather than an unverified performance claim.
Place the camera close to eye level and avoid strong light directly behind you. If the room is dim or your face is partly blocked, tracking may be less consistent. Keep your usual speaking distance and posture during setup; calibrating while leaning towards the camera and then sitting back can leave the avatar’s neutral position feeling wrong.
Choose a 3D avatar that VTuber Maker currently accepts. The research establishes webcam tracking in the manual, but does not verify a current list of supported model formats, import routes or avatar catalogue contents. Check Live3D’s current product documentation or in-app choices before spending time adapting a model. Do not assume that a model file intended for one application will automatically work in another.
Think about the avatar’s intended use before choosing its appearance. A waist-up character in a small corner of a study stream needs different framing from a full-body character in a chat scene. If you plan to use a keyed or plain-colour background, consider whether the avatar’s outline and colours will remain distinct against that scene. These are presentation decisions, not guarantees that a particular background-removal or compositing method is available in every current version.
Before opening VTuber Maker, close other applications that may be using the webcam. If the camera is already occupied, the tracking application may not receive its image. This is a practical check rather than a Live3D-specific compatibility claim; on a shared computer, video meeting software is a common reason a camera appears unavailable.
Set up webcam tracking in VTuber Maker
Launch VTuber Maker and locate its tracking controls. The manual describes enabling tracking, then using face, head and upper-body movement to animate the selected character. Menu labels may differ in the current release, so use the application’s current help or controls rather than relying on an old screenshot. Once tracking is active, sit in your normal position and make small, deliberate movements: look left and right, nod, and speak at your ordinary pace.
If the avatar does not sit naturally at rest or seems to follow an unintended position, use the calibration control. The manual identifies a Calibrate action for cases where the avatar does not match the user’s natural state. Calibration is most useful when you are centred in the frame, facing the camera and holding the posture you expect to keep while streaming. Avoid calibrating while turning, leaning or gesturing, since that gives the software an unhelpful reference position.
Voice-based lip-sync is separate from face and body tracking. If you want it, find the Audio Based Lipsync control described in the manual and enable it; otherwise, leave it off while diagnosing camera movement. Speaking a short sentence is enough to see whether mouth movement follows voice. If the mouth does not move, first check that the audio input reaches the application and that the relevant control is enabled. Do not mistake this feature for speech recognition: the documented behaviour is audio-driven mouth movement.
Select the avatar and scene that fit the intended presentation. Historical walkthroughs of VTuber Maker describe avatar selection, rooms or backgrounds, poses and quick actions, but those details may change. Treat the current interface as authoritative, and do not rely on old claims about which options are free, how many models are included or what a plan costs. If an action panel or shortcut is available, learn how it behaves before adding it to a live scene.
A dependable first pass is deliberately small: one avatar, one scene, tracking enabled, and no optional effects. Once that combination behaves predictably, add voice lip-sync or an action control and check it separately. Changing one thing at a time makes it easier to tell whether a problem comes from camera input, calibration, the avatar or a presentation setting.
Check movement and on-screen framing
Before routing anything to a live audience, look at the avatar in VTuber Maker and check the movement that matters to your format. Turn your head gently in both directions, look up and down, and speak. The avatar should respond in a way that is readable at the size you plan to show it. If the response seems exaggerated or delayed, avoid making several changes at once; return to a neutral posture and establish whether the webcam view itself is clear.
Check the frame around the avatar as well as the tracking. Leave enough space above the head and at the sides for natural movement. If the character will sit beside a chat window, logo or captions, confirm that the layout does not cover those elements. A layout that looks fine in a large preview can become cramped when reduced to a small overlay.
Background choices need a separate check. If you intend to use a scene inside VTuber Maker, verify that it is the scene you want before routing the output. If you intend to key out a background in another application, check the current output and receiving-app controls rather than assuming a transparent background will be sent automatically. The VTuber Maker manual describes a background-free widget, but that is not the same as a claim that every virtual-camera path carries transparency.
Make a short private recording or preview with the receiving software before you go live. Watch it back at the size and layout your audience will see. This catches practical issues that a preview window can hide: a clipped avatar, a camera source that is not selected, or a face that moves outside the frame when you turn. If your eventual channel uses a looped video rather than a live avatar, the planning questions are different; see how to build a looping product-video stream for that separate format.
Route the avatar into a stream or video chat
The manual describes a virtual broadcast feature, spelled “Virtual Barodcast” in that version, and says it requires driver support. The feature label may have changed. Find the current virtual-camera or broadcast control in VTuber Maker, enable it, and then open the application where you want the avatar to appear. In that application’s camera or video-source menu, select the VTuber Maker virtual camera if it is available.
The receiving application is a distinct step, not a continuation of avatar tracking. VTuber Maker must first be producing the avatar output; the stream or chat software must then capture the correct camera. If the receiving software shows your physical webcam instead, the avatar may still be tracking correctly in VTuber Maker. Change the selected input to the virtual camera and inspect the preview again.
For a livestream, add the avatar camera as a video source and size it in the scene. For a video call, select it as the camera input in the call application, then check its preview before joining. Exact menus vary by product and version, so look for the camera-selection control in the current interface. If no VTuber Maker virtual camera appears, confirm that the feature is enabled, that any required driver support is present, and that the receiving application has been restarted if it was open before the camera was enabled.
If your format is a pre-recorded continuous stream rather than a person performing live, an avatar-tracking setup is not the same as a video playout workflow. You can compare the trade-offs in ways to keep prerecorded video streaming continuously. For an on-camera VTuber performance, keep the tracking computer and its receiving stream application available for the duration you plan to broadcast; do not assume that a camera output continues if the applications or computer are closed.
VTuber Maker’s manual also describes a background-free, always-on-top widget and a Quick Action Panel. These may help if you want an avatar visible over other desktop content or want to trigger an action, but check their current behaviour and shortcuts in the application. A widget on your desktop and a virtual camera sent to another application are different output routes. Choose the one that serves the task, and test it in the actual receiving software rather than inferring its behaviour from the desktop preview.
Troubleshoot tracking and output
When the avatar does not move, work through the chain in order. First check that VTuber Maker is receiving the intended webcam image; close other applications that might hold the camera and verify that the correct device is selected. Then check that tracking is enabled and that your face is visible in the camera view. Only after those steps should you recalibrate or change avatar and scene settings.
If movement is present but feels wrong, return to your normal posture and use Calibrate as described in the manual. Check that the camera is steady, your face is not obscured, and you are sitting where you expect to sit during a stream. Make a modest change and test again. The point is to identify whether the issue is the reference position, camera view or avatar presentation rather than changing several settings without a clear reason.
If the avatar tracks but the mouth does not follow speech, check whether Audio Based Lipsync is enabled and whether the application receives voice input. Keep that diagnosis separate from face tracking: the manual presents voice-driven lip-sync as an optional control, not as a necessary condition for the avatar to move. If it is not needed for your presentation, leaving it off removes one variable.
If the avatar looks right in VTuber Maker but not in your stream or call, check output routing. Confirm the virtual broadcast or camera feature is enabled, any required driver support is available, and the receiving software has selected the VTuber Maker virtual camera. The manual’s version says driver support is required, but does not establish a current list of supported operating systems or receiving applications. Check Live3D’s current documentation if a camera is missing, rather than assuming compatibility from another user’s setup.
If the output appears but the framing is poor, adjust the avatar’s position and size in the receiving scene, then preview again. If you need a persistent broadcast that does not depend on your local computer staying switched on, that is a different operational problem from getting a webcam-tracked character into a live scene. StreamNeo removes the specific burden of keeping your own computer running for a file-based YouTube broadcast, but it is YouTube-only and does not replace a live webcam VTubing setup.
For a channel that is meant to remain live overnight, distinguish a continuous pre-recorded stream from a live performance requiring your camera and tracking application. The first can be designed around a prepared video source; the second depends on your live input and the applications carrying it. The 24/7 streaming cost guide can help you compare the operating implications of local equipment and cloud-based approaches without assuming they solve the same problem.
Make a short pre-flight check
Before a public session, confirm the webcam image is clear, the avatar is selected, tracking follows your ordinary movements, and calibration feels natural. If you use audio-based lip-sync, test it with your real microphone input. Check that the chosen background and framing suit the scene, and that the receiving software is using the virtual camera rather than the physical webcam.
Then inspect the final preview as a viewer would see it. Confirm that the avatar is not hidden behind a logo or chat panel, that its size is legible, and that the camera feed does not show an unintended window or desktop. For a video call, make this check before joining; for a livestream, use a private or unlisted test where appropriate and verify the scene before announcing a public start.
The official references are useful when a menu does not match a walkthrough: consult the VTuber Maker manual for the described tracking and output workflow, and the VTube Studio documentation only for understanding its separate Live2D model workflow. Product labels, drivers and supported receiving applications can change, so confirm those details in the current versions rather than transferring settings from another application.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Is Live3D the same as Live2D?
No. Live3D is the product name relevant to VTuber Maker in this guide, while Live2D is a separate 2D character technology. VTube Studio’s maintainer documents it as a Live2D application, so use VTuber Maker’s current instructions for this 3D tracking workflow.
Do I need a special webcam for VTuber Maker?
The manual describes VTuber Maker as using a webcam, but the reviewed research does not establish a specific camera model or minimum specification. Start with a camera that gives a clear, steady view of your face, then check Live3D’s current requirements if you are considering new hardware.
Why can my stream software not see the avatar?
The avatar can track inside VTuber Maker while the receiving application is still using the wrong camera. Enable the virtual-camera output, check the manual’s driver-support caveat, and select the VTuber Maker virtual camera in the receiving app’s camera menu.
Can I use this setup for a 24/7 prerecorded stream?
VTuber Maker’s webcam workflow is for tracking a performer in real time, so it is not the same as continuous prerecorded video playout. A live VTuber session depends on the camera and applications remaining available; choose a separate workflow if your channel needs a prepared video to continue while you are away.