TubeSum ← Transcribe a video

How To Make a 2D VTuber Model | Art - Rigging - Tracking [All In One Guide]

Transcribed Jun 14, 2026 Watch on YouTube ↗
Intermediate 15 min read For: Aspiring VTubers and digital artists with basic drawing skills who want to create their own 2D model.
57.1K
Views
3.7K
Likes
112
Comments
7
Dislikes
6.7%
🔥 High Engagement

AI Summary

This video provides a comprehensive guide to creating a 2D VTuber model, covering the entire pipeline from art creation and rigging in Live2D Cubism to facial tracking and streaming setup. The creator shares a free test model and walks through each step, including software recommendations, drawing techniques, parameter rigging, physics, and integration with V2 Studio and OBS.

[0:00]
Software Overview

Needed software: any art program that exports PSD (e.g., Photoshop, Clip Studio Paint, Procreate), Live2D Cubism (rigging, ~$96/year or $14/month, 42-day free trial), V2 Studio (tracking, free with watermark, $6.7 to remove, iPhone version $ for Pro), and OBS (free streaming/recording).

[2:28]
Art Preparation

Canvas size: small avatar 2048x2048 px, full body 4096x8192 px. Separate body parts into layers (hair bangs left/right, face features, clothes, etc.). Name layers clearly, combine line art and color, keep lip and face line art separate. Use clipping masks for shadows with multiply blending mode.

[6:29]
Live2D Interface and Mesh

Import PSD into Live2D, reassign clipping masks. Use automatic mesh generator (deformation heavy for important parts, standard for others). Copy mesh from face skin to face line art for synced movement. Edit texture atlas to arrange parts without overlapping or rescaling.

[14:36]
Deformers and Parameters

Warp deformer: green grid to move mesh. Rotation deformer: rotate parts. Stack deformers (e.g., eye deformer under face deformer). Parameters: Angle Z (head rotation), Angle X/Y (look left/right/up/down), link X and Y for corners. Use key forms (three green dots) to set poses.

[22:46]
Eye and Eyebrow Rigging

Rig left eye first, then copy-paste and reflect horizontally for right eye. Use deform pad edit for eyelashes. Set eye open parameter (0 closed, 1 normal, 1.3 surprised). For eyebrows: Bro Y (up/down), Bro L form (-1 sad, 0 normal, 1 happy). Transfer parameters from left to right using change/adjust.

[29:34]
Lip Sync and Body

Link mouth form and mouth open parameters. Use deform pad edit for lip line and mouth skin. Body X/Y/Z: create one big deformer for all body parts, link parameters, add key forms. Breathing parameter: create deformer on top, move up slightly for automatic breathing.

[32:56]
Physics

Add physics to hair and other parts using deformers and custom parameters. Use hair move front/side/back parameters. In physics settings, add input (head) and output (parameter), adjust pendulum settings for movement intensity. Multiple pendulums for more flow.

[35:44]
Export and V2 Studio Setup

Export as OC3 file with physics settings. In V2 Studio, import model, connect webcam or iPhone (same Wi-Fi). Set up expressions and hotkeys for toggling extra objects. Enable transparent background for capture. In OBS, add game capture source, select V2 Studio window, allow transparency, crop if needed.

The video successfully delivers a complete, step-by-step tutorial for creating a 2D VTuber model, from drawing to streaming. It emphasizes using free trials and provides a free test model, making the process accessible to beginners.

Clickbait Check

95% Legit

"Title accurately describes the comprehensive guide covering art, rigging, and tracking."

Mentioned in this Video

Tutorial Checklist

1 2:28 Choose canvas size (2048x2048 for half-body, 4096x8192 for full body) and draw character with separated layers for each body part.
2 6:29 Import PSD into Live2D Cubism, reassign clipping masks, and generate meshes for all parts using automatic mesh generator.
3 12:33 Edit texture atlas: arrange parts without overlapping or rescaling, ensure quality remains unchanged.
4 17:22 Rig head rotation (Angle Z): create rotation deformer, add key forms, rotate for left and right.
5 19:01 Rig look direction (Angle X/Y): link parameters, create deformer for head, add key forms for left/right/up/down, adjust individual features like nose.
6 22:46 Rig left eye: set eye open parameter (0 closed, 1 normal, 1.3 surprised), use deform pad for eyelashes, deformers for upper/lower eye skin. Copy-paste and reflect for right eye.
7 29:34 Rig mouth: link mouth form and open parameters, use deform pad edit for lip line and mouth skin.
8 30:44 Rig body X/Y/Z: create one deformer for all body parts, link parameters, add key forms for movement.
9 31:34 Add breathing: create deformer on body, add key forms, move up slightly for automatic breathing.
10 32:56 Add physics: create deformers for hair, use hair move parameters, set up physics settings with input (head) and output (parameter), adjust pendulum.
11 35:44 Export model as OC3 file with physics. In V2 Studio, import model, connect webcam or iPhone, set up expressions and hotkeys.
12 41:02 In OBS, add game capture source, select V2 Studio window, enable transparency, crop if needed, and start streaming.

Study Flashcards (15)

What are the four main software needed to make a VTuber model?

easy Click to reveal answer

An art program (e.g., Photoshop), Live2D Cubism, V2 Studio, and OBS.

What is the cost of Live2D Cubism per year and per month?

easy Click to reveal answer

Approximately 14,000 yen ($96 USD) per year or 2,000 yen ($14 USD) per month.

0:28

How long is the free trial for Live2D Cubism?

easy Click to reveal answer

42 days.

0:40

What is the recommended canvas size for a full-body VTuber model?

medium Click to reveal answer

4096 x 8192 pixels.

3:03

Why should you separate body parts into different layers when drawing?

medium Click to reveal answer

To make rigging easier and allow independent movement of each part.

3:30

What does the multiply blending mode do for shadows?

medium Click to reveal answer

It removes white color from the layer, making it easy to create shadows.

5:46

How do you reassign clipping masks in Live2D after importing?

hard Click to reveal answer

Select the shadow part, go to inspector, and choose the body part to clip to from the clipping ID dropdown.

7:54

What is the purpose of a warp deformer?

medium Click to reveal answer

To move the mesh below by manipulating its green grid points, allowing flexible deformation.

14:36

How do you make the model look towards corners (e.g., upper left)?

hard Click to reveal answer

Link Angle X and Y parameters, then use the 'set size corner' function to auto-generate corner poses.

19:01

What is the parameter range for eye open (surprise eye)?

medium Click to reveal answer

0 (closed), 1 (normal open), 1.3 (surprised wide).

23:24

How do you copy rigging from left eye to right eye?

hard Click to reveal answer

Select left eye parts, copy-paste, then use 'reflect horizontally' to flip. Transfer parameters using change/adjust options.

26:49

What is the purpose of the breathing parameter?

medium Click to reveal answer

To make the model automatically breathe by moving the body deformer up and down slightly.

31:34

How do you add physics to hair in Live2D?

hard Click to reveal answer

Create a deformer for the hair, use hair move parameters, then in physics settings add input (head) and output (parameter), adjust pendulum settings.

32:56

How do you enable transparent background in V2 Studio for OBS?

medium Click to reveal answer

Go to background settings, select color picker, and turn on 'transparent in capture'.

40:36

What OBS source type should you use to capture V2 Studio?

medium Click to reveal answer

Game capture, set to capture specific window (V2 Studio), with allow transparency enabled.

41:25

💡 Key Takeaways

💡

Complete VTuber Pipeline Overview

Provides a clear roadmap of all steps from art to streaming, setting expectations for beginners.

🔧

Layer Separation for Rigging

Emphasizes the importance of separating body parts into layers, a crucial technique for efficient rigging.

3:30
🔧

Stacking Deformers for Complex Movement

Explains how nesting deformers (e.g., eye under face) creates realistic combined movements.

14:36
🔧

Copy-Paste Rigging for Symmetry

Demonstrates a time-saving method to rig one eye and duplicate it for the other, reducing workload.

22:46
🔧

Physics Pendulum Settings

Details how to adjust pendulum parameters to control the intensity of physics, giving life to the model.

32:56

✂️ Creator Tools: Viral Hooks

AI-generated clip ideas for Shorts based on the transcript

How to Make a VTuber: Software & Costs

51s

This segment provides a clear, actionable breakdown of the essential software and costs, which is highly valuable for beginners and likely to be saved and shared.

▶ Play Clip

Free VTuber Model & Art Tips

60s

Offering a free model and practical art tips (canvas size, layer separation) is a generous and useful hook that attracts aspiring VTubers.

▶ Play Clip

Live2D Rigging: Mesh & Deformers Explained

60s

Demystifying complex rigging concepts like mesh and deformers in a simple, visual way is highly educational and likely to engage viewers struggling with the technical aspects.

▶ Play Clip

Copy-Paste Eyes: Save Time Rigging

60s

The time-saving tip of copying and flipping eye rigging is a clever hack that appeals to efficiency-minded creators and is highly shareable.

▶ Play Clip

VTuber Physics: Make Hair & Clothes Move

60s

Adding physics to a VTuber model is a visually impressive and satisfying step that demonstrates the 'magic' of making a model come alive, which is highly engaging.

▶ Play Clip

[00:00] I'll try to go through everything in

[00:01] this video from modeling, rigging,

[00:03] tracking, and recording. What software

[00:06] is needed to make a VTuber? First, for

[00:08] art program, it can be any application

[00:11] as long as it can export as Photoshop

[00:13] file program like Photoshop, Cliff

[00:16] Studio Paint or Procreate will work just

[00:19] fine. Second, live 2D cubism to rig the

[00:22] model for the art program. You can

[00:24] download the software at

[00:26] robblifeet2d.com

[00:28] English. It costs around 14,000 yen or

[00:32] 96 USD per year or 2,000 yen or 14 USD

[00:37] per month by the time of recording. I

[00:40] downloaded the 42day free trial version

[00:43] and managed to finish rigging my model

[00:45] before the trial end. So, unless you're

[00:47] going to be using this application for a

[00:49] long time, I suggest you try out the

[00:51] 42-day trial first. Third, B2 Studio.

[00:55] After you finish rigging in Live 2D, you

[00:58] can import the file to V2 Studio to

[01:00] track your facial expression with your

[01:02] model. It is free on stream, but cost

[01:05] around 6.7 USD to remove the watermark

[01:08] using webcam.

[01:10] Otherwise, you can use an iPhone and

[01:13] install the app to link to your iPhone

[01:15] camera. The iPhone version would cost

[01:18] for the US to upgrade to Pro. If you

[01:21] don't upgrade on the iPhone, the

[01:23] connection between iPhone and PC will be

[01:26] lost every 10 minute or so. The benefit

[01:28] of using an iPhone instead of a webcam

[01:31] is that it can track the facial motion

[01:34] more accurately than a webcam does.

[01:37] Fourth, OBS is a software for live

[01:40] streaming or recording videos. I highly

[01:43] recommend this software because it is

[01:45] free and what most people use. It can

[01:48] record models from V2 Studio and

[01:50] recognize a transparent background.

[01:53] Overall, this is what you need and how

[01:55] much it costs to make and use a V2

[01:57] model. Mainly does cost you money if you

[02:00] want to go professional, but if you want

[02:02] to just test it out, you can try using

[02:04] all the trial version. Now, I'll go over

[02:07] what asset you need and how to use the

[02:10] software. If you don't want to learn

[02:12] what to draw and skip to the rigging

[02:14] tutorial. I drew this model for free

[02:17] that you can use. I tried to make this

[02:20] model as simple and easy to rig as

[02:23] possible so that we can test it out

[02:25] together. Link is in the description.

[02:28] You can use this model for anything

[02:30] after it as well. Art program. First

[02:33] thing that we have to decide is what

[02:35] canvas size will it be on the screen is

[02:38] how many pixel for a small avatar half

[02:41] body full body super detail and alter

[02:45] detail is the file that I gave for the

[02:48] test is48x48

[02:52] pixel wide it is enough if you just

[02:55] using it to stream on the corner since

[02:58] you don't need a leg for that just need

[03:00] your upper hand meanwhile This full body

[03:03] model that I've been using to stream is

[03:06] 4,96 pixel x 8,192

[03:11] pixel. You can see how far you can zoom

[03:14] in and out and how detailed it is. Now

[03:17] it's time to start drawing. Start off

[03:19] with sketching symmetrically. You can

[03:22] make a model distorted, but it would be

[03:26] harder to rig. When we start drawing in

[03:30] detail, we need to make sure that we

[03:32] separate the body parts in different

[03:34] layer. Here are the different parts.

[03:37] Different hair bangs, left and right,

[03:39] separated face features like the blush,

[03:43] nose, eyebrows, eyelashes, upper eye

[03:48] skin that we're going to be using to

[03:50] cover the eye when we move later on.

[03:53] Lower eye skin as well.

[03:56] I shine the pupil, the eyeball,

[04:00] upper lip,

[04:02] upper mouth skin to cover the what's

[04:06] underneath it. Lower lip, lower mouth

[04:10] skin, the mouth inside,

[04:13] face line, face skin, and ear.

[04:18] And then we got the clothes, the

[04:21] neckline, neck skin, the neck shadow,

[04:25] body, clothes behind the body, hair

[04:28] behind the body. Make sure to name the

[04:31] layer so it doesn't get confusing later

[04:33] on. Label the left and right side of the

[04:35] body part as L underscore and R

[04:38] underscore, or you can just type left

[04:41] and right. The line art and the color

[04:44] should be combined when you finish

[04:46] drawing. Like the arms should just be

[04:49] the arm layer. The hair should just be

[04:52] the hair except for the lip line and the

[04:56] face line art. Those should be

[04:57] separated. Now, this part is optional.

[05:00] We can draw a mask to erase the face

[05:04] line art during rigging when the

[05:06] character faces a different direction.

[05:09] You can use an airbrush to make this

[05:11] mask and then turn off the visibility.

[05:14] As for the face shadow, at the moment it

[05:17] look like a mess because it's out of the

[05:19] border. But don't worry, this is

[05:21] intentional since most art program have

[05:23] the ability to use clipping mask. When I

[05:26] click on the layer and turn on clipping

[05:28] mask, the shadow will only be inside of

[05:31] the layer underneath. Now it's not

[05:34] messy. And during the rigging, using a

[05:37] clipping mask will make the shadow look

[05:39] pretty dynamic. Then proceed to change

[05:42] the shadows blending mode into multiply.

[05:46] What multiply does is that it removes

[05:48] the white color from the layer that you

[05:50] set it to. This is what a lot of digital

[05:53] artists use to make easy shadow. And

[05:56] that's it for the main parts of the

[05:58] model. Here are question that I had

[06:00] while running like what about different

[06:03] clothes or expression. You can rig to

[06:06] change expression like blushing, crying

[06:10] or draw extra hand that turn on and off

[06:12] later. If you want those part in, you

[06:15] can draw those part and then turn off

[06:17] the visibility for now. And now we're

[06:19] done. Save the file as Photoshop file

[06:22] and it's time to head on to the next

[06:24] program or you can take a rest, get some

[06:27] snacks and come back later.

[06:29] Live 2D cubism. Overall, I'll be going

[06:32] in these orders. Interface, mesh, tools,

[06:37] parameter for rigging, physics, then

[06:40] exporting. The timestamp will be in the

[06:43] comment if there are specific parts that

[06:45] you want to skip to. Open up live 2D

[06:48] cubism.

[06:49] Click new and on this box drag the

[06:53] Photoshop file in. It will ask what

[06:56] resolution for displaying the model

[06:58] image. This doesn't affect the actual

[07:01] model. So just click one out preview.

[07:07] Now you see this interface but make sure

[07:09] that you're in the correct interface.

[07:11] Over here, there will be a dropown that

[07:13] you can pick modeling, animation, form,

[07:16] animation. Make sure you're in the

[07:18] modeling because this is where we'll

[07:20] begin rigging. All right. Now, looking

[07:23] at the part window, this is where all

[07:25] layers that you had in the art program

[07:27] will be shown. The part window will be

[07:30] on the top left of the screen by

[07:32] default. However, if you can't find this

[07:35] window there, go to the window menu here

[07:39] and it will show all the windows that

[07:41] are available that you can turn on and

[07:43] off if you can't see it. Back to the

[07:47] model, the recognizable difference is

[07:49] that my shadow is not being clipped

[07:51] masked. We'll need to reassign all the

[07:54] clipping mask again. Go to where the

[07:56] shadow is in the parts. Once you click

[07:59] on the shadow part, the setting for this

[08:01] part will be shown in the inspector

[08:04] window. This is where you can edit the

[08:06] name and clipping ID of the part. Go to

[08:09] the pop- down of the clipping ID and

[08:12] find the body part that the shadow is

[08:15] going to be clipped to. In this case,

[08:17] the neck. Once selected, this shadow

[08:21] part will be clipped inside.

[08:23] Now repeat this process for the other

[08:26] shadows or part that have to be cliff

[08:28] too like the face shadow and the

[08:31] blushes.

[08:34] Now before I move on, if you have edited

[08:37] or drawn in new parts in the art program

[08:40] and want to update it here, you can just

[08:43] drag the newly updated Photoshop file

[08:46] into live 2D over here and it will give

[08:50] you an option to create new file or

[08:52] choose the already existing file. Click

[08:55] that and after that the part that have

[08:58] the same name will be replaced and the

[09:00] part with names that didn't exist before

[09:03] will be added in a new folder. So you

[09:05] don't need to be afraid if you want to

[09:07] make changes to your model. Another

[09:09] window I want to explain is the project

[09:12] window. It looks similar to the parts

[09:15] window but it's there for a different

[09:17] purpose. You can see what Photoshop file

[09:20] is in the current live 2D model. The

[09:22] useful thing about the project window is

[09:25] that if you accidentally delete a part

[09:28] in the parts window and you want it

[09:30] back, you can head to the project window

[09:33] to find it and add it back in by

[09:36] right-clicking and clicking create art

[09:39] mesh. This part will then be readited

[09:42] into the part window. Now that you have

[09:44] gotten the model in live 2D, don't

[09:46] forget to save. Press Ctrl S on your

[09:50] keyboard or go to file save as. The file

[09:54] for editing in live 2D is C O file. So

[09:59] if you're tired and want to edit another

[10:01] day, this is what you want to open to

[10:04] edit it again. And with that, we move on

[10:07] to the mesh. To be able to make the part

[10:10] flexible, you have to give every parts a

[10:13] mesh first. What is a mesh? a mesh or

[10:16] these spiderweb looking thing. When you

[10:19] move the dot, it will move the image

[10:22] that it's assigned to. This is what help

[10:25] the make part looks a lot more flexible.

[10:28] Now to add a mesh, click on a part. Then

[10:32] using this two tools up here. The first

[10:34] one is called manual mesh edit. You can

[10:38] use this to add mesh point one by one.

[10:41] However, it's very inefficient. So,

[10:43] we're going to be using the second tool

[10:46] more often, the automatic mesh

[10:49] generator. Instead of creating a mesh

[10:51] point by point, the program will do it

[10:53] for you. When pressed, the toolbox will

[10:56] appear. You can choose from the press

[10:58] and drop down standard deformation

[11:01] little deformation heavy. The less

[11:04] certain mesh is, the less fluidity there

[11:08] will be.

[11:09] So I suggest using deformation heavy for

[11:12] important parts like the face, the eye,

[11:16] the hair, the lips and standard for

[11:19] parts that won't move as often. Now make

[11:22] sure that every part you have in the

[11:24] file has a mesh by selecting all the

[11:27] parts and clicking on the automatic mesh

[11:30] generator.

[11:33] Now we want the face part and the face

[11:35] line art part to have the same exact

[11:38] mesh so that when we move one the other

[11:42] move the exact same way. If it have a

[11:44] different mesh it will not sync with

[11:47] each other. So create automatic mesh

[11:51] then go to manual mesh edit.

[11:54] Select all of the mesh and on your

[11:57] keyboard press Ctrl C to copy all of it.

[12:01] Then click the shack button. Now we go

[12:03] to the faceeline arter parts. Click

[12:06] manual mesh edit and if there's already

[12:08] a mesh in there, select them all and

[12:11] delete it. Click by clicking the delete

[12:13] button. Then press Ctrl +V to place what

[12:17] we just copied in. And after that, click

[12:19] the check button. And now these two

[12:22] parts have the exact same mesh. Do the

[12:25] same for the neck. If you separated the

[12:28] line art and the skin,

[12:32] if you're done with adding mesh for the

[12:33] parts, we can move on to edit texture

[12:36] atlas. This is what texture atlas look

[12:39] like. This is where we'll lay all the

[12:41] parts into a page for export. When press

[12:44] the edit texture atlas button, it will

[12:47] ask name, size, and default layout. For

[12:50] the name, you can just leave it as

[12:52] texture atlas one. For the size, it

[12:55] depends on how big your canvas size is.

[12:58] Once again, there's different pixel size

[13:01] for each one. Recommend that you make it

[13:04] square size. If you're using this test

[13:06] model, you can just use48

[13:09] *48.

[13:11] For default layout, picked do not place.

[13:14] And now we go okay. And this is a window

[13:17] that will appear. On the right here is

[13:20] the model image list. It is all the

[13:23] parts that you have in your model.

[13:25] Double click on the part to place it in.

[13:28] You can move it around in this texture

[13:30] atlas. What we want to do here is place

[13:33] the textures into this sheet without

[13:36] rescaling and without overlapping the

[13:38] parts.

[13:40] [Music]

[13:41] Now, if all the parts aren't enough to

[13:44] fit in a single page, you can add

[13:46] another page by clicking the add texture

[13:49] button up here. With the second page,

[13:52] you can place the remaining parts that

[13:54] couldn't fit in.

[13:58] [Music]

[14:01] And once you're done, press okay.

[14:04] As said on the bottom right, if you

[14:06] press T on your keyboard, you can swap

[14:08] between what the quality look like

[14:10] before and after texture atlas. If

[14:13] nothing change when you press the T

[14:15] button, that's mean you did a great job.

[14:18] But if it looks different and it looks

[14:21] weirdly wonky, then maybe you

[14:23] accidentally overlap some parts or

[14:26] descale the parts to make it look a lot

[14:29] more blurry. Now, after you're done with

[14:31] texture atlas, we move on to the other

[14:33] tools you use to rig your model.

[14:36] Starting with the warp deformer, select

[14:38] on any part and click on the warp

[14:40] deformer. This panel will pop up. Just

[14:43] click create. Now, there will be this

[14:45] green grid box. This is a deformer. If

[14:49] you move the point on this deformer, it

[14:51] will move the mesh below. You'll be

[14:53] using this tool a lot. Now, when you

[14:55] click on the deformer, the setting for

[14:57] the deformer will appear on the

[14:59] inspector. You can redeem deformer here.

[15:02] Trust me, you won't regret naming things

[15:04] since there will be a lot of deformers.

[15:08] Another thing in the inspector is

[15:10] changing the number of division. So if

[15:13] you want to make there be more division,

[15:16] you can play around with these numbers

[15:19] and it will make you have more control

[15:22] points. To see every deformer that has

[15:25] been created, you can look at this

[15:27] deformer window on this bottom left.

[15:30] This is like the folder system for the

[15:32] deformers. It help you see and rearrange

[15:35] what deformer is in each other. It's a

[15:38] very helpful window and you'll be using

[15:41] it a lot. The reason to use deformer to

[15:44] wreck instead of controlling the mesh

[15:46] directly is because you can create a

[15:48] deformer on another deformer. It works

[15:51] quite like a folder. Like I can have a

[15:54] deformer for the eye and another

[15:56] deformer for the face turning which make

[15:59] it to where the eye deformer which is

[16:02] under the face deformer will get

[16:04] deformed along as well. And by stacking

[16:08] things up like this, that's how we make

[16:10] our model become alive.

[16:12] Next is the rotation deformer and

[16:15] rotation deformer creation tool. Just

[16:17] like the warf deformer, it will control

[16:19] the mesh and be put in the deformer

[16:22] window. However, instead of moving the

[16:24] points, you rotate instead. The

[16:27] difference between rotation deformer and

[16:29] rotation deformer creation tool is that

[16:32] the first one will spawn the rotation

[16:34] deformer in the middle of your parts

[16:37] while the second one lets you create the

[16:40] rotation deformer wherever you want. And

[16:43] that's the tree deformer tools. Now we

[16:46] head to arrow lasso and brush tool which

[16:49] is pretty self-explanatory.

[16:51] You use the arrow to select different

[16:54] tools. You use the lasso to select

[16:57] multiple things at once and you use the

[17:02] brush selection to select things like a

[17:04] brush.

[17:06] And now we go to deform pad edit. This

[17:11] allows you to make a part used to edit

[17:13] the mesh in a line easily. This is great

[17:16] for the lips, the eyelash and the

[17:18] eyebrow. Now we move on to actually

[17:22] rigging the model. For that knowing the

[17:24] parameter is very important. When you go

[17:28] to parameter window you can see these

[17:30] defaults parameter given. These

[17:33] parameters are empty and you'll have to

[17:35] fill them up. Let's start with angle C.

[17:38] This is for rotating the head. So what

[17:41] we have to do is select all of the head

[17:44] and create a rotation deformer creation

[17:48] tools. Now drag the rotation deformer

[17:51] from the shin to the head.

[17:54] Now go to the arrow tool instead. Click

[17:57] on angle C and press this tree green dot

[18:01] button. They are the key forms. The

[18:03] middle is looking straight. The left is

[18:07] for rotating to the left and the right

[18:10] is for rotating to the right. So when we

[18:12] move the red dots around to the left, we

[18:15] can rotate the deformer to the left. And

[18:19] when we move the red dot to the right,

[18:22] we can rotate the deformer to the right.

[18:25] And now when we move the red dot around

[18:27] the parameter, it will rotate the head

[18:30] around as it should. And this is how

[18:34] parameters work and what we're going to

[18:37] be doing for the rest of it. To

[18:39] summarize again, select all parts and

[18:42] create a deformer. Click on the

[18:44] parameter. Add key forms. Go to the key

[18:49] form you want to edit. Move the deformer

[18:52] around. And there we go. That's how you

[18:55] rig every part of the model. Now, angle

[18:58] C is the easiest one. Let's move on to

[19:01] the harder ones like angle XY. Angle X

[19:06] is for looking left and right, while

[19:08] angle Y is for looking up and down. Now

[19:11] we want the model to be able to look

[19:13] towards the upper corner as well like

[19:16] upper left, upper right, lower left or

[19:18] lower right. To do that we have to

[19:20] combine angle X and Y by pressing this

[19:23] link button right here. When we press it

[19:25] the parameter will now look like this

[19:28] graph. Now let's get to rigging by first

[19:30] selecting all the head parts and create

[19:33] a new deformer. Call it BF angle XY

[19:38] head. Then select the angle XY parameter

[19:41] and click this three green dot icon

[19:43] again. Go to the left in the parameter

[19:47] and make sure you snap to it by right

[19:49] clicking. Right clicking is how you snap

[19:51] to these dots. Now turn the face to the

[19:55] left. It will look pretty bad right now

[19:57] cuz we have a lot of adjusting to do.

[20:02] And after you're done turning it left,

[20:05] you'll want to turn it right.

[20:09] And now your face is turning. It doesn't

[20:13] look so good right now though because

[20:15] you want to move individual facial

[20:17] features to create more dimension like

[20:19] the nose. So we can go to the nose and

[20:22] create another deformer. Call it df

[20:27] anglexy

[20:29] nose. And then when you look to the left

[20:33] the nose may protrude a bit more.

[20:36] [Music]

[20:40] And now you're going to write it may

[20:41] protrude a bit more as well. So now it

[20:44] looks better.

[20:46] And do the same for all the other side

[20:48] as well. Looking to the top

[20:51] and looking to the bottom. Just adjust

[20:55] it around. This will take quite some

[20:57] time. And then after you're done with

[21:01] left, right, top, bottom, instead of

[21:04] working on the corners right away, we

[21:07] can have the program try to work on it

[21:09] first. Go to the center of the

[21:12] parameter. Then in the menu, click

[21:14] modeling parameter, set the size corner,

[21:19] and okay. So the software have now

[21:23] created the corners for you. When you

[21:26] move around the angle XY parameter, the

[21:30] model will face towards that direction.

[21:32] You can fix them manually if it looks

[21:34] bad though. And with a lot of adjustment

[21:37] and on a lot of time, angle XY is kind

[21:41] of dumb.

[21:44] Now remember the face mask, we can use

[21:48] that here to make the face line

[21:50] disappear a little at a certain angle.

[21:53] Turn on the visibility of the mask.

[21:55] Second, select the face line art and on

[21:58] clipping ID, make it clip to the face

[22:01] mask.

[22:02] [Music]

[22:04] Then we click this reverse mark button.

[22:07] Turn the opacity of the face mask to

[22:10] zero.

[22:12] We can create a deformer for the mask

[22:15] and in angle XY add the key form and

[22:18] move the mask around. This isn't

[22:20] necessary, but I see most Btuber use

[22:23] this technique. You can check out the

[22:26] preview of the model by going to the

[22:29] menu modeling physics settings. Ignore

[22:33] the weird setting on the left right now.

[22:35] Looking at the screen on the right, when

[22:37] you click and hold the screen, the model

[22:39] will look toward your cursor. This is

[22:41] where you can go to check whether your

[22:43] model is working or not. Moving on to

[22:46] the eye parameters.

[22:48] First, we'll only be doing the left eye.

[22:51] And then after we're done rigging the

[22:53] left eye, we'll then copy paste the left

[22:56] eye to the right and flip it. So that

[22:59] will be less work to do. Looking at the

[23:02] IL open here, you will see that the left

[23:05] is zero and the right is one. Previously

[23:09] in angle X, Y and Z, the parameter can

[23:13] go to the negative to indicate left or

[23:16] bottom and positive to indicate right or

[23:19] up. For the eye, however, there's only

[23:21] close and open. Close being zero and

[23:24] open being one. Now, if you want a

[23:27] surprise eye, you have to increase the

[23:30] parameter to 1.3 so that the eye would

[23:34] widen more. To change that, you can

[23:37] right click the parameter and edit.

[23:40] Change the maximum to 1.3.

[23:45] Starting with the eyelash, select the

[23:48] eyelash and since it's a line instead of

[23:51] a deformer, we'll use the deform pad

[23:53] edit tool instead. Click along the

[23:56] eyelash to create control points. These

[23:59] control points will let you move the

[24:01] eyelash freely. Go to the arrow tool on

[24:04] the box over here. Type one so we make

[24:07] sure we're in the one and add key form

[24:11] at this exact location. And instead of

[24:13] clicking the three dots in the

[24:15] parameter, click the two dots instead.

[24:18] So now zero is I closed, one is I open

[24:24] normally, and 1.3 is I being wide. And

[24:29] we just move it around at these

[24:31] locations.

[24:34] And now we go to the upper eye skin.

[24:38] Create a deformer for the upper eye

[24:40] skin. and add the same key form to the

[24:43] parameter

[24:45] 0 1 1.3 and drag the deformer so that it

[24:49] covers the eyeballs when the eye is

[24:51] closed or open.

[24:53] [Music]

[24:54] Now go to lower eye skin and do the same

[24:58] process

[25:01] and after that that's it for open

[25:04] parameter.

[25:06] Next is IIL Sumal. Here we can make the

[25:10] blush more visible. Go to the blush part

[25:13] and turn it on if it's off. Now add two

[25:16] key forms using the two green dot button

[25:19] here. On zero we'll turn the opacity of

[25:22] the blush to zero. And on one we'll turn

[25:26] the opacity to 100. So that's how we

[25:30] make it seems as if the model is

[25:33] smelling. Next is eyeball XY.

[25:36] X is for looking left and right. Y is

[25:40] for looking up and down. Connect the X

[25:42] and Y together so that we can look at

[25:44] the corners as well. Select the pupil

[25:47] parts and create a new deformer. Click

[25:50] on the three green dots and start moving

[25:53] the pupil around. This is basically the

[25:56] same thing we did for face turns of

[25:59] angle XY timeline.

[26:03] [Music]

[26:05] And there we go. And then regrue the

[26:08] eyebrows. Create a deformer for eyebrow

[26:10] Y for moving up and down.

[26:15] And another deformer for bro L form.

[26:19] Negative one is sad, zero is normal, and

[26:23] happy is one. We can just delete bro

[26:27] angle and bro X because YouTube Studio

[26:31] doesn't really detect eyebrow moving

[26:34] left and right or bro angle. Also, you

[26:37] can make a deformer for the entire eye

[26:41] and move it around ankle X line to see

[26:44] if different direction the eye looks

[26:46] good and or not. Okay, now that your

[26:49] left eye is nice and done, we're moving

[26:52] on to copy pasting it to the right.

[26:55] Select all of your right eye asset and

[26:58] just delete them cuz we don't need them.

[27:03] And then select all of your left eye,

[27:06] copy paste.

[27:08] [Music]

[27:11] Now after you're done copy pasting it,

[27:13] go to modeling, edit form, reflect,

[27:18] click reflect horizontally, then click

[27:21] okay. Now we managed to create the right

[27:24] eye. Okay, we have to transfer the

[27:28] parameters from left to right as well.

[27:31] Starting with IIL open, click the number

[27:34] box and choose change.

[27:38] Go to IR open parameter. Now when you

[27:42] move the IR open parameter, it should be

[27:46] the right eye. Now also do for same for

[27:49] eyebrows. Click BR L Y change to Bro R Y

[27:56] and also Bro L form change to bro R

[28:00] form.

[28:02] [Music]

[28:04] Now as for the eyeballs, it's now

[28:07] looking at the opposite direction. In

[28:10] eyeball X parameter box, click adjust

[28:13] instead. In value after change, put in 1

[28:18] 01

[28:21] and click okay. Now the eye should look

[28:24] the same way. If it looks weird, you can

[28:27] adjust it around. If you edited the IL

[28:30] open parameter to 1.3 for surprise eyes,

[28:34] don't forget to do the same for IR open

[28:36] as well. And you can also make a

[28:39] deformer for the eye shines when it

[28:41] looks at different angles as well. Make

[28:44] sure to rename the folder into I write

[28:47] instead. And we want to rename

[28:49] everything else in this folder from left

[28:52] to right as well. To change it easily,

[28:55] go to edit search and replace on

[29:00] condition. Type L underscore or left.

[29:04] And now we have to select all of the

[29:05] copied left eye to make sure that it is

[29:08] the copy part. on the left parent column

[29:11] it should say that the parent is R I

[29:15] select all of it and on the bottom here

[29:18] before replacement as L underscore and

[29:21] after replacement to R underscore then

[29:24] click replaced now all the left part

[29:27] that got copied will be named right so

[29:30] that it won't be confusing but now the

[29:32] eye should be completed

[29:34] now we move on to lip sync if you manage

[29:38] to come this far then the mouth isn't so

[29:40] different. Link the mouth form and mouth

[29:44] open together. The bottom middle 0 0

[29:48] should be mouths closed. Up one should

[29:52] be mouth open normally. To the left in

[29:55] negative is sad and to the right is

[29:59] happy. Now use the deform pad edit tool

[30:02] for the lip line and move it as you

[30:04] want. You don't have to make it exactly

[30:08] the same. and can go freestyle with it.

[30:14] After you move the lift, then you move

[30:16] the upper lip skin to conceal the mouth

[30:18] inside. You can also use the deform pad

[30:21] edit for the mouse skin as well by

[30:24] creating the pad along the lips. And

[30:26] then timelapse. It's going to take some

[30:29] times to move things around, but I think

[30:32] now you get what to do. It's all the

[30:36] same process, just um different parts.

[30:39] And tada, we we did it. The mouth.

[30:44] And moving on to body X, Y, Z. Body X is

[30:49] moving the body to left and right. And Y

[30:52] is moving body up and down. And C is if

[30:55] you tilt your body like the hips and the

[30:57] shoulder. And since we don't have the

[31:00] lower body here, we will just ignore

[31:03] body scene. Now go to the deformer

[31:05] window, collapse the deformer tabs and

[31:09] select every part. Now make one big

[31:12] deformer and call it deformer_bodyxy

[31:18] underscore body. Link the power meter

[31:21] together. Create key forms and move the

[31:25] deformer just a bit left, right, up, and

[31:28] down. Synthesize the corner. And now

[31:31] we're done with body X and Y. Next is

[31:34] the breeding parameter. This is a

[31:36] parameter that will play on

[31:38] automatically. Zero is normal, one is

[31:41] breed in. So select the body XY deformer

[31:45] and create another deformer on top.

[31:48] Create two key forms and on one just

[31:52] move the deformer up a little bit. And

[31:55] now when we check in the physics

[31:56] setting, we can see the character

[31:59] breeding automatically.

[32:01] And with that, you filled in all default

[32:04] given parameters. Don't forget to save

[32:08] extra segment. If your model has extra

[32:12] objects like different hand, a mic, or a

[32:15] pet that you want to be able to toggle

[32:17] on and off, you can make a toggle by

[32:20] creating a new parameter.

[32:22] Click on the plus button down here and

[32:25] name it whatever extra object that is

[32:28] and make the minimum zero. All there is

[32:31] to do is select those extra parts and at

[32:34] zero make it zero opacity and at one

[32:38] make it 100 opacity and that's it to

[32:41] create the toggle which we can use to

[32:44] later turn it on and off in YouTube

[32:46] studio. And there we go. The model will

[32:49] work. In fact, you don't even need to

[32:50] use this next step. But if you do, it

[32:53] will give your model some life to it.

[32:56] Physics. Physics is where the magic

[32:59] happen. It makes this turn into this.

[33:02] [Music]

[33:05] First, the hair is where there is most

[33:07] physics. Seect the front hair and create

[33:10] a deformer. Name it something like um

[33:13] DFP

[33:15] front hair. Our hair is really simple,

[33:17] so we can just use a default hair move

[33:20] front parameter. Negative one is sway

[33:23] left and one is sway right. To make it

[33:26] easy after moving negative one parameter

[33:29] go to modeling motion mirroring.

[33:34] Okay. And the program works on the right

[33:36] side of the parameters for you. We then

[33:40] continue to do the same for hair move

[33:42] side and hair move back. Go to physics

[33:46] setting. That's on the modeling physics

[33:49] setting. Now we don't ignore the setting

[33:52] on the left. Click the add dot dot dot

[33:54] button and name it hair

[33:58] input present as head input physic model

[34:01] presence as hair short. Then click okay.

[34:06] Now over here click the output setting.

[34:09] Then click the button add underneath it.

[34:12] Search for the parameter that we just

[34:14] made for this part. Check it on. Then

[34:17] click okay. Now you'll notice when you

[34:19] move the model left and right, there is

[34:21] now physics on your part. If you think

[34:24] it moves too much, then you can adjust

[34:25] the number in the pendulum setting down

[34:27] here. Tweak the number around and see

[34:30] what you like. Repeat this step for

[34:32] every parts that you want to have

[34:34] physic. Like this jacket strings doesn't

[34:37] have a default parameter, so we need to

[34:39] create a new one. Click on the plus

[34:41] button below the parameter and okay. and

[34:45] do the same thing where you sway it left

[34:47] and right and then add it into the

[34:49] physics settings. This time it's the

[34:52] body input.

[34:58] [Music]

[34:59] Now if you want the hair to flow more,

[35:02] we can create another deformer on top of

[35:04] the same hair part. Let's try this with

[35:07] the back hair. Then create a new

[35:09] parameter for another physics on the

[35:12] part.

[35:13] Move it a little on the tip of the hair.

[35:16] Then add this new parameter into the

[35:19] physics settings.

[35:26] Create another pendulum.

[35:28] Change the pendulum number of the new

[35:30] parameter to two. And now you can see

[35:34] this hair part moves more than before.

[35:36] If your model has many long hairs, you

[35:39] can add as much pendolums as you want to

[35:41] give it more physics. And with this,

[35:44] we're pretty much done with the model

[35:46] and can move on to exporting. To export,

[35:49] just go to file, export for runtime,

[35:53] export as OC3 files. Make sure the

[35:57] export physics setting file is checked

[36:00] on as well. Then press okay. Choose the

[36:03] folder that you want to save to. Create

[36:06] a new folder, named it, and save it

[36:09] there. This is what it should look like

[36:11] containing the texture atlas and

[36:13] everything. And we're good. We studio.

[36:17] Now we move on to the third needed

[36:20] program, YouTube Studio. This is it,

[36:22] guys. We're getting to the finish line.

[36:24] On PC, download the software from

[36:27] Stream.

[36:29] When you open the software, it will be

[36:31] blank. Double click on the screen so

[36:33] that you can open the menu. First button

[36:36] is for choosing the model. Click on the

[36:39] button and a character selection will

[36:41] appear on the bottom. The software

[36:42] already has a default model for you to

[36:44] test out. If you click on any of them,

[36:46] it will show that model up. To put in

[36:49] your own model, click import your own

[36:51] model and open folder. Now in the

[36:55] folders as open, copy paste the stuff

[36:58] that we exported from live 2D wherever

[37:01] you saved it.

[37:04] Now go back to the solver

[37:06] and click unload model to refresh.

[37:10] Clicking the model button again and

[37:12] scrolling through there should be a new

[37:14] model and choosing it. Our model will

[37:17] appear. There will be a popup to s auto

[37:19] setup. Okay. Now your model is in the

[37:23] server. If it's not moving, that would

[37:26] be because you have not connected a

[37:28] webcam or your iPhone yet. If you're

[37:30] using a webcam, first connect your

[37:32] webcam. Then click on the setting here.

[37:35] On the upper left, choose the webcam

[37:37] button. Then on the right, choose the

[37:40] blue choose camera button. Choose your

[37:43] camera and just click okay to whatever

[37:45] pops up. After that, it's done. Click

[37:48] the blue camera on button. It will

[37:51] calibrate your face.

[37:54] [Music]

[37:55] Close the calibrate. Now you have a

[37:58] working V2 model. And if you're

[38:01] connecting through an iPhone instead,

[38:03] first download V2 Studio on your iPhone.

[38:06] Go to setting and head down. You'll see

[38:09] connect through Wi-Fi or connect through

[38:11] wires. This is your preference. I picked

[38:14] Wi-Fi so that I can still charge my

[38:16] phone while streaming. Have your

[38:18] computer connect to the same Wi-Fi as

[38:20] your iPhone and it will automatically

[38:22] detect. Turn the button on and off if it

[38:25] doesn't work. Now, when you move or make

[38:27] facial expression while looking at your

[38:29] phone, the software on your PC will also

[38:31] detect that and follow along. Now that

[38:34] you connected it, it worked. There's

[38:36] still a little more things like changing

[38:38] the icon. To change the icon, you have

[38:41] to put an image into the same folder

[38:43] that contain your V2 file.

[38:48] Then back on V2 Studio, go to settings

[38:51] and on top click on it. Click on the

[38:54] image you put in the folder. Now the

[38:56] icon will be updated.

[39:00] Moving on to expression. This is how you

[39:03] can add a trigger if you have an extra

[39:05] object in your model. If you don't, then

[39:07] you can ignore this part. On the setting

[39:11] menu, you'll see this clap or button.

[39:14] Clicking on it will lead to the

[39:15] expression editor. Click on the blue

[39:18] button called expression editor. Then

[39:20] click create new expression.

[39:25] I have a mic on and off parameter for my

[39:28] own model. So I'll name it mic. Now then

[39:30] I scrolled all the way down where that

[39:32] parameter is. When we turned on the

[39:35] switch, it lets us adjust what that

[39:38] parameter will be stuck as. So I turn

[39:40] the slider all the way up and click

[39:43] save. Now going back, we have to add the

[39:46] hot key. So to tell the software how to

[39:48] trigger this expression, click on the

[39:50] plus button and a new hotkey panel will

[39:53] appear. Name it.

[39:56] Then click on the blue hotkey action

[39:58] button. Select set and unset expression.

[40:03] And then on the blue button under

[40:05] expression, select your expression.

[40:09] Key combination is if you choose a

[40:11] button on your keyboard, it will trigger

[40:14] the expression on and off. Then below is

[40:16] a screen button hotkey, which is what

[40:18] we'll be using. Click on the slot you

[40:20] want. Then make sure the switch on top

[40:23] of the hotkey setting for use screen

[40:25] button hotkeys is turned on. And now the

[40:29] screen there will be a button that you

[40:31] can click to trigger this expression on

[40:33] and off. Finally, how to get transparent

[40:36] background. First, on the menu, click

[40:39] the background button. On the list of

[40:41] backgrounds, scroll down until you see

[40:43] color picker. Select it. As you can see,

[40:48] under the color picker, there is a

[40:49] switch to turn on transparent in

[40:52] capture. What this does if if you turn

[40:54] it on when you connect this to OBS or

[40:56] other recording software you won't see

[40:58] the background only the model which move

[41:01] us over to the next and the last

[41:02] software OBS open the software that will

[41:06] be this window scenes if you can't see

[41:10] this window on the menu above docs is

[41:12] where all the window is kept at first

[41:16] add a scene by clicking on the plus

[41:18] button name it whatever you want it to

[41:20] be I'll name it YouTube scene

[41:23] Now in the next window called sources,

[41:25] click the plus button and select game

[41:27] capture.

[41:30] Click on create new and name it YouTube

[41:33] source.

[41:34] [Music]

[41:36] Now in the property here, change the

[41:38] mode to capture specific window. Then in

[41:41] the window, select YouTube studio. Under

[41:44] the check box, turn on allow

[41:46] transparency and turn off capture

[41:48] cursor. And there we go. The model

[41:51] should be shown in OBS. If you want to

[41:54] add a display behind the model, click

[41:56] the plus button again and select display

[41:59] capture.

[42:03] And now it will appear in the scene. You

[42:06] can click and drag the source up and

[42:08] down to arrange what appears on top and

[42:10] bottom.

[42:11] Now if you want to crop out the buttons

[42:13] on the side, you can right click V2

[42:15] source and select filter. Click plus

[42:18] button and select crop and pad.

[42:24] Now on the right, you can enter how much

[42:26] pixel you want to crop out to look nice

[42:28] and clean.

[42:33] And voila, it works. Now you can stream.

[42:36] Test it out just by recording to

[42:38] yourself first. Hit that record button

[42:40] and move around.

[42:43] [Music]

[42:49] Once you're done, stop recording and you

[42:51] can check where the file your recorded

[42:53] file is in the setting.

[42:57] [Music]

[42:59] Go to output and you'll see where the

[43:01] location is linked to. You can change

[43:04] the location to where you want it to be

[43:06] as well. And then you're actually done.

[43:09] That was a bunch of information that I

[43:12] wouldn't blame you for taking a break or

[43:14] even giving up. Now with this hard

[43:17] earned knowledge, what else can Live 2D

[43:20] use for? It can be used for animated

[43:23] movies. You can go farther with the

[43:25] animation feature and export as video

[43:28] instead. Make it like looping

[43:30] illustration for waiting screens or

[43:33] animating stickers and stuff. As for how

[43:36] to use the animation, here's a quick

[43:38] tutorial by going back to live 2D cubism

[43:42] in files, new animation.

[43:45] SDK Unity is what we use for V2 Studio.

[43:49] And just drag and drop the model you

[43:52] have into this window over here. And

[43:56] then drag it all the way down here.

[43:59] And now the model will be shown on the

[44:02] screen and all the parameters going to

[44:06] be here. And if you know how to edit

[44:08] videos, it's the same thing where you

[44:11] got the timeline and then you can adjust

[44:14] the parameter

[44:15] [Music]

[44:19] and then save the animation and export

[44:22] the file by going file export image

[44:25] video video

[44:27] [Music]

[44:28] and there we go. You've got a video now.

[44:32] Otherwise, if you want to export the

[44:34] animation to use in V Studio, you can

[44:37] click file, export for runtime, export

[44:40] motion file,

[44:43] [Music]

[44:47] drag it into your V2 model folder,

[44:52] [Music]

[44:53] and just like the expression, just add

[44:56] in the animation instead. Click on that

[44:59] plus button. hotkey action. Go to play

[45:04] animation, select animation,

[45:08] and then choose the one that we just

[45:10] paste in.

[45:12] Assign hotkey. And there we go.

[45:16] [Music]

[45:19] Thank you so much for watching the

[45:21] video. I've been planning to create this

[45:22] video for a very long time, but it takes

[45:25] some time to learn the software and make

[45:27] sure I didn't miss anything important

[45:28] myself since there are a lot to cover.

[45:31] If you want to commission me from

[45:33] drawing characters to rigging, I have a

[45:36] fiber page or you can just DM me on

[45:39] Instagram, Discord, X, etc. I hope the

[45:43] video wasn't too complicated. If there's

[45:45] any question, you can ask in the

[45:47] comments below. I stream from time to

[45:49] time on YouTube and Twitch. So, if you

[45:51] come to hang out, I'll be very happy.

[45:53] And thanks so much for watching. Hope

[45:55] you have a great day. Bye-bye.

⚡ Saved you time reading this? Transcribe any YouTube video for free — no signup needed.