TubeSum ← Transcribe a video

How I Make 2D Models for Streaming | Live2D Tutorial

0h 16m video Published Dec 27, 2024 Transcribed Jul 18, 2026 L LunaLase
Intermediate 6 min read For: Aspiring VTubers and digital artists interested in creating 2D Live2D models for streaming, with basic drawing skills.
Views
⚡ —
VPH
V/S

AI Summary

This video provides a comprehensive tutorial on creating 2D models for streaming using Live2D, covering the entire process from sketching and dividing a character into movable parts to setting up physics, animation, and motion capture. The creator shares personal tips and workflows to help beginners navigate the complexities of Live2D model creation.

[00:05]
Introduction and Model Showcase

The creator shows a finished Live2D model and promises to demonstrate how to create such a model.

[00:29]
Starting with a Sketch

Begin with a simple sketch, dividing the character into parts that will move. Use references from Pinterest and client wishes.

[00:57]
Character Division

Divide the character into hair (front, side, back), face (eyebrows, eyes with white, upper/lower lashes), mouth (inner part, tongue, teeth, lips), and body (main part, hands, decorative elements). Draw elements slightly deeper than visible part.

[01:40]
Exporting and Importing to Live2D

Save the drawing as a PSD file. Open in Live2D; blending mode masks may be lost and need to be re-added.

[01:54]
Adding Clipping Masks

Two ways: copy the ID of the mask object and paste onto the element, or select from a drop-down list. Example: adding colored highlight and soft light near face.

[02:24]
Eye Mask and Blending Modes

Add mask to eyes so they don't go beyond whites. Use blending modes like 'ti' and 'multiply' (similar to other editors).

[02:54]
Adding New Elements

Return to graphic editor, make changes, delete unrelated layers, save new PSD. In Live2D, add all layers as a new folder, then drag elements to correct positions.

[03:36]
Creating Art Mesh

Art Mesh makes 2D drawing pseudo-3D. Two options: automatic (suitable for most cases) and manual (for stretching objects while keeping line clarity).

[04:42]
Grouping Parts

Organize character parts into groups (hair, face, body). Subgroups allow independent movement. Layer order does not affect display.

[05:38]
Deformation Tools

Universal tool for dragging points to deform elements. Rotation former allows scaling and rotation. Last former smoothly changes shape.

[06:07]
Animation Basics

Animation occurs by changing parameters. Create keyframes: select parameter, set extreme values, program interpolates. Combine parameters for head turns, eyes, mouth.

[07:30]
Adding Physics

Create new parameter with range -1 to 1. Set keyframes for movement. Choose preset (e.g., short hair). Add parameters for axes, set influence (e.g., 100 for X and Z).

[08:43]
Exporting and Motion Capture

Compile textures (set size about 2x original). Export as .moc3 file. Use VTube Studio (free on Steam) for motion capture via webcam or phone app.

[10:55]
Final Touches and Additional Emotions

Change model name and icon. Add accessories like glasses: draw, add to PSD, set parameter for emotion (min 0, max 1), animate appearance, group with eyes. Export and add expression in Live2D Viewer.

[12:48]
Tips and Tricks

Separate eyes if accidentally on same parameter: remove parts from parameter, move to new parameter. Use reflect to fix positive/negative values. Copy animation with Ctrl+Shift+C/V.

The tutorial concludes with the creator offering the model for free download and encouraging viewers to join their Telegram and Boosty for more content. The video aims to help beginners overcome the initial learning curve of Live2D model creation.

Clickbait Check

90% Legit

"Title accurately describes the tutorial content; it's a genuine how-to guide for creating 2D Live2D models for streaming."

Mentioned in this Video

Tutorial Checklist

1 00:29 Draw a sketch of the character, dividing into movable parts (hair, face, body).
2 01:40 Save the drawing as a PSD file and open it in Live2D Cubism.
3 01:54 Re-add clipping masks for elements that need them (e.g., highlights, hair layers).
4 02:24 Add masks to eyes to prevent them from extending beyond the whites.
5 03:36 Create Art Mesh for each part: select all elements and use automatic generation, or manual for specific stretching.
6 04:42 Organize parts into groups and subgroups (hair, face, body) for easier animation.
7 06:07 Create keyframes for parameters (e.g., head turn, eye blink) to animate the model.
8 07:30 Add physics to hair and accessories: create a parameter with range -1 to 1, set keyframes, choose preset, and set influence.
9 08:43 Compile textures (set size 2x original) and export the model as .moc3 file.
10 09:33 Open VTube Studio, load the model, and set up motion capture via webcam or phone.

Study Flashcards (10)

What are the main parts to divide a character into for Live2D?

easy Click to reveal answer

Hair (front, side, back), face (eyebrows, eyes, mouth), and body (main part, hands, decorative elements).

00:57

How do you add a clipping mask in Live2D?

medium Click to reveal answer

Two ways: copy the ID of the mask object and paste onto the element, or select the object from a drop-down list.

01:54

What is Art Mesh used for in Live2D?

easy Click to reveal answer

It makes a 2D drawing appear pseudo-3D by creating a grid that deforms the image.

03:36

What is the standard range for physics parameters in Live2D?

medium Click to reveal answer

From -1 to 1.

07:30

How do you add a new element (e.g., glasses) to an existing Live2D model?

hard Click to reveal answer

Draw the element, add it to a new PSD file, import into Live2D as a new folder, set a parameter for emotion (min 0, max 1), animate appearance, and group with relevant parts.

11:23

What should you do if both eyes are accidentally on the same parameter?

hard Click to reveal answer

Click on the parameter, remove the parts that should remain, then move them to a new parameter using the change option.

13:03

How can you copy animation from one parameter to another?

hard Click to reveal answer

Select the position to copy, add a keyframe, then use Ctrl+Shift+C and Ctrl+Shift+V.

13:44

What is the recommended texture size when exporting a Live2D model?

medium Click to reveal answer

About twice the size of the original canvas to preserve detail quality.

08:58

How do you set up motion capture via phone in VTube Studio?

medium Click to reveal answer

Download the VTube Studio app on phone, ensure same Wi-Fi, start server on computer, enter IP and port in app, and connect.

10:12

What is the purpose of grouping parts in Live2D?

medium Click to reveal answer

To organize the model and allow independent movement of subgroups while maintaining parent-child relationships.

04:42

💡 Key Takeaways

🔧

Character Division Strategy

Provides a clear, systematic approach to dividing a character into movable parts, essential for beginners.

00:57
🔧

Art Mesh Creation

Explains the key concept of making 2D art appear 3D, a core feature of Live2D.

03:36
🔧

Physics for Hair and Accessories

Demonstrates how to add realistic movement to hair and earrings, enhancing model quality.

07:30
🔧

Motion Capture via Phone

Shows a practical, accessible method for real-time animation using a smartphone.

10:12
💡

Fixing Parameter Mistakes

Offers a solution for a common error (both eyes on same parameter), saving time and frustration.

13:03

✂️ Creator Tools: Viral Hooks

AI-generated clip ideas for Shorts based on the transcript

No viral clips found for this video, or they are still being generated.

[00:05] seen each other for a long time, I know I want to show you something, you won't be scared, right, then on the count of three, three Well, what do you say, not bad Really, let me show you

[00:17] how to create such a model I will say that this video is not exactly a tutorial, I will show how I create my model and maybe my advice will be and maybe my advice will be useful for someone, we start with a sketch,

[00:29] nothing complicated Draw a circle and finish drawing everything else But seriously, when we Which elements will have to move When I first received an order for a YouTuber, to be honest, I was confused, it was difficult to understand What

[00:44] parts to divide the character into correctly and I used guides from Pinterest and also relied on the wishes of the customer By the way, if you are interested, in my Telegram channel I will leave a guide according to which I created my first YouTuber, but a

[00:57] myself. The character is divided into hair, face and body, the hair is divided into the front part, side strands and the back part, the face includes eyebrows, eyes

[01:10] consisting of the white of the eye itself, the upper and the lower eyelashes, the mouth consists of the inner part of the tongue, teeth and lips, the body can be divided as meticulously or as I do into the main part of the hand and decorative elements. It is also important to draw

[01:25] elements a little deeper than their visible part. Since when we, for example, a hidden shelf, part of the face may become visible. And otherwise, draw as usual. Your unique style is what will make the model truly special. The

[01:40] model is ready for export, no matter what program you drew in. You need to save the file in psd format. Go to the program and open the file you just saved. We see that all the

[01:54] blending mode masks have flown off. Now we will fix all this in 2D. A clinching mask can be added in two ways. The first is to copy the ID of the object that will serve as a mask and paste it onto the element that needs a mask, and the second is to

[02:09] select the object from the drop-down list. The first way I added a colored highlight, and the second is a layer that will add hair. Soft light near the face. Now you need to add a mask to the eyes so that the eye does not go beyond the whites. There

[02:24] can be as many elements inside the mask. The only thing is that they do not contradict each other, otherwise it will turn out like in the example. We do the same with the highlights at the same time. Let's return the Shine to them in 2D, in addition to

[02:39] the standard, there are two blending modes, ti and multiplay, they work the same way as in other graphic editors. I drew the highlights separately so that I could create a beautiful overflow during movement. Here I noticed that the line near the lips is

[02:54] too light, so at the same time I will show you how to add new elements to the model. We return to the graphic editor, make the necessary changes, and when everything is ready, I delete the layers that do not relate to the elements that I am going to add

[03:08] so that they do not interfere and save a new PSD file. Open a new file in the window, select the model on which we are now working, then select add all layers as a new MS folder. From where we pull out the added elements and

[03:23] drag them where necessary. In my case, I delete the old line and replace it with a new one. Before conquering the world, there are two preparatory steps left: the first is the creation of Art Mouse, the second is the distribution of

[03:36] character parts into groups. Art Mouse is a grid that makes our 2D drawing pseudo-trides. We have two options for adding it: automatic and manual. The automatic method is suitable in most cases, but sometimes the manual one is

[03:49] simply irreplaceable. For example, if you need I strongly stretch the object while maintaining the clarity of the Line. I manually add an area around the Line and Voila, everything except it is stretched. Let's create an automatic arch. With the left

[04:02] mouse button, select all the elements in the work area and click on automatic generation of the sha from the drop-down list, select the standard the number of points created. Well, the

[04:15] standard option usually copes well. The mouse is created and now let's move copes well. The mouse is created and now let's move

[04:42] can optionally change the name of the group and set the value for the number of main and additional points by simply dragging them. If necessary, we can split this group into subgroups in the same way as we created the main

[04:54] group. I select all the front strands in a subgroup. And now I can move them without affecting other elements of the main group. But if I move the main part of the hair, the front strands will To begin with,

[05:08] I created the same groups as during drawing, divided the characters into hair, face, body, and so on. As the work progresses, if necessary, you can add additional groups and subgroups. This is a flexible process. By the way, the detail is important here. The

[05:24] order of layers in this window does not affect how the elements will be displayed on each This means that you can focus on the semantic structure and not be afraid to disrupt the visual order. Let's consider the tools available to us for

[05:38] changing the shape. It was through it that we created groups, and this is a universal tool with the widest application. You simply drag the points and the elements are deformed. Rotation former. A miniature Carousel for your elements

[05:53] also allows you to scale them, and the last former allows you to smoothly last former allows you to smoothly change the shape of the object being created. Finally, we can begin animating in Life 2D. Animation occurs

[06:07] due to changes in the parameters. A visual representation of what each of the basic parameters is responsible for will appear on your screen. It is not at all necessary to use them all, and you can also add as many new parameters

[06:20] add as many new parameters as you want.

[06:36] Now that we know what each slider is responsible for, we begin creating keyframes. Select a parameter. Select objects and set frames. There is an extreme animation value, the program will automatically draw

[06:50] some parameters between them. You can combine this is useful, for example, for head turns, useful, for example, for head turns, eyes, and mouth animation. To do this, click on the chain next to the parameter to accurately hit the keyframe,

[07:04] you need to right-click somewhere near it and don't forget to set keyframes on each element that is being edited. If you are not careful enough with this, you risk wasting a lot of time and nerves.

[07:17] In general, the animation process looks like this: set a keyframe, edit objects, set a keyframe again, and edit objects again. Nothing complicated, but long and painstaking.

[07:30] Now let's figure out the physics so that the hair sways softly with each head movement. To do this, first create a new parameter with a range from minus one to one. This is the standard value for

[07:45] physics parameters. For regular parameters, the range is most often from -30 to 30. Then, as usual, create keyframes and set the movement. When everything looks as intended, go to the

[08:03] name, and select a suitable preset. In my case, a preset for short hair will do. We add the parameters at which we want to see the physics. I set the movement along the axes. I also added another segment to the parameter so that the Earrings would

[08:19] swing more. This is not a mandatory step, but you can experiment with the parameters here. In the outputting section, we indicate where the physics will take the animation from and we see that nothing

[08:31] happens. This is because I forgot to set how much the movements along the axes will set how much the movements along the axes will affect the physics. We set 100 for the affect the physics. We set 100 for the X and Z axes. And now, finally, it worked.

[08:43] We are at the finish line. The only thing left is to export the model to the motion capture program. Before that, we need to compile the textures for this. Go to the ATL settings texture menu here. I recommend setting the texture size to about

[08:58] twice the size of the host on which you drew the model. This will help preserve the quality of the details. Please note that the textures do not overlap each other. If this happens, just [music]

[09:19] move over. You can export the exp for Run file. Er export the exp for Run file. Er AF in the window, just click o. Create a new folder, save our model in it. Go to the YouTube Studio program.

[09:33] You can download it from Steam here. Go In the settings, if we click on Open Data Folder, the folder from which the program takes resources will open. Open Life 2D modules and

[09:45] copy our previously saved folder here. Now our model is among the others. Select it and click OK. The model has appeared. But while it does not move, VOP

[09:57] Studio can capture motion through a webcam or phone. Let's start with the webcam in the settings. Select the camera, turn it on and calibrate it. And voila, it's alive. Let's try the second

[10:12] method. For this, you need to download the YouTube Studio application from Asterisk or Google. It is important that the computer and phone are connected to the same Wi-Fi network. Go to the settings and scroll down. Switch the toggle switch to start the

[10:27] server. Here we need the port and IP values ​​​​in the application on the phone. Open the settings, enter the same values ​​​​that you saw on the computer and click Find North. After a successful search,

[10:42] connect to the computer. Keep in mind that in the free version, the connection will drop every 5 minutes, but just click Connect again. Now let's just see how happy I am with the new model.

[10:55] new model. [music] the final touch I change the name of the model and put an icon for this in the folder with the model I add the desired picture and in the character settings I click on the empty

[11:08] icon and select the picture that I just added to the folder also If desired, the model can be sent to the phone, provided that the connection with the phone has already been established Congratulations, the basic course is complete But that's not all, I'll

[11:23] show you some useful points further Let's add additional emotions I'm an accessory but the principle is the same I drew glasses and added them to new psd

[11:35] drew glasses and added them to new psd files as at the beginning of the video I set the parameter for additional emotions the minimum and default value should be zero and the maximum is one I create an animation of the appearance of glasses I throw them into a group with

[11:48] eyes so as not to Riga separately and correct where necessary before exporting you need to add new textures to the map I add new textures to the map I just

[12:01] then re-save the model further We open 2D VI this program is installed together with the main one from the folder Where you just saved the model by dragging the M 3 file into the window that opens then click the file add

[12:17] Expression and Name the folder and select the emotion among the parameters, look for the one responsible for the appearance of this emotion and set it to Maxi, save the emotion and export the model with a

[12:32] replacement in YouTube Studio, open the IDM model in the settings, set what the emotion will be turned on, select Which one exactly and set a hotkey either by selecting from the list or by clicking on the red

[12:48] record button and pressing the desired button on the keyboard, check the work. Congratulations, keyboard, check the work. Congratulations, you are great at showing a few moments, if I had known from the very beginning, you would have saved me a lot of time, let's say you accidentally

[13:03] set both eyes on one parameter, they can be easily separated, click on the parameter next to the add and remove keyframe icons, add and remove keyframe icons, hover over the arrow, press sec, holding down Ctrl,

[13:17] remove all the parts that should remain on this parameter. Then return to the parameter again and click change, now select the parameter where we will move the animation [music]

[13:30] [music] now everything works correctly in 2D, it is important not to confuse positive and negative parameter values. But if suddenly this happened, it is enough to simply press reflect if you need to copy animation

[13:44] to another parameter, select the position that we want to copy, press Select, add a keyframe in this place, also add a keyframe on the parameter where we will copy and copy using Ctrl Shift C

[14:00] and Ctrl Shift V And now let's see what we got. I created this model specifically for this tutorial, so anyone can download it from the link in the description. It seems to me that Tube models should not be

[14:14] so expensive, so let this model be available to everyone model be available to everyone [music] [

[15:00] to the end, I hope this video was useful for you. I know how difficult it is to start, especially when you are faced with a bunch of new tools, so it was especially important for me to share with you everything that I learned myself so that you

[15:13] could move forward faster and more confidently. Taking this opportunity, I would like to once again invite you to my Telegram, where every day I post sketches, spoilers for future projects and exclusive shots that do not appear on other

[15:26] social networks. The channel is still small and I want there to be more of us. I will be extremely happy to welcome every new participant to conduct various activities, and for those who want to see the full process of creating an

[15:39] aventurine model, as well as the top-secret process of creating a model of the Moon, I have process of creating a model of the Moon, I have prepared exclusive content on boost. This video is coming out 5 years after the last video on this channel. I

[15:52] hope the next video will be released much sooner. Bye-bye everyone.

⚡ Saved you 0h 16m reading this? Transcribe any YouTube video for free — no signup needed.