[0:00] Hello, I’m Hanacue, your local digital 
artist. So I’ve been really interested in  
[0:05] making my own VTuber model these days, and 
I thought I would take you guys through my  
[0:10] process of making a VTuber model from scratch. 
And as an added challenge – because I sure do  
[0:16] enjoy making things harder for myself – let’s 
try to make the whole model in a week. Also,  
[0:21] this video is going to be a two-parter, with 
this first video covering the process of  
[0:25] drawing and cutting the model, and the second 
video will be about rigging the model.
[0:32] BUT first! I need to call out some people. Do 
you have an original character that you’ve been  
[0:38] neglecting to come up with a backstory for? 
Or maybe you’re one of those people that have  
[0:42] hundreds of OCs just sitting around, waiting 
to finally see the light of day? You know who  
[0:47] you are. Well, I am calling out all artists with 
OCs, whether one or a bunch, to give them a home  
[0:54] on UnVale! UnVale is a new social art sharing 
platform that emphasizes a focus on artists,  
[1:00] writers, and worldbuilders. You can 
create pages for your original characters,  
[1:04] fill in details about their backstory, add tons 
of images to their pages, and even add character  
[1:08] traits to them. I love using the pages as a 
way to write about the characters I’ve made,  
[1:12] and color coding them to their exact aesthetics. 
You can even create worlds for your characters  
[1:17] to categorize them, and there’s even a diverse 
community that you can socialize with on your  
[1:21] pages! I hope you can’t tell how excited I am 
about this site, because I love OCs with all my  
[1:27] heart. I’ve even added a bunch of story tidbits 
to some fan favorites like Sitrina, Enceladus,  
[1:32] Zori and Gin, if you want to go ahead and check 
them out on my page below! I love art websites  
[1:37] like this because they’re vast yet tight knit, so 
it really feels like you’re part of a community.  
[1:42] It’s time to pull your OCs out from the cupboards 
and really give them a proper home and showcase  
[1:47] them on UnVale! If you’re interested, please feel 
free to check out my link in the description. And,  
[1:52] if you do create a profile, make sure to tag 
your characters with the hashtag #hanasquad,  
[1:57] so we can all find each other. And so I can 
gush over your adorable characters. Thank you  
[2:02] to UnVale for sponsoring today’s video. 
Now, let’s head back to the video!
[2:08] So some of you might be familiar with my past 
VTuber work on the channel. I actually started  
[2:13] my YouTube channel to share my findings on 
making VTubers as an artist. I came a long way  
[2:18] since then, and have shared hundreds of short form 
videos on five models I had a part in making from  
[2:23] 2021 to 2022. The first three models I made from 
scratch, meaning I did everything from design,  
[2:29] art, to rigging/animating, were the iconic 
wolf boy, Gloomy, wolf girl and cat duo,  
[2:35] Rhi’a and Kurone, and of course the purple fox 
boy, Crazyfox. I’ve also done rigging work,  
[2:40] which means all I did was make the model move, 
for pink magical girl, Squeakychiona (who I will  
[2:45] also be making a model from scratch for later 
this year), and white cat girl, Starsmitten.
[2:51] Keen observers might notice that all of these 
models were made for other people. Despite having  
[2:56] my own persona for years, I have never attempted 
to make a VTuber model for myself. And the reason  
[3:01] is, well, I’ve just never felt compelled to. 
I just love making work for other people,  
[3:06] so I simply never found the time to. But 
of course, during my occasional streams,  
[3:11] this was a point of curiosity for people. It 
was ironic that as a VTuber maker, I, myself,  
[3:18] never made myself a VTuber to use. But today, that 
all changes. I am going to make myself a VTuber  
[3:25] and I will finally become a VTuber. 
Before we go ahead, I do have to put out a  
[3:30] disclaimer that this is not a tutorial video. 
Consider it more of a general process video,  
[3:36] where I’ll loosely explain what I’m doing and how 
I’m doing it. And if you learn something from it,  
[3:41] that’s great! Just please don’t expect a full 
fledged tutorial, because really it’s more of  
[3:45] a join me along as I make something kind 
of cool and intensive. Alright,  
[3:49] nice. Let’s move on ahead to the first step.
[3:56] For the VTuber’s design, I obviously have my  
[3:59] current persona that I use in all my videos, 
which is the green-haired girl with braids and  
[4:03] a little white dress with a fuzzy cardigan. This 
design came about after many iterations of past  
[4:08] personas I’ve had. Before green braids, I was a 
generic brown-haired twin tail girl in a similar  
[4:14] little white dress. This version actually had its 
own 3D VTuber model during my Twitch days that  
[4:19] a friend got for me, which was pretty cool for a 
while. Before that, I was a pink haired girl with  
[4:23] braids and a simple get up. And even before all 
that, my past identities were… well… that’s all  
[4:31] classified for another video, hehe. 
I really enjoy my current design. It’s  
[4:35] unabashedly girly, with all my favorite colors, 
and things like green, pink, pearls, ribbons,  
[4:41] and more. And most of all, I love all the fanart 
you guys have made of her, which has made me come  
[4:46] to love her even more. But that also makes me 
feel a little guilty to say that I am changing  
[4:52] the design just a little bit. In my first iteration of the design,  
[4:55] I loved everything but felt like some parts were a 
bit redundant or unclear. So for day 1, I created  
[5:01] a new sketch to simplify a lot of parts, and make 
sure the shape language was clear. Stay with me  
[5:06] here, I promise the final VTuber is going to be 
just as unabashedly detailed and pretty as before,  
[5:12] but sometimes you need to reduce things to their 
simple forms before you can add more details.
[5:17] The parts that were important for me to clarify 
were the messy bob hair that was tied into  
[5:22] heart-shaped braids, and the overall silhouette 
of the character. In fact, this entire design uses  
[5:28] heart shapes everywhere. The top is heart shaped, 
the ends of the hair split into hearts, and the  
[5:33] silhouette is also meant to reflect a vague upside 
down heart. To make sure the silhouette in general  
[5:38] was clear, I had to emphasize the most important 
design elements. Messy bob, braids, fuzzy jacket,  
[5:44] short skirt, and long legs. Basically, I want this 
character to read like a doll, because the goal is  
[5:49] everything girly, cute, and doll-like.
As you can tell, I removed a lot of details  
[5:54] from the original designs. Some of them will 
make their way back, but others were redundant. 
[6:00] For example, these arm sleeves, and arm 
accessories weren’t really needed because I intend  
[6:05] to have the character wearing the large fuzzy 
cardigan at all times. I also clarified shapes  
[6:09] and the logic of design choices like the placement 
of the stitching on her dress and opting for a  
[6:14] thicker ribbon. This is very important for later 
when we work on drawing the actual model.
[6:19] I’m pretty happy with the clarified design, and I 
consider this more a bare bones design. If I had  
[6:24] to only keep the essentials, what would I keep? 
So this is what I came up with for my VTuber  
[6:28] design! Similar to before, but a little clearer. 
However, it has just now occurred to me that maybe  
[6:34] some of you don’t know what a VTuber is.
[6:39] So what is a VTuber? I’m going to explain to  
[6:43] you what VTubing is to me, from strictly an 
artist perspective. A VTuber is a digital  
[6:49] avatar that you use to represent yourself, which 
simulates your movements in real time using a web  
[6:53] camera. Ever seen mocap actors and how that then 
translates into games? It’s like that, but live.  
[6:59] The term VTuber is short for Virtual YouTuber, 
but this term now encompasses a large range of  
[7:05] content creators outside of just YouTube alone.
VTubing found some of its first introductions on  
[7:09] YouTube with popular 3D VTubers like Kizuna 
AI. But it reached the real height of its  
[7:15] popularity during the pandemic lockdown, when 
most people had to be inside and thus turned to  
[7:19] online entertainment. During this time, with the 
introduction of VTuber corporations and the growth  
[7:24] of streaming in general, VTubers became extremely 
popular. In my opinion, the VTuber community also  
[7:30] grew quickly due to its intersection with other 
communities like anime, gaming, streaming, art,  
[7:37] voice acting, singing, and many others!
Now, you may have seen some VTubers that are 3D,  
[7:42] which are avatars you can sometimes use in apps 
like VRChat. We often refer to these as models,  
[7:47] or 3D models. However, a new type of VTuber 
model would begin to rise during this time as  
[7:53] well. These were called Live2D models, which were 
a type of model that used flat 2D art, and rigging  
[7:59] and morphing technology in a program called Live2D 
Cubism. This type of rigging is an animation style  
[8:05] that is popular in newer animation shows, because 
it utilizes the ability to keep a character on  
[8:10] model and adding what we call a skeleton under, 
to move body parts and change facial expressions  
[8:16] instead of redrawing them over and over.
This by no means makes it easier, it just has  
[8:21] different benefits that a show that uses the same 
characters and similar shots may benefit from.  
[8:26] But this technology became useful as the idea of 
creating 2D VTuber models was popularized. You  
[8:32] could essentially draw a flat 2D image, and morph 
the different angles to give it the illusion that  
[8:38] it was turning left and right, or up and down. 
This opened up many avenues for artists who were  
[8:42] skilled at art, but not so much at 3D modeling, a 
chance at creating their own VTuber models. And,  
[8:48] ding, ding, you guessed it. I’m one of 
those artists who jumped on the train!
[8:52] Now there are many benefits to making a Live2D 
model. 1: Live2D models have that organic beauty  
[8:59] of being a drawn image that is different from a 
3D model. Of course it comes down to preference,  
[9:03] but it is a different appeal. 2: You are able to 
break more things and hide things in 2D art than  
[9:09] you can in a 3D model. 3D models tend to have 
the need to be logically sound because you are  
[9:14] unable to 100% control the way the model operates. 
But in a 2D model, you ARE able to control every  
[9:20] aspect of the model to appear only how you like. 
3: This is just my opinion, but I think Live2D  
[9:27] models are easier to get into because most of us 
will have general skills in art, but not many of  
[9:32] us will have learned 3D modeling or sculpting 
skills. Now, I say easier to get into, but this  
[9:38] does not mean a Live2D model is easier to make. 
It is still a lot of hard work, but essentially,  
[9:44] doable for anyone who has some art skills.
Of course, the program in question to make this  
[9:49] all happen is Live2D Cubism. I’ll talk 
to you more about that in the next video,  
[9:53] but as I am currently aware, Live2D Cubism 
is the only program as of right now that  
[9:58] you can make a 2D model in.
So now you know what a VTuber is  
[10:01] and the different kinds there are, now let’s 
move on to start sketching our model!
[10:10] When it comes to sketching and planning the model, 
there are a few things to keep in mind. Firstly,  
[10:14] and this is not a hard rule, but it will make 
your life a lot easier if you keep the model  
[10:19] symmetrical on the left and right side. It’s 
easier to rig the model when your two sides  
[10:23] are approximately the same because you can 
just repeat what you do on one side to the  
[10:27] other. So I’ll just draw the body and head 
using the symmetry ruler tool in Clip Studio  
[10:32] Paint. Most art programs should have this tool. 
For parts like the hair that are asymmetrical,  
[10:37] I’ll create a guide for placement, then draw 
in the hair. Secondly, you’ll want to draw your  
[10:41] model in a program that can save to a .psd file, 
has layers, and also some clipping functionality.  
[10:47] Again, Clip Studio Paint works for me so I’ll 
be using that. Thirdly, the way you’ll want to  
[10:51] think about 2D models is as if they’re paper dolls 
you can dress up. Our paper dolls will be quite  
[10:56] advanced as they will have joints that allow you 
to change their pose or position. So you’ll want  
[11:00] to imagine that you’ll be building the dolls up 
layer by layer. This is hard in the beginning,  
[11:05] but gets easier as you do it more. 
For the second half of day one and for day two,  
[11:09] I’m going to sketch and plan the model. As 
I’m working on my model, I’m really aiming  
[11:13] for a doll type of pose, because I personally 
really like that aesthetic. I also plan to have  
[11:18] all the body parts completely detached from the 
clothes, so that I can draw more outfits for my  
[11:22] character in the future if I want.
One defining feature of VTuber models,  
[11:26] other than the bouncy boobas and jiggle physics, 
is the addition of many types of accessories.  
[11:31] This is definitely on purpose, because the more 
interesting accessories you have on your model,  
[11:36] the more physics you can add later on. For 
example, adding ribbons can allow you to create  
[11:40] a lot of dynamic physics when the character 
moves. Adding things like earrings, will add  
[11:44] a lot more interest around the hair and face 
movement. Having strands of hair in many areas  
[11:48] will also help create dynamic movements. So while 
these design elements may seem random at times,  
[11:53] a lot of VTuber artists will add them because 
they help create the illusion that the model  
[11:57] is more advanced and dynamic, with more physics 
and movements. This is of course up to personal  
[12:02] preference, but it’s certainly something I try 
to keep in mind when making a VTuber design.
 
[12:06] To continue my planning, I add colors to get an 
idea of how things will look in the final model  
[12:11] before I jump straight into lining and rendering. 
This is extremely helpful because sometimes you  
[12:16] aren’t able to tell if the model is unbalanced 
in color or line distribution until the colors  
[12:21] are on, so I really recommend you create this 
color plan first. The colors can be simple,  
[12:25] just enough to give you an idea how it 
will look in the final artwork.
 
[12:28] Now, it’s time to plan some poses for my 
character, which in my case are just going  
[12:32] to be different arm attachments. I wanted to 
create a pose for all my existing avatar images,  
[12:37] so it didn’t feel like anything was lost when I 
begin using the model. I really don’t recommend  
[12:42] you add too many arm poses if you’re a beginner 
because truthfully it is extremely time consuming  
[12:48] and hectic. You will see later how much 
I regret this decision. I will also take  
[12:53] the time to explore different facial expressions 
and mouths to imagine how the model will look in  
[12:57] rigging later on. And that wraps up everything 
I wanted to do for days one and two!
[13:05] Hello, hello, it’s day 3 and I’m making sure to be 
very caffeinated for this process because today we  
[13:11] will be lining the model. Sketching the model is 
kind of like the appetizer to the process, but  
[13:15] you really get a full taste of what it feels like 
to be a VTuber mama once you get to the lining.  
[13:20] By the way, VTuber mama and papa are the terms 
of endearment that we use for people who make  
[13:24] VTuber models. Pretty cute, right? Remember how I 
said earlier that you want to think of the models  
[13:29] like paper dolls, where you build up pieces? 
Well, I really mean it. I’m going to line  
[13:33] every part separately, and what needs to be 
separated is really up to your intuition. I  
[13:38] like to keep a lot of parts separate though.
So here I am outlining the whole head, then a new  
[13:42] layer, and do the ears. If you’re a beginner, 
I recommend you keep this simple and intuitive  
[13:46] for yourself. You can always go back and do more 
cutting in the future if needed, but you certainly  
[13:51] don’t need to be excessive. However, since this is 
my sixth model, I’ll be doing some more advanced  
[13:56] techniques in my lining and cutting process. 
For the lines, this includes some layer parts  
[14:00] having an outside line and an inner line, so I can 
create more 3D looking rigging in the future.
[14:05] Lining is easy to explain, but a lot of work to 
do. We’re making sure to line to get a clean idea  
[14:10] of each layer part of the model. You’ll also 
want to imagine and overline areas like the  
[14:14] neck for example. We make the neck extremely long 
because the head turning in rigging requires a lot  
[14:19] of area to make it look natural. I also made my 
layers somewhat visible here so you guys can get  
[14:24] an idea of my workflow. As a reminder, this is not 
a tutorial, just my process, so please don’t feel  
[14:30] like you have to do exactly what I’m doing. The 
amount of layers is pretty intimidating and yes,  
[14:35] it is hard to keep track of. But to make VTuber 
models, we will need to get used to it. I  
[14:40] recommend naming your layers when you can and even 
color coding them with this button in Clip Studio  
[14:45] to help your eyes with the process. I’m making 
sure to draw the bare body, then the clothing  
[14:49] items on top. And I’m using the symmetry tool 
again for a lot of this to save myself as much  
[14:54] work as possible. I like to line my work even if I 
intend to paint or render the model in the future  
[14:59] because it also helps create clean bases which 
is important for our VTuber model to work well.
 
[15:04] Here on out is just me lining the model, so 
I’ll leave you guys to a short timelapse.
[15:34] So that entire lining process has taken me 
about 3 to 4 hours now, and that’s a lot of  
[15:39] drawing in a day so I’m going to return back to 
you guys with some base coloring tomorrow.
[15:47] Base coloring is a pretty essential process 
to making 2D Vtuber models because it ensures  
[15:52] that no blank spaces or weird pixels are out of 
place. So for day 4, I’m going to be adding the  
[15:57] base colors and also fixing up some lines. The 
thing to consider when it comes to adding colors  
[16:01] is understanding the layer hierarchy - that is, 
which layer goes belocw or above another layer,  
[16:06] so that the layer parts show in their appropriate 
order. This is part of what we call cutting  
[16:10] in the 2D model art. The second part to cutting 
is making sure that all the layer parts have their  
[16:14] corresponding “back layers” as needed. For 
example, for the cardigan, it is a bit weird  
[16:19] to have it as just a floating part with no back 
part to the cardigan, so I will draw that in and  
[16:24] make sure the layer order is to the back behind 
everything. You will need to use your intuition  
[16:28] here for this, but typically parts that may need 
this are sleeves, and undersides of clothing.
[16:33] I’ll just continue coloring and organizing 
all the layers in a timelapse now,  
[16:37] so I’ll see you guys in a bit!
[16:53] We’re dipping into day 5 here, as I’m  
[16:55] working on lining and coloring all those arm poses 
that we talked about earlier. This requires a lot  
[17:00] of redrawing and imagining how the arms will work 
in motion and with the layer hierarchy. And it’s  
[17:05] about 8 different arm and hand sets I’m drawing, 
so wish me luck. This entire process was actually  
[17:11] extremely painful, and took a lot of brain power. 
In the end, it took the whole day and a whopping  
[17:16] 4 hours which is the same amount of time I took 
to line the entire model itself already. And I’m  
[17:22] actually not sure if it will be worth it, because 
most people have enough trouble remembering to  
[17:27] press toggles to change poses to begin with. So 
if you guys take anything from this video, learn  
[17:32] from my mistakes, and maybe skip the alternate arm 
poses or save them for a later upgrade after you  
[17:38] finish the model because it is really draining. 
I’m going to get a lot of rest before this next  
[17:42] part because I am gonna need the energy to 
power through the rest of this model.
[17:50] Ah, it’s finally time to do some rendering. I 
mentioned earlier that one of the best features  
[17:54] of a Live2D model is the fact that you get an 
organically drawn look to your model. And there  
[17:59] is no better time to sell that effect than in 
the rendering part. So the way I’m doing this  
[18:03] is that I’m going to add some detailed coloring 
on a different layer clipped to the base layer.  
[18:07] This is the time to make it as detailed as you 
like, and for me, I like to take the opportunity  
[18:12] to put extra care because people will be looking 
at these models for long hours on end. So I like  
[18:17] to try and make the work stunning for people 
to enjoy and find new things each time!
[18:21] I want to mention that the traditional way to 
make Live2D models is to merge parts down so  
[18:25] it’s the base color, line, and rendering all 
in one layer. This is a really convenient way  
[18:30] to simply rig the art later. And I recommend this 
method especially if you’re planning to work with  
[18:34] another Live2D rigger. However, since I plan to 
rig this model myself, I have a vision where I  
[18:39] want to keep the lines, base color, and detail 
coloring all in separate layers. I’m doing this  
[18:44] so I can create a more 3D look in my model as I 
can manipulate the parts at different speeds. An  
[18:49] additional way of looking at this is that, I’m 
essentially creating a wrap for my parts using  
[18:53] the detailed coloring where I’m overdrawing the 
shading in a way that makes it flat. This way  
[18:57] when the layer is moved to the right or left, it 
will have the extra art to create a more 3D look.  
[19:02] It’s a similar concept to wrapping texture in 3D 
modeling. Again, this is not necessary at all,  
[19:08] and you can simulate a 3D look with one layer, 
but I’m an experimental person, and I really  
[19:12] wanna see what kind of look I can achieve with 
this additional effort. There will be a lot more  
[19:16] layers, but hopefully the result is worth it.
I wanted to include my whole coloring process  
[19:20] here, but as I explained earlier, I drew the bare 
model so I can swap out her clothes in the future.  
[19:25] Which means I was coloring some skin areas that 
might not be completely appropriate for YouTube,  
[19:30] so sadly I didn’t include those parts! But if 
you guys are interested in the skin coloring,  
[19:35] I may release part of it elsewhere if you like!
Now, I’m going to continue my process of rendering  
[19:39] and overcoloring all the way until we’ve 
reached the completion of the coloring. This  
[19:43] was the bulk of the work I did for days 6 and 
7. In the meantime, enjoy the timelapse!
[20:16] Alright, so I may have grossly overestimated how 
much I could do in a week. I was planning to do  
[20:23] some cutting as well this week, but somehow 
I overlooked how challenging it would be to  
[20:28] draw an entire model, try and plan and draw an 
animation to go with the reveal because yes,  
[20:33] I’m crazy, and also edit the entire video in 
a span of one week. As it turns out, I am only  
[20:40] human and for some reason can’t spontaneously 
grow four additional arms to draw, animate, and  
[20:45] edit at the same time. So I tried to finish the 
coloring and rendering to the best of my ability,  
[20:50] but ultimately I think the art is only about 75% 
done, at the moment. So I’m going to be kind to  
[20:55] myself and take it easy, and save the rest of the 
work for the next video where we’ll tackle all the  
[21:00] goodies like cutting the model, finishing the arms 
and emotes, setting up the model in Live2D Cubism,  
[21:05] and rigging and animating the model. I am dead 
tired as I am recording this, and while the art  
[21:10] isn’t yet finished, I feel like there’s enough to 
give you all a little bit of teaser in the final  
[21:14] reveal. So please do enjoy the final reveal 
of all my hard work and effort so far!
[22:35] Thank you so much for watching! I’ve always 
been a little bit unsure how to make a video  
[22:40] on the process of making a VTuber model because 
it is really complex with lots of steps. It’s  
[22:45] definitely a lot easier to capture just my process 
of making a single artwork. So I thought I might  
[22:49] do it in a little bit of an informative process 
vlogging type of style. It’s different from my  
[22:54] usual type of video, but I hope you can find the 
process enjoyable to watch! Again, not a tutorial,  
[22:59] because I know this entire thing can be really 
overwhelming. My main goal is to just expose  
[23:04] you guys to the process and familiarize you 
with it and the terms. Hopefully it helps a  
[23:08] little! And if not, I’ll try my best to get 
a little better next time! As always, here is  
[23:13] a showcase of all of the lovely fanart you guys 
have made for me lately. I love them very much,  
[23:17] and I hope this new model can live up to the 
beauty you guys always capture me in! Until  
[23:22] next time, I’m wishing you all the best on your 
journeys of achieving your goals and dreams. Bye!