Arcane came out last month and, me being the
old man that I am, didn't know about it until people started asking me to make a video about
it! So let's talk about how to paint characters like the show; concepts that make the show look
good in general; and how to train ourselves to make concept art for shows like this! This video
is in partnership with Acer's ConceptD: a new brand for creator PCs. I'll be using the ConceptD 7 Ezel to make all the art in this video. And I'll show you some of my favorite highlights of this
computer along the way. Let's talk Arcane! Arcane is a fairly painterly looking show, so let's see if
we can decipher some of that style. Here's a screengrab I took, which I think will work as an overall
case-study. I'm making a new layer here, filling it with white, then decreasing its opacity. Basically
I've created a sheet of tracing paper. Yes - for this study i am going to trace the character! I want to
study - exclusively - the painterly style of this show. I want my drawing to be as exact as possible,
so I can get right down to the shapes of value, color and texture, knowing that if I get any of
that wrong, I can't blame the drawing. So in this case tracing is a strategic decision. On a layer
underneath the lines I'll fill in the base colors then on a new layer set to 'multiply' mode, I'll
start finding the shadow shapes. It's important to note that I'm not at the style stage yet! I'm
just establishing the painting fundamentals that I would need for any painting. In this case,
a single average light value, responsible for all the light, and a single average shadow
value, responsible for anything in shadow. The style of the shapes in Arcane is quite graphic,
so I like to use the lasso tool to determine where my shadow shape is, then I can fill it in almost
like a stencil. On the ConceptD, I have the easel feature tilted halfway down. That's so i can
access shortcut keys. I was worried that the screen would lack stability floating in the air
like that...but it feels great! Here I'm starting to add some complexity to the skin's color with
a soft brush set to 'color' mode, and just glazing in some of these more purple-y hues. Notice that
I'm doing this on the original skin color layer. That way my shadow layer, which is set to 'multiply,'
will update its colors automatically. Okay it's time to start tackling one of the main staples of
Arcane's style. At least with the characters, the show does not shy away from having the appearance of
lines in the renders. But the lines are not around the character's silhouette like in a drawing,
they're inside the form - often defining major plane changes or anatomy. So for example there's
a line here defining the bottom of the eye socket, or the 'bags under the eyes,' as we commonly know
them. The line defines the contour of it, but then there's a subtle bit of shading here defining the
form of it. Arcane is not the first production to do that ... but it remains uncommon to see this in CG
shows. As a painter, I appreciate the design choice! So that's what I'm concerning myself with at
this stage of the process: finding where the lines go. Now, I can't just use my tracing for this
because my tracing was rough. But the flow of the lines on these characters is exact. It's almost
like calligraphy: it announces itself. So I have to slow down and find it. And now I'm putting on my
'RossDraws' hat, and on a layer set to 'linear dodge' mode I'm throwing in that dramatic rim light on
the left. Whether it be a shadow shape like this, or a rim light shape like that, the shape is really
what I'm in service of. Notice how those two shapes, despite one being light and one being shadow,
they're connected thematically. They look alike! The way your shapes look plays a huge part in
delivering style! Even a non-artist who may not know this stuff intellectually ... like we do ... they
will be able to sense it when it's there. They'll also be able to sense it when it's NOT there. Okay
here's another interesting part of Arcane's style. That whole calligraphy thing I just talked about?
It's also echoed in the light shapes. So here on a new 'linear dodge' layer, I'm stenciling out some
shapes and with a textured brush, laying them in pressing lightly on the stylus though to implement
them in a semi-transparent way. I should probably point out now that when I laid in that original
color, I did it with a medium value - knowing that at this stage I would be going lighter. The light
shapes in Arcane are a big piece of connective tissue of style. Check out this tweet I found. Most
of the characters have this particular shape on them. That shape is true to life! It defines another
piece of the orbital that often does catch a highlight. But it's not required to be there! Arcane,
however, uses it throughout the show as a way of connecting the characters stylistically. Again a
strong choice, in my opinion. I just want to show you some of the calligraphic shapes and rhythms
that i'm tracking here. For example, this one: a flowy connected shadow shape. A very similar
wormy shape connects the eyebrows, but then we get some contrast with this rhythm i've already shown
you. Stylistically, there's something pleasing about these rhythms: the first two rhyme and the third
one breaks. Now buckle up and look at this one: there is a rhythm that connects the cheekbones
to the jowls and back up the opposing cheekbones! Now watch as that dotted line goes away...it's a
subtle one, but functionally, it ties together the entire design! All right, so here's my finished style
study. I think we made a few good insights there. Now I want to show you how to study readability.
Readability is at the core of what makes a picture understandable to us. It concerns value structure,
and there are a lot of values, so this is difficult for a lot of people. So watch this. I'm going
to take this section of the frame, and I'll start with the background. The background is on the
dark side, somewhere around here on the value scale. Knowing that I have so much room above that on
the value scale, I'll block in this ball, which is a stand-in for the character. And for that I'll
use a value here. Now, take my word for it, that this is a dramatic value difference! Even though
it may not look like it on the value scale, the eye has no trouble picking that up. So what
our study represents at this stage is if our frame from Arcane looked like this. You know, simple
yet strategic one-value decisions for each element. This puts us in the right ballpark right away
with contrast. Okay but now we have to think about the light on our character. I have a
lot of room above me on that value scale. The character is already popping from the
background, but I can make it pop even more with dramatic light. So I've jumped to a value up here.
Our little ball there is crystal clear; it's an instant read. But there's even something more I can
do. Where the character casts the deepest shadow, that value can be way down here, but now I have to
be careful. These two values both represent shadow ... but they are very different. Shadows in real life
tend to group together more than that. Now, this is art! I can change nature if I want to! And I *do*
want to, for the sake of clear and fast readability. So what I'm doing here is trying to bridge
the gap a little bit while staying true to my overall contrast strategy. And I think that works.
One thing I could do to group the shadows more - and this is happening in the Arcane frame - is merge
the character's cast shadow with some of the background elements. 'Merge' just means bring those
values closer together. And you notice I only did that where the shadows are, not where the light
is. And that's the story of that part of the frame! All right but there's a whole other part of
this frame that reads completely differently! This character here. This time the background is
lighter, around here on the value scale, and she is darker against it: somewhere around here. She
still has light though, and that light is still lighter than the background - that hasn't changed.
Though the contrast between the two is less. Less ... but still plenty punchy. We're
looking at a value distribution like this. Now, remember what I said about shadows earlier?
Look how merged those shadows are. So I need to adjust my study to accommodate that, playing with
shadow values in this range, blending freely from character to background, though making sure I
don't totally lose the character silhouette. Of course, the light retains a perfect silhouette.
On an emotional design level, this character is sad. It makes some kind of poetic sense for shadows
to share each other's boundaries like that. So I'll continue that aesthetic but I've dropped the
values down here for a more full range of contrast. I didn't have to do that but that's what I see
in the reference. And here's our two readability studies, from that one frame! Two different
characters, two different approaches, both very readable. All right, let's try the same thing for this
frame. It looks immediately less subtle but there are still some subtleties in there. The background
eats up the most value range, say, like this, and the character in its entirety is darker than that.
Now, there is some ambient light on the character, but it's very subtle, just a half step up maybe.
Then the character has some ambient occlusion... which is different than shadow - I have a whole
video about that ... but essentially it bleeds right into the character. So I'm sticking with this value
range. Because this is ambient occlusion, however, I'm much more willing to lose the silhouette
there, because that's how ambient occlusion behaves. And then lastly, because I have so much value room
above the ball I'm just adding some light accents which in the reference is inspired by that puff
of smoke. And I just want to say that the ConceptD screen is a joy to work on! It reproduces values
and colors reliably, which is something I cannot say for many monitors, laptop or otherwise. All right!
So, maybe you want to work on shows like Arcane, doing backgrounds and concept art. If that's the
case I really recommend you learn some basic 3D. Blender is free and does everything we need. Upon
opening Blender you'll see the famous default cube, and a camera, and a light. You will probably also
see a menu full of stuff you probably don't understand. And yes, Blender can be intimidating! But,
as painters, we only need to know how to do a few basic things. We need to be able to make an object,
scale an object, rotate it around, move it around, and maybe even take little sections of that
object and further model that. But that's it! It's basic ... and here's how to do it! The first thing
you'll want to know is how to navigate your view. Use the middle mouse button to orbit around,
shift + middle mouse button to move left and right, control + middle mouse button will pull in and
out, or you can scroll your mouse wheel for that. By the way you will want to use a mouse with
Blender. So this is the default perspective view but if you press the tilde (~) key, you can change
your view to, say, the top view. I can push tilde again and switch it to the right view, and you
get the idea. If I orbit the camera now it'll flip back into perspective view so that's something
to keep in mind. It's helpful to have different windows that have different views. To do that, click
way up here in this corner drag out a new view. Then in that view you can switch to top view or
whatever you like. If I want to get rid of that top view, I go back to the same corner click and drag
until I see this arrow, and just let go. Objects are selectable simply by clicking on them. You can
also see what's selected in the outliner here. To move an object click on the move icon right
there and you'll see these three axes: x y and z pop up, I can just simply click on one and move it
in the desired way. If you click on the circle it's freeform move. The shortcut for that is 'g' which
I use all the time. I just push g and then, say, 'y' and I'm now moving on the y axis. Then click the
mouse to accept the movement. Rotating is done in a similar way with this button here and again
you've got the three axes you can rotate around. Alternatively I could press 'r' for rotate and, say, 'y' rotate on y, rotate on x rotate on z ... and then click to accept. And the third most-common
operation you'll want is scale, which is this button here. Same thing: you've got your axes and
you can scale away. Now, one quick note: let's say I rotate my object like this ... and now I want to scale
it like this. Well my icons here don't really allow me to do that. That's because it's referring to
what's called the 'global axes,' which is the world view. Because I've rotated the object off of the
world view, I can simply go up here switch it to 'local' axis and now I can transform my box exactly
how I want. Okay, to delete anything in the scene select it and then you can right click and say
'delete.' To add objects to the scene, go up to the add menu, go down to mesh, and if i want that box
back, I'll go 'cube' - and there it is. Okay I've added another object to the scene. I'm going to go back
to this marquee tool here - which allows me to multi-select things, and then I can move them
all at once. If I want to group them so Blender considers it one object, with them both selected,
just right-click and go up to 'join' or push ctrl+j. And now you can see up here in the outliner it
just says 'sphere.' And by the way I could double click and rename it. Now, the next major thing
you'll want to do is edit the individual elements of an object. Look up at this menu here: it says
'object mode,' which is the mode we're in right now. Click that and go to 'edit mode,' or press TAB
to toggle between the two. Now I have access to individual points, which I can interact with in
three ways, controlled by these three icons here. By default we're on vertex mode and I can select
a vertex or shift-select several vertices. I can click on my move tool and move them or I can push
'g.' And that same principle applies to rotating and scaling as well. If I click the second icon, which
is edge mode, I can select the lines or edges that connect two points. Again, I can shift-select
multiple, and move them. One handy feature here is to turn on the x-ray mode by clicking there, and
now I can click on edges that are otherwise hidden. By default that mode is off and I can't access
that back edge unless I actually rotated my view to go find it. So that comes in handy. And then the
third method of editing an object is face mode, and now I can select individual polygons, or faces, as
blender calls them. Okay, in conjunction with moving and rotating and scaling, the main tool we'll be
using to block in a scene is the extrude tool. So I'll make sure I'm in face mode, select a face,
and it's this button right here. Or you can press 'e.' Click the plus icon and extrude that face up.
You could click the plus button again to extrude again, maybe scale that new face, extrude it, extrude
it again, scale it, extrude it, and look at that - I've made an arcane medicine vial! See how powerful
that is? A similar tool is the 'inset' tool, which is the one below extrude. Click on the circle here and
the inset face will stay on the same plane, but you do have new geometry there. And if I wanted a top
for my medicine vial, I can push 'e' for extrude. It defaults to the z axis - you can see that blue line
there - and I can drop that down...and look at that! Sorry, that's my camera getting in the way. Might be
useful to go up to the outliner and hide that guy. Okay, you see how this object is kind of wonky? That
was just a byproduct of me playing around here... but let's say I actually wanted to intentionally do
that. Well, I can go to edge mode, pick an edge, and move it. But this is too minor. I want something
more impactful over the object as a whole. Enable this button here, which is proportional mode. Now if
I push 'g' for move, you see that circle? That's the radius of influence. I can scroll my mouse wheel,
increase the radius of influence, and now I can manipulate this object as if it were made of soft
clay or something. And I'll just click when I'm done. Sometimes you'll want more points. The tool
to add points is the 'loop cut' tool, right there. Clicking it and moving around your object will
give you some options. I can click my mouse and it'll give me some points there. If I want more
control over where those points land, click and hold, and you can drag your loop cut wherever you
want it, releasing the mouse when you're happy. Now I have more edges and vertices and faces to play
with. All right, so let's do this for real! I've got in mind a colorful science lab environment. You
know, something befitting Arcane. But I'm not going to copy or trace this time. This will be purely my
own creation. So I've just taken the default cube and added a bunch of edge loops. Then in top view
with x-ray mode turned on, I'm shaping this cube into a curved desk. I don't need all these edges at
the top so I'm multi-selecting them then pushing 'x' which is the delete menu, and saying 'dissolve
edges.' Now in face select mode I'll select that fresh face and start extruding it into a beveled
desk. I'll take some of these faces too and extrude them for a little construction detail. Now, I like
to do this with render mode turned on. So just click on this button here then go down to this tab
and change the render engine from 'Eevee' to 'Cycles.' Cycles is a more accurate ray tracer. So here's a
cool little modeling thing: I made a cylinder, then in edit mode i'll select both the top and bottom
faces. I'm pushing 'i,' which is the shortcut for inset, then with both of those new top and bottom
faces still selected, I'll go up to edge and say 'bridge edge loops.' This gives me a ring that I can
play around with, and I'll duplicate that a few times because i think i'll use this as a thematic
shape. The shortcut for duplicate is 'shift+d,' which is exactly how i'm placing these columns as well.
To make a chemical flask I'll start with a sphere. In x-ray mode, delete these top faces then extrude
out this remaining edge ring until I have a shape that I like. Now I've switched back to render
view, just placing elements on this desk. Those books are just boxes with a few extrusions. And if
you're wondering, the ConceptD is plenty powerful! It's actually more powerful than my pc. Even as
the geometry started to add up on this scene the render engine preview was able to keep up with
me in essentially real-time. I'm sure you'll find that combining 3D with 2D unlocks a part of your
creative brain that may not be there if you had to draw all this. Now, let me get ahead of something
that I'm sure will show up in the comments. A lot of people will call this 'cheating.' And to an extent
I agree. If you can't draw this stuff yourself then using 3D will likely become a crutch and I
think you can cheat your own artistic development that way. But if your drawing skills are sound, then
working this way is efficient. And the smart thing to do if you want to work in this industry. I
mean, you still have to have good composition aesthetics, good lighting aesthetics, you know,
texture, detail ... all of which we're going to be completely repainting after this! The only thing
the computer really solves for you is the correct perspective from your given camera view. That's it.
Everything else is still up to us. So here's the 3D block-in I ended up with. That's all super basic
modeling, and just the one default light source. Now we'll convert this baby into painting
mode. And here's how I like to start things off. I'll duplicate the render, then crush the levels,
both black and white. I'll go to filter -> stylize -> find edges. That gives me a line drawing. Set that
line drawing to 'multiply' mode of course, then on a layer beneath that ... start painting! I like to really
decrease the opacity on my line layer and I'm also going to start this with a noise texture applied -
just pasted onto my paint layer. Then using the lines as reference I grab the lasso tool and start
blocking in the shapes that are there. So in this case i started with the focal point desk then
the shadow on the floor around the desk and then other little objects that are close to camera,
just filling in my paint layer so I can start using it as reference and not the 3D. The 3D is
there more as a guideline - a way to jumpstart the scene. Once I get into painting it's the painting
that tells me what needs to be done. Of course, the 3D being there affords me the luxury of zooming in
a little sooner than I otherwise would, But as with all paintings, I'm really interested in getting
a strong read right away. If I were to analyze this right now using our readability exercise from
earlier - that study would look something like this. Essentially three simple groups of value that
help this thing pop. Here's a cool trick: I'll go back to the original render and magic-wand select
some light pixels. I'll push ctrl+j to put that on its own layer, set that layer to 'screen' mode,
and give it a black layer mask then paint white into the layer mask and reveal all kinds of cool
shapes! So, I wasn't going to do this originally, but let's throw a character in there. And this guy is
inspired by one of the henchmen in the show. Just scribbling in a simple line drawing on a new layer
and painting it up on that same layer. Readability-wise, that character shares its value with the
desk, because I consider that character part of that general group of the picture. Here's some
floor texture being laid in on a new layer set to 'screen' mode, and distorted into place. You know the
Instagram 'vignette' filter? That's what I'm doing here ... just a little more artistically. Another value
structure that's going on here is this alternating light/dark/light/dark/light. That really helps
create depth by keeping you accountable for the shapes that happen as depth increases. The
stylus on the ConceptD uses Wacom technology, and it really responds well to a wide range of
pressure applications. All right, at this stage, I feel like there needs to be another character in
there; I really wasn't planning for this! So here's maybe a doctor character working away. It's not
easy to come up with a character design, a pose, and lighting all in one shot, and I'm struggling
here. So I'll keep his head and try something else. Putting in those shafts of light helps integrate
him. Of course the characters are all on layers so I have the ability to manipulate them. And yeah,
here's what I'm calling the final! This represents about one workday. And of course, with my 3D file,
I could easily render out a different view and paint that up; it's very efficient. So I hope you
got something out of this video! Thanks to ConceptD for sponsoring this; I'm really enjoying that
ConceptD 7 Ezel. All right, see you in the next video!
