Rotohail's "Reclusive" SketchBook

[Gliger]

Before anything,  if my eyes are messing with me and my version is just shit and I don't realise until a few days later, ignore the actual paintover and simply focus on the fundamentals that I think you should practice. Also I may be off about everything I say so, yeah, you decide what sticks and what doesn't. 

First of all, sorry for fucking up her hair/mane, it looked dope, but the pose and angle weren't working for me, it merged super hard with the rest of the silhouette and the rendition wasn't as iconic as it could be given the design (which conceptually is fucking dope, love that mane), second of all, I went in a roman soldier/gladiator direction just to have something to go off from, and some elements to use and reference, I'm not AT ALL saying that it's a better concept or that you should go in that direction; saw her design, thought of gladiators, lions and sphinxes, so yeah, thought of romans.



EDIT: Fixed up the cape overlap value a little bit






lighter cape



I don't know if there's final work that you aren't posting here, but going off of what I see in your sketchbook you have 2 big "problems", and those are how you show form and structure, and how you make elements "sit" within your design, you have plenty skill, but it feels like you draw expressively and eyeball most things without having picked up a list of best practices when it comes to showing structure and solidity/unity.


I used to kind of draw like that (don't get me wrong, you are waaaaaaaaaaaaaay better than I was by a long shot, and your painting studies from reference are fantastic, I can't paint for shit).
BUT, a problem that comes with eyeballing these things without a solid mental checklist is that you fail more often and harder than necessary, and you contantly have to backtrack, paint over, move things around, until they feel in perspective; and even when you do it correctly you may not know exactly which elements made the image succesful, at least not in a way you could easily reverse-engineer and repeat without much trial and error.


I mentioned the hypothetical final work you aren't posting because I often see people succesfully get past these problems 1 of 2 ways, 1 is just spending a ridiculous amount of time on each image and picking good reference.
The other is just sticking to a specific tradition and just doing it the way those people do it from the start, I made my choice based on the type of work I have to do and the things I have to achieve with it in terms of design and general perspective, I look a lot at french, american and argentinian inkers and Madhouse/gainax animators,
plus other people who were trained or influenced by them, stuff where you really have to be conscious of what performance you need to get out of your model in terms of action, shadows, sectional perspective(basically overlaps that show space) and whatever else.


If you haven't picked a specific style or tradition that solves the problems you need to solve, do it asap, and rip them off, copy their problem solving, their visual logic, their staging and posing, their lenswork, their surface treatment, don't worry if people think your work is derivative, if you think other people's isn't, you just aren't familiar with the traditions they follow.


Part of the symptoms of these problems are:


1. Your costumes always feeling like wrinkly cotton or smooth plasticky leather with decals on, and even when I look at your weapons, either blades or guns (those look very nice btw), I just keep thinking "you could've made that part look more solid", "you could've reinforced that transition" through either a value overlap (from a cast shadow for example, or just a change in local value), a gradient, subordination or some careful placement of texture.

2. Large areas of jumbled nothingness without much direction, they lack form, they don't imply any physical forces like squash and stretch or pulling and falling of the clothing.

First of all, always transition between different elements, ALWAYS, AND I FUCKING SAID, FUCKING ALWAYS, if it's an unimportant background element, transition with a texture gradient or with a soft value gradient, if you leave it flat it'll be too graphic, it'll jump at you, it will float in 2 places at once. If you have a shape that touches another shape, think of that transition, if the things they represent are really far away from eachother, imply the air in between them, throw some white airbrush in between, imply that aerial/atmospheric perspective, draw the air, draw a line of dudes going behind the first shape, half a tree, a car being cut by that overlapping shape, if they are immediately behind eachother add stronger ambient occlusion where back shape is overlapped, a cast shadow that is cut by the overlapping object..., IF THEY ACTUALL ARE TOUCHING EACHOTHER IN REAL SPACE TRANSITION BETWEEN THEM, blood, scratches, buttons, a guy, fallen off paint, the cast shadow of a bird, put a window in the middle, a decal, a hole, a graffiti tag, sewing, something that really makes you feel like they exist in the same plane.


If there's a sleeve on top of a wrist, IMPLY THAT SPACE BETWEEN THEM, a cast shadow, a value overlap (light on dark or otherwise), reflected light, ambient occlusion, a crosscontour, if you need to be very minimal and conservative then simply use opposition and proportion, different things have a different general gesture, so show that.



If something is directly on top of another thing, throw a cast shadow in there, one that reinforces the form of the overlapping object, but still fits with the shape language of it, if you have trouble finding the right angle for small cast shadows in the form, pick diagonals, don't make any angle parallel within each shadow, and repeat the angles of different cast shadows, i.e. make them all gesturally feel like they are more or less at 53º but don't repeat angles within the same cast shadow, tapering creates direction in a shape/composition, if your shape has 4 lines, you gotta taper it twice, that's not the end of it, then you gotta figure out in WHICH DIRECTION they need to go, but by that point you've done most of the work.
Frazetta was great at this, burne hogarth was great at this.
They also were really good at showing a leg or an arm was behind the body by drawing them in shadow.












Subordination and framing not only reinforce a design's main shapes (its roundyness, squaryness etc.), they (especially when the element is repeated) also help show form and perspective, because it's an extra reference point from which
your brain can relate angles, materials, highlights, overlaps, AND if they have form THEY GIVE YOU FREE CAST SHADOWS, your brain doesn't interpret form just because you got it in perspective right, or because you got the lighting right, plotting physically correct lighting and reflections and getting things in correct linear perspective is great practice, but it's not what makes things feel real and solid, posing, placing the lights in the general right place, and accentuating the formal (as in form) elements of your design is what is going to give you that spatial or 3dmensional feel, people in makeup know this shit either monster makeup or just everyday stuff, it's not about the actual 3d form , it's about what 2d shapes you can get out of it, and pushing and pulling them where you need to.

I didn't just render things more, I added elements that helped show the perspective, that casted shadows, that helped strengthen those transitions and overlaps.





Buy the Alex Senechal design tutorial on gumroad if you want more design theory stuff and more in depth.

https://gumroad.com/acms#SDNyD

I should go into color and value, how you can simplify yours a lot. As well as accents, and how yours don't subordinate right, and they just kinda feel disordered and disoriented,  but this tutorial ^ does a much better job of explaining that stuff than I ever could, I'm not particularly good at any of it. If you got no money just DM me.

When it comes to representing that structure, you gotta pose right, gotta light right, and you gotta break your design right, look at bridgman, look at harry carmean, top plane, bottom plane, top plane, bottom plane, left plane, right plane, left plane, right plane, find those, look at andrea blasich or try sculpting a little bit with a top light tilted a little bit forward, you can do it with an eevee preview in blender, or with clay irl, look at how you show the model through those bottom facing planes, I think of it the way I think of normal editing in 3d, I don't worry much about where the light is hitting or where it stops, I don't visualize the light rays,  I just decide "these planes have to be in shadow, these in the light", and I break things a little bit if I need to, try to get shadows that make the form read. People in anime are nuts at this, they have to be, they often only get 1 shadow value to show the whole form with.

Wish I could find the video, but Myron Barnstone had this rule that if you wanna show form, you gotta show the main (somewhat) opposing planes and the THICKENSS between them, most people fuck that last part up, they say " I can draw a cube" and then they do a nice sharp cube with a good 1-2-3 value read, but they never bevel it, they maybe quickly draw a highlight in that edge because they know enough to know there are planes there, but they never really try to show exactly HOW sharp or dull it is, and in failing to do that, the miss out on the chance to reinforce the form and perspective with that beveled element, not always necessary, but a useful tool to keep at hand.

Also, to show something is behind something else, remember, different elements have a different direction, show opposition in the gesture, you can break this if you wanna make things very graphic and patterny, but generally if you wanna show an arm in shadow behind the body, draw the body and arm along a different direction line, that extra opposition aids clarity, since it's just another form of contrast.













Also, I know you were going for like a cat/panther thing and that's why you did the paw pads, but look at how (I think Sadamoto) did the palms of the hands and the soles of the feet, another way of showind those top vs bottom plane divisions is with changes in local value, it's not a coincidence in lots of animation art, the inside an lower parts of a costume are darker and less saturated (often grey), it almost like getting a free shadow, kind of like countershading in animals, except instead of for camouflage, is so you actually feel the form better. You did paw pad designs, they look cool, but especially since the grey blends with the bakground, and because of the way you posed the hand, I don't get a clear read of the form. I know it's not a design for 2d animation, so it's got different goals and limitations, but I think you could've posed or lit it so that those features sat a bit most powerfully.


Look at the EVA Rebuilds model sheets, or tetsuya nishio's for naruto, look at toshiyuki inoue's stuff on sakugabooru, his magnetic rose and paprika shots, gits as well.

I made a (pretty mediocre) post on neopatogen's sketchbook going a bit over value overlaps, you can check that one out, it was a while back and I'd change a bunch of it now but eh. Also that discord thing I posted on my sketchbook from peleng's server on what I look for in figures.

Just to reiterate, you already do most of these things here and there, but not consistently, so most of your pieces have this feeling of "wooah hold on, what's up with *THAT* part", like you couldn't resolve it, or at least get it correct right away.