How I Drew One of My Best Sonic Fan Art Pieces Yet | Digital Process Walkthrough
TO WHOMEVER IT MAY CONCERN: Hi! If you’re finding this page that was created by me in mid-2022, I think it’s only fair that I warn you my art process has changed and improved dramatically since the writing of this post. You may find some useful information, but just know that this process approach was very long. I’ve since moved on to not using as many individual layers, and I’m not saying this approach was “wrong”, but there are ways to work faster. Thank you! Jesus loves you!!!
If you think my art looks nice, thank you, but it was NOT at this level when I first started. It took hard work and determination to get to where I’m at now. I’ve seen early Sonic art by others that looked pretty decent to me, because at least there was substance. And the artists are like, “that was terrible back then”, and I’m like, “Well, yeah, you’ve obviously gotten better, but your roots got nothin’ on mine…”
You see my early Sonic art? There is no such thing as lack of talent in drawing, only lack of determination. And it’s NEVER too early or too late to start learning. #art-philosophy
Anway, you’re here to improve? Awesome! I personally learn a lot by imitation and observation, so hopefully this breakdown helps ya’ out a bit! (BEWARE: sometimes I don’t even know if my process is “correct”)
I use a Wacom Intuos Pen Tablet with Clip Studio Paint as my primary art program. At whatever point in this process you create your digital canvas file, I use 300dpi not 72dpi(resolution) and a general size of 3000x4000px/4000x5000px works good. (actually, this Also, with digital art you have the power to set your base canvas color to whatever you want. With that in mind, setting your canvas to a gray color can be easier on your eyes, rather than the default white canvas provided.
Once the file is created, I opened references with the Sub View window in Clip Studio Paint. They will serve as my guide not only in the pencil stage, but in the steps that follow, so it’s nice to have. It’s also a good convention to name your layers as you go. This keeps the document organized, while also preventing you from getting scatterbrained when you need to find a specific layer.
Also, I drew a guideline with the ruler for when the half Eggman/Mr. Tinker split(he’s a completely different person when he’s not a villian!), which is used as a reference “do not cross” line layer later on.
SAVE YOUR FILE OFTEN, MORE THAN YOU THINK! (but be careful not to crash the program either)
Step 1: Concept
This can be as simple as: idea in your mind, BOOM, next step. (gotta go fast)
However, for this particular image–(case-by-case basis, because my process isn’t even clear-cut; it ever changes for different styles, experimenting, etc; #art-philosophy: Art is subjective, make it yours!)–I didn’t have much experience with drawing Metal Sonic “technically correct”; In retrospect, I really had no idea with Eggman; The best chance I had out of the three was Belle after her 3D journey, but I still couldn’t draw her properly without relying on reference. What I DID have was a sketchbook and an idea, so even though I lacked ANY references at the time, I went ahead and sketched a concept from memory.
BEAR IN MIND: Even if the concept involves a character(s) that I’ve drawn so much to the point where I CAN draw her/him without reference(ex: Tangle & Mighty), sometimes I’ll still draw a concept(perhaps a VERY messy one…) This helps me to better envision how it will look, ensure that it lays out correctly, fix anything that I thought would work when the idea was in my head, also serves as a warmup, etc.
In addition, this step can be done on paper, so for me, that’s always a plus(less screen time AND there’s somethin’ about pencil and paper that digital just can’t capture!)
Step 2: Pencils
You: Didn’t we just do pencils?
Sonic: I just ran with my idea to this step. What took YOU so long?
I’m here to clear things up. The pencil stage is the visual foundation for your ENTIRE artwork, so you want it to be as refined as possible. Yes, concept involves penciling on paper, unless you flew past it like Mr. Needlemouse over there–(Sonic: Hey…)–but that(for this particular image, can’t stress that enough) was essentially to jot the idea down, not create a detailed basis(FOR THIS PHOTO).
BUT–there are SOOOO many ways that the transfer from concept to pencils can go, a few examples include:
- You had an idea(concept); jump straight to pencils, I’M CONFIDENT! (this IS NOT a bad option, I work this way sometimes!)
- Idea, messy sketch(concept); pencil on a blank canvas, keeping what you learned from your concept sketch in mind; you’ve got this!
- Idea, average sketch(concept); scan in/photograph concept sketch to use as somewhat of a base; compare it to references; might need to change a lot or disregard most of it, but the substance is there!
- Idea, pretty refined sketch(concept); scan the sketch in, pretty good base; still comparing it against some reference; might need to fix a few details, but a strong start!
Besides that, there are different levels of penciling. Pencils can be anywhere from VERY refined to VERY sloppy, in a way it’s kind of like handwriting. Sometimes my pencil base isn’t as refined, but I know I’ll take care of it by winging it in the next step. With Metal Sonic AND Belle having hard-surface robotic parts, it was better in the long run to sketch them out well, but I probably could/should have spent more time on it. (also, it’s helpful to create a pencil layer for each character)
Now, in a nutshell, the pencil phase is subdivided into three main categories, but the transition between each step can be a very thin line:
- BLOCKING, which is drawing out your subjects using basic shapes to create the skeleton of everything. Using shapes rather than detailing everything out at first, makes it easier to adjust parts and make it proportionate. If you don’t do this, you risk drawing something like a nice hand, only to find out it’s in the wrong location–(which isn’t as big of a deal on digital, but still…)–or it’s at an incorrect angle/perspective. But hey, some folks can gracefully lay down their pencils without the blocking phase, however I find that this works better FOR ME. (#art-philosophy: OWN IT!)
- ROUGHING is giving definition to the shapes that you’ve previously defined. In other words, you’re fleshing out your drawing in more detail. For example, Belle’s pigtails may have been represented as ovals in the blocking phase, but now they need to be refined closer to the shape they should be. This step can be done on a new layer, which I recommend if possible, but sometimes it blends into the blocking phase, so it doesn’t always end up as a completely separate layer.
- REFINE your drawing even more than you did in the roughing stage. Add things like buttons or little miniature details that weren’t major at first. This could also mean doubling BACK over your pencils to make everything really clear, defined, and obvious. I’ve seen pencils that were VERY clean, and I suppose that’s extremely important, especially if you’ll be handing it off to someone else in the next stage. As the only person in the pipeline, I keep mine at a point where the lines are legible, sort of mid-clean, but I’m not aiming for complete perfection, since I’ll kind of have in mind what I’m doing. (might not be “correct”, but y’know)
In summary, pencils create the frame for the rest of your drawing. My pencils definitely could have been smoother, especially on Metal Sonic & Belle’s arms/legs, but I spent the time I wanted to on it. You obviously want optimal quality on ALL of these steps, but in the end, you’re the one determining how much you need to rely on the pencils, as well as the time you’ll spend on each part. It can range from quick and careless to spending 80 hours on this step alone(a bit exaggerated). I’m not saying to be lazy, because you do want structure. It’s a matter of finding balance and knowing your goal. Is it practice or are you trying to complete the BEST piece you’ve ever done? Whatever you do, make sure you actually FINISH! (#art-philosophy: Finish what you start!)
To me, what I had was good enough for this one, so I trucked ahead!
Step 3: Inks/Lineart
This is the first step of many that will remain in place to be seen for the final piece. As hard as you’ve worked leading up to this moment, the truth is, the pencils WILL NOT be visible in the end, it’s the lineart that’s traced FROM it.
Because you’ve already poured SO much in to this, you obviously want to keep it nice, right? Well, the first thing you need to know is that choosing the correct program can make this a load easier! As I said in the beginning, the primary program I use is Clip Studio Paint.
See the difference in line art quality? Let me tell you about two VERY helpful features…
- Stabilization is a feature that causes your lines to lag behind your cursor as you draw, generating smooth, clean, non-jaggy inks with ease. It’s something to get used to, but after seeing the improvement in your lineart, you’ll quickly learn how to use it. Crank that thing up to 50, and your lines look like you have a very steady hand(which I don’t, so yay for this)
This feature can also be found in Blender 3D(free). (–what, we’re using a 3D program for 2D art?? what’s the world coming to….) Yep, Blender now has a 2D animation area, which you can utilize to draw smooth lineart with stabilization! (NOTE: If you use Blender, you should probably move the lines to another program for coloring; make sure you keep the lines high quality.)
- Vector Lines are different than the regular “raster” lines you draw by default. There’s a technical explanation as to why, but I’ll keep it simple. With raster lines, you have to manually erase or change your lines, trying SO hard to keep them looking right. With vector, you can actually edit the points after they’re drawn. Be it pulling them into place or slightly reshaping the existing line how you want, editing is a cinch! ANOTHER thing vector can do that raster can’t is epic! Using the “vector eraser” on the “erase up to intersection” mode, ALL you have to do is scrub parts of intersecting lines you don’t want and POOF! Gone. Of course, it’s not ALWAYS going to be perfect, but I think this method is much easier AND faster, especially if you’re not extremely skilled with drawing lines. (vector can also be scaled up or down as much as you want with no quality loss, because the points are mathematically calculated)
There are obviously other helpful features, like post-correction, but stabilization and vector lines are two of the major game-changers for me.
Now, based on my limited knowledge of the subject, inking is the art of drawing lines and controlling the thickness of them to convey and define shadows, depth, etc., prior to any color. I don’t necessarily think THAT deep about it, and I don’t even know if the correct terminology for what I do is “inking” or just “lineart”. Like, the brush I use in Clip Studio Paint is set to be pressure-sensitive, producing lines that vary in thickness depending on how much pressure I put on my drawing tablet, rather than staying one uniform size. Also, I know things such as, the lines for fingers might slim down as they come in toward the hand, or sometimes I can just tell, “Oh, that line is WAY too thick”, but is this enough to be considered inking? Either way, that’s what I generally call this step.
Furthermore, there are many different kinds of brushes. Smooth, rough, pencils, pens, SO MANY. In this piece I used the “textured pen”, which I LOVE! It adds texture to the lineart, rather than leaving it super smooth, which I think brings more variety to the finished piece.
My “inking” process can also be simplified to three steps(I inked each character separately, afterwards erasing what I didn’t need and merging them to one lineart layer):
1 – Base Lines
Basically, this is just me going on a new VECTOR layer to start tracing my pencils. (very important, DO NOT draw your lines on a raster layer; if you do, your lines will be raster and NOT vector, meaning you’ll have to re-ink it all over again; the countless times I’ve done that…) I’m not always very strict with tracing, sometimes I loosely follow the pencils as a guide, it just depends on how I want each part to look. Hands are a place where I particularly end up being loose sometimes(because hands are…hands) But since I will “erase up to intersection” after this, I’m almost always extending my lines so that they intersect. (if your lines don’t intersect, the program can’t erase them at the point of intersection, because there is no intersection; in this case, you have to:
1. lengthen the lines, 2. manually push them in, 3. delete the extra points, 4. possibly try to erase with “touched areas” mode, but the eraser doesn’t work how it would on a raster layer, which may be a thing some people dislike;
I should also note that the third mode for erasing is “whole line”, so if you want one particular line gone, like the entire thing, you can do that) It can be hard to draw one continuous line, so I draw some parts that might be one piece in the end as multiple lines(ex: muzzle/mouth area), because I know I’ll come back to stitch it together later. (+1 point for VECTOR) Make sure you’re hiding your pencil layer as you go to see if you missed anything, how it looks, etc. (and make sure you stay on the vector layer!)
I recommend either duplicating your lineart layer or saving a copy before proceeding.
2 – Erasing Overflow
I hide the pencil layer, looking over my inks to erase any overflowing, intersecting lines. (turning down the opacity temporarily can help you see WHICH layer you’re working on) If there are lines not intersecting that need to be erased, I just leave those for the next step, no big deal.
(you could honestly duplicate the lineart again, just in case, because who wants to back track more than necessary)
3 – Refinement
I pass over my lineart, probably more than once as I go through this step, using the handy vector tools to go around:
- connecting lines(even if they LOOK like one line, try to connect ones that SHOULD be continuous; ex: muzzle = one line; place where head meets muzzle = doesn’t have to be one line(in my opinion))
- deleting extra points that are created after connecting lines; trying to make the lines flow and appear as if they were continuous from the get-go
- pulling points
- deleting more points
- use the textured pen to occasionally add new lines that were missed, need to be redone, etc.
- redrawing existing lines(rarely)
- messing with the line thickness(also kinda rare)
I repeat that process(not always that EXACT order), until I have a finished lineart that I’m satisfied with.
Your inks define the placement of your colors, so it’s in your best interest to have them finalized as much as possible before continuing. The further you go, the more tedious it becomes to make edits, so changes to the outline of your artwork after this step is not desirable. REALLY try to make sure that you’re happy with it! (honestly, even I might change something in the middle of things, but TRY to avoid it)
Step 4: Flats
Flats are the base colors of your subjects BEFORE you shade or highlight. Tinting your colors can be good to help fit your artworks into different settings or specific environments(ex: blue-lit aquarium). For this photo, I mainly swatched some references, then later I apply a blue-ish gradient over the characters. (and flats that were 100% black got changed to a dark blue later on)
I create a new layer for each color, usually as I go, UNDER the line layer. For colors I use a raster layer, NOT vector. (which does mean the color layers won’t be scalable while still retaining quality) I think raster is better for coloring because vector processes different. The mathematical nature of vector is amazing for lineart, but I want more natural flow and control over the colors.
This stage used to be VERY time-consuming, because I would manually fill with a brush, trying to stay in the lines like a coloring page. Thankfully, there is a faster way to do flats! In Clip Studio Paint(probably others, as well), there is a function you can enable called “Do not cross lines of reference layer”. With this option checked on, you can set your line layer as the reference. Now when you color with a brush(ex: milli pen), your colors stay within the lines! (if there are tight areas you may want to fill it manually with the “do not cross” setting turned OFF)
There’s more fine details you can tweak like “fill up to vector path”–(it’s good for filling tight areas that you would otherwise have to manually come back to fill, but sometimes my vector isn’t perfect so I uncheck the “fill to vector” option for that color/portion to manually correct/fill it)–and “area scaling” (I generally turn it up to the max, 20)
The most annoying part about this method is that it’s ALMOST lineless if you hide the line layer, but there’s like a pixel of space that doesn’t get filled under the lines. So if you’re looking to do lineless art, this is a good start, but I’m just warning about that. (there may even be some quick solution to it) Also, some people say you use the paint bucket with area scaling to color, which I may experiment with more as I continue to grow, but for this piece I used the “do not cross” method.
Lastly, although it doesn’t directly affect what is visual, it’s good to name your layers accordingly. This can be something simple like “yellow” or “cuff-yellow”, or if you want to make it unique “eggman-cuff-yellow”. I also generally don’t do a separate layer for the same color swatch(ex: Metal Sonic has blue parts that are technically separate pieces: head, left arm, right arm, left leg, etc.; I keep it all the same layer because it’s all blue) I compile each character’s color layers into a group named something like “belle-colors”.
Step 5: Shading
There are many different types of shading, but you’re really not forced to stick to only one as you create new pieces. In fact, I think it’s good to try and see which ones you like, observe if one kind works better for this piece, etc.
Cel shading is a style I’ve used a lot in the past. It’s a hard-edged style, used in 2d animation/anime, and it’s styles similar to it appear in some comics. With this style, sometimes I feel like there’s no rhyme or reason as to why shadows don’t appear in certain places where they would in real life. On top of that, you generally have to keep the edges smooth and clean, unless you’re doing it different. I think I might understand cel shading better the more I do it, but it’s not my favorite method.
The style I used in this piece is actually my new favorite method. (and this doesn’t apply to everyone, I’m just sharing MY preferences) It’s more 3D-like, and if the edges are messy, fine, because the brush(noise airbrush) is already textured AND I’ll blend it out at the end anyway. (the blending step, which is up ahead, can also be used to dynamically change areas that you may realize the shades aren’t working in)
I would say steps 5-8 crossover each other, so it’s alright to bounce in between them.
Before beginning, you need to determine your light source. In this photo I kept it simple, just one. Writing “Light” and having multiple arrows pointing in the direction the light is coming from/to on a separate layer can be a helpful to look at as you shade and highlight.
I think for this one I used a gray-ish color–(MAINLY; I did use additional colors)–from my references in combination with blending effects. (I save the main shade color to my swatches) However, I find that sometimes the gray-ish color doesn’t work so well to shade, like with Eggman & Belle’s skin tones. In those instances, it’s better to shade with the color you’d like it to be instead.
Make a new raster layer ABOVE the color you want to shade(that’s right, I gave each color it’s own shade layer, sometimes multiple, like Belle’s eyes; I normally do it as I go through each color) Then, “clip” that layer to the one beneath it, the flat color layer. Clipping is a feature that lets you draw without going outside the boundaries of another layer(ex: when shading Eggman’s mustache, the clipped mustache-shade layer will only be visible on that area.)
Blending effects were also applied to the layers, mostly color burn or multiply, and I messed with the opacity here and there. I didn’t name the shading layers(being honest), I should have, but I didn’t.
Step 6: Highlights
The process for highlighting is pretty similar to how I shade. I use the same brush(noise airbrush), create a new layer for the highlights(yes, one for each color again), and clip it. (you can clip multiple layers to the same flat color layer) My layer stack is usually: highlight layer > shadow layer > flat color layer (but both the highlight AND shadow layers are clipped to the flat color one; this happens automatically when you clip them in a row; the highlight layer is NOT clipped to the shadow layer, even though that’s how it looks)
The main differences from when I shade are–(aside from the fact that highlights hit in different spots than shadows)–I set the blending effect to glow, overlay, or a similar one to brighten it up–(messing around with the opacity as I go, once again)–and I used a light color(did I swatch for this one? it was a light gray-ish white; saved this swatch as well).
The last thing I’ll mention in this section is that sometimes I might want to see some of the base color showing through, not having the artwork completely engulfed by shadow or light. That’s not really a strict rule of thumb, nor is it necessarily my preference, but you can keep that in mind.
Step 7: Bounce Back Light
Bounce back light is an effect caused when light bounces from the ground plane back to the object, appearing even in shaded areas. It’s something that’s not too complicated, but it makes a huge difference in adding believability to your artwork.
You could definitely go and create a new layer for bounce back on each color, clip it, so on, but I didn’t here. Instead, I cheated and only created a bounce back layer for each character, using the same color I used for highlights. I clipped a new layer to the color FOLDER of each character, set the blending effect to glow, and lowered the opacity, even more than the highlights.
My solution didn’t turn out too bad, since the bounce back is an extra detail that’s sort of secondary in comparison to the shade and light. Anyway, this changes everything, and you can tell just by turning the bounce layers on and off.
Step 8: Blending
This one is pretty self-explanatory, but first I duplicate ALL the colors, shade, light, and bounce layers. (or you could save a copy, which is probably more efficient for the computer to process)
I use the “textured blender” blending brush to blend the ends of each layer created in steps 5-7. It makes the effects not as harsh and creates fall-off, similar to a dithered 3D render. Once again, steps 5-8 blend together. So if I see something that’s not working or that I don’t like with the shade, light, or bounce, it’s perfectly fine to erase, add, and tweak those things, then blend it again. (just make sure you’re working on the correct layer with the right color)
Step 9: Line Art Color
(notice I made a very late edit and removed Belle’s lower eyelids…it was do-able, but you should try to avoid these types of situations.)
I select any lines that’s supposed to stay it’s color and cut/paste it to a new layer(ex: Belle’s eyelashes and pupils). I changed lines that were solid black on that layer to a dark blue. (apparently I left the other lines(not the layer with Belle’s lashes) 100% black in this one; I would maybe change that in a future piece) I changed the blending effect on my lineart layer(the non-Belle lashes one) from normal to overlay, turning the opacity down a little bit. This is a really quick way to achieve a lineart color base that’s not just solid colors, but has shade/highlight. It’s close to lineless, yet still there to reinforce the artwork.
You could leave it here, call it done, but you’ll see those little pixel lines mentioned above. To resolve it, you can fill those spots in with different techniques, maybe touch it up with the correct swatches. I stopped here and resolved it with Pretty Surge the Tenrec, but the lines aren’t perfect. Another possible solution is to duplicate ALL of your flat, shade, light, and bounce layers–(please save before doing this, it uses a good bit of memory)–merge them to one layer and blur it. Place that layer UNDER your initial color layers, and select the area outside of your lineart to delete it from this layer. (I learned this from the internet)
Of course, you can also change the line colors by selecting them and changing it manually, or you can draw on a new layer clipped to your lineart layer with different colors. I’ve done both in previous pieces, but this time, I went a different route…
Step 10: Back to the Flats
To help eliminate the empty pixel spaces, as well as give priority to lines that should be in front, here’s what I did(this photo was my first time really doing it this way, and I’m not even saying that it’s the right way or that there isn’t a better way to do it):
- I changed my eyedropper to pick colors from the editing layer(the currently selected layer)(this saved time so that I didn’t have to hide the shade layer every time I wanted the base flat color swatch)
- Ctrl-Click on the lineart layer to highlight that as my selection(in hindsight, this may have caused some extra leftover color crumbs; solutions may have been expanding the selection, or just not using the selection at all)
- Went through each flat color layer and erased/colored where the lines should/shouldn’t be on that color(ex: Eggman’s hand needs to be in front of Metal Sonic’s head, so erase any blue that was overlapping that area)(working from the top layers, downward is better because you’ll take care of the layers in front first; only what’s visible matters; ex: if Eggman’s red jacket layer is ABOVE his black pants layer, any red that’s not supposed to be there needs to be erased, but if I hide the red layer and see the black goes where the red belongs, that’s fine because when the red layer is visible it covers the black up)
- Deselect my selection, color/erase anything that was left behind from when I only had the lineart selected. (and take care of any “crumbs”)
If you look closely at the finished piece, you’ll see the lineart wasn’t perfect, but it’s not bad! As I create more art, I’ll keep experimenting different methods, but this was the process for THIS ONE.
Step 11: Finalizing
Almost there! Here I’m looking over my artwork and making sure it’s good. Double-checking my shades, lights, bounces, opacity levels, blending effects(multiply vs color burn, etc…), line color is in order, and all that stuff. Yep! Looks great! I hide the gray background canvas layer and anything else I don’t need(like the light source reference).
Step 12: Export
I hope that you’ve saved your file as a .csp(Clip Studio Paint file) or a .psd(Photoshop file) or whatever kind of file that can store all your layers if you need to make edits, a million times in this process. (#art-philosophy: SAVE, SAVE, SAVE, DO NOT RELY on autosave!!!) (even though sometimes it might cause the program to crash, ugh, so annoying when that happens)
With your file saved in a editable format, you can now go ahead and save it as a PNG or JPG (careful, if you make any changes after saving it as a .png or .jpg in Clip Studio Paint without closing and reopening the document, “Save” will only save it as a .png or .jpg again; Answer? You have to “Save As” a .csp or .psd to keep the changes as an editable file)
Open the PNG/JPG file in the art program(or you could keep working in the file, skipping the export process, but I think this helps conserve memory and processing power, which can help prevent things from crashing; however, if you need to make any layer changes, you would have to go back, fix it, export again…which is tedious, so whichever works for you) (also, you might want to “Save As” a .csp or.psd once you open the PNG/JPG before you start compositing in step 13)
I enlarged the canvas size(not the image size), to give me more wiggle room. Position the artwork layer to fit in the center of the larger canvas.
Step 13: Compositing
- Clipped a light blue to medium blue gradient, Soft Light blending effect, above the artwork, which provided atmospheric color
(the following layers were UNDER the artwork layer and are displayed in the correct stack order)
- Ground shadow w/ noise airbrush, Linear Burn blending effect(I think I used the same shade color swatch)(I came back to fix it more after receiving insightful feedback)
- Random blue swirl w/ noise airbrush
- Circular vignette-ish surrounding w/ noise airbrush, Subtract blending effect
- Solid light gray-blue fill layer, Multiply blending effect
- Blue to white-blue/white gradient
Step 14: FINISHED!
I estimate that this picture took over 10 hours, but I’m very proud of it!
Well, that’s my process! I hope you enjoyed this walkthrough of all the steps that went into creating this digital art piece! (as you have seen, it’s an effort)
If you want to get better at drawing in general, you just gotta keep doing it! Studying art by others, observing real life, 3D modeling(Yup.), and getting advice from family, friends, etc. are all great ways to improve, I know it’s helped me out! I kept drawing, and I WILL keep drawing, bettering with every step, loving all these moments, knowing there’s still lots to discover!(which makes it fun!) YOU CAN TOO!
If you want to see more art from me, be sure to subscribe to my email list on the right so you never miss a beat! (only if you’re over 18)
Thank you for reading! Jesus loves you!!!