I actually wasn't intending on doing a Patreon post - Process of this one but then I was like why not. And it also helps me digest this while doing it as well as after. The video just covers the Adobe Fresco portion from approved sketch to final color sans overlay + adjustment layers.
The premise of the commission was two characters chilling on the tower from the tower control mode in Splatoon on one of the night stages.
These are initial thoughts before looking into reference (proportions) and without caring about perspective.
Gathering in-game reference is obvious. I spent about 1.45 hours gathering reference mostly stressing the background of the map and the size and proportions of the towers to the inklings/octolings and weapons (rollers)
Internalizing the reference:
(This is with the Wally World tower control reference)
(This is after learning that the size of the tower varies per map - and making said adjustments)
Essentially trying to figure out how large the tower is and how small the inklings/octos are on it. I took over probably 100 reference photos, overall, but these were the reference photos I used for the tower:
There is also reference I took of the rollers but I didn't transfer them to the Pureref file since that was further down the process. Pureref is free and lets you basically drag and drop images into one compartment so you have it all in one place. I had it prior but never used it because the controls/commands were odd. They're still odd but I find it much useful for commissions.
So I go into Sketchup and model out a mock tower. I tried to get what I thought was a square shape for the base. I did have to readjust it once I learned the Skipper Pavilion tower was slightly shorter than the Wally World one.
Checking it to the game's one. And then when I'm done enough, I rotate and move the camera/view to views that I want. Then I will screenshot it, take it into Photoshop and draw over top.
Because this is more of a fully illustrated commission - with a background, I do multiple thumbnails/views for options.
(This was like a throwaway option just to see what it could look like but it died here)
So those are 5 thumbnails. I narrowed it down to 4. Then because I hear stories of like the commissioners being unable to read thumbnails, I choose to clean them up before presenting them.
Thoughts on the thumbnails:
I favored a looking up camera angle because it features the fireworks and night sky on the stage. It also gives me less grief regarding detailing the background. I did one with the camera angled down just to see what it could look like. I wanted to make sure the lanterns were in there as well since that's specifically unique to that map, I think.
Roller placement was dictated by the fact that they needed to be seen but also shouldn't look too same-y regarding the pose. Ink tanks were a given; also preferred them not to be too symmetrical as well.
The reason to have one posed with a leg up was simply for variety; the last thumbnail kinda shows that the same pose is visually boring.
Final cleaned thumbnails submitted:
I colored in the objects with a flat color just to make it easier to read.
I'm not sure whether I should put the spheal here about how using 3D modeling isn't considered cheating or "improper." The TL;DR is: if you know what you're doing, using 3D modeling makes the process more time efficient, which in this case, is essential when you know the time is going to cut it close. (And I don't want to spend 7+ hours of pain from that inkling apartment piece with perspective issues.)
You still need to have some vague understandings of perspective or using 3D models would still fail regarding proportions and angles (the most) - when you draw in the characters and details.
So this was the one picked:
And now I take it into Adobe Fresco to clean it up and do most of the work in there. I did the thumbnails in Photoshop specifically because i had screenshots from Sketchup and that was quicker than transferring those screenshots to Fresco.
I did allow myself to adjust the size of the characters because it is a character illustration. Overriding technicalities for artistic license.
Start refining the sketch. I did check for the ink tank placement on the rollers.
With this hero roller, I intentionally made the holding handle part a bit longer than it actually would be because it looked too small when it was technically correct. Because my proportions of inklings/octolings are more stretched out than the ingame models, I allowed to adjust the rollers so that they would appear "normal" as well.
And then start fleshing out the characters:
I always try to be aware of the contour of the cuffs and whatnot to make sure they're in the correct perspective-ish.
Shirt design was requested to be a toasted marshmallow. Which was much easier than attempting to do the actual design on the shirt.
And onto the ink tanks:
It was a little hard to figure out where the rings should be that show up in the back. I think I had to fudge it a bit for the one on the right but it's such a small thing that I don't think it's greatly noticeable.
I also realized that there's a little divot in the Splatoon 2 ink tanks but it's not that noticeable and by the time I realized that (around coloring) that it wasn't worth the time to fix such a small detail.
After the second pass of cleaning up the thumbnail, I send this off to get approved and put down placement ink colors that get changed later so that it fits the color comp. I also add the ink splotches to check if it was enough or not enough:
It gets approved and then I go to the final linework pass and set up the detail areas (like shoelaces, color on the pants, etc.):
I think there might be some slight perspective issues with the roller but it wasn't enough to distract from the piece overall and wasn't worth the time.
Because doing linework all in one go can be boring, I switch to doing flats as I please, ahhaaa
I try to clump the flats together so that I avoid having a separate layer for everything like the Traveling to the Splatlands piece. Also because the clothes are 90% all black, it works just to leave as much as I can on one layer (Hat, shirt, pants, shoes are all on one layer).
Take a break from flatting because it can be boring and switch back to doing the linework.
And then flat:
And then linework the tower:
I wanted to make sure the hand was correctly gripping the grating of the tower.
Now I need to take it to Photoshop and make the designs for the logos on the bottom part of the tower as well as the grate:
With the arrows, I made one then flipped it.
I think I actually then took one of the arrows back to put in the little squid tentacle icons and then reduplicated it so it was even.
I thought I could get away with just selecting the area and then brush a dot brush over top. However the dots weren't even so I had to manually clean up the area around the tentacles for both sides. The screenshot above shows a version with the manual edits and the other arrow without the manual edits.
So right now, if you worked on a file in Photoshop, you cannot take that file back into Adobe Fresco without all the layers being merged. So I put the tower onto a separate file, export it as PSD, upload it to Google Drive and then download the PSD file into Fresco and that will give me the layers. It's a janky workaround and I don't think they're fixing it anytime soon.
Onto the grating:
Because I couldn't find a plain diamond pattern brush, I opted for this fancier diamond-ish pattern.
Because grating is a pain and then drawing the stuff underneath it, I intentionally chose to leave it opaque. The main reason being I was fairly certain it wouldn't be noticeable at first glance and that means the goal is still achieved so no need to waste excessive time on that.
Distort it and try to take a gross guess on where the edges are on the other side so it looks relatively correct.
Make the grates a clipping mask layer so it doesn't go outside of the intended area.
Save it as a separate smaller file and open that in Fresco so it will let me open the layers. And then I copy those layers over to the actual file.
Delete areas where it overlaps, like the roller overlapping on the tower gratings.
Start cleaning up:
Since everything is now flatted, time for color comping.
Tried to do n overlay layer in Fresco but I wanted to check in PS, so I took it to PS and then transferred the overlay layer into Fresco.
Used Photoshop for overlay layers since it has the triangle color wheel.
Sent that last photo for color approval before going in and cleaning things up.
I duplicated the color comp and made it smaller as something to reference when zoomed in.
I did a loose wireframe just to make sure the designs were steady.
Compare the next 2 images:
The lantern designs and little indents - I'm not that concerned about them being uneven as well as they don't look too out of place.
And then the lanterns are finished.
Slight unnoticeable gradient.
Adding shadow areas where I think they would be on the tower.
Lowered opacity of the signs even though they glow bright just so they don't steal the spotlight.
Thankfully the shoes were basically all black so there was no fine detailed coloring involved.
And onto the rollers:
Thankfully I had a dot brush, from Kyle Webster's brushes, that made it easier to do the dot-work.
Same process laid out for the Hero Roller.
Started the tanks then realized I forgot the little strap. I didn't think it was a big enough detail to add on but it bothered me enough to add it in.
I just tacked it on top so it would be easier.
(Says trace / line out fireworks)
I try to tie in the colors where I can to keep the color palette centralized.
I look over it to see if there's anything I need to do before taking it into PS for final work.
(says soft brush)
Scattered dust and gaussian blur.
Look at B&W to see what things could be adjusted/nitpicked:
I knock down the lanterns contrast because it's not the main focus.
There's what I seem to understand about the color wheel and how the placement affects how the overlay layer works (to saturate or change value of the area).
Blur the tower building silhouette so it appears far away-ish.
And then the finished product:
I'm relatively quite pleased with this one, especially with the handling of basically all black outfits. I intentionally chose warm grays for the octoling and cool grays or the inkling to give it variation rather than straight deadpan grays.
I also consider the workflow that I did to be relatively efficient as well. The lanterns were maybe the most inefficient part - or it took longer than I thought it would have. The workflow of taking it to Photoshop when I knew it would be quicker helped make better decisions with overlay rather than fumbling with the square color wheel in Fresco.
I also think working on the tower patterns in Photoshop and using the distort tool was also more efficient than trying to do guesswork straight on the illustration.
Specifically for this one, gathering reference was more time consuming than normal. Since I had to get references for the map, for the tower, and for the detailed splatoon clothes. I was also relatively concerned about the request for purple and yellow ink colors since they're complementary colors and that color combo can get dirty/gray/dull fast. Thankfully I think making it a warm purple and warm yellow worked well.
We won't talk about how I forgot the rim lighting on the masks and will throw that under the rug of "the mask material is different" category.
I feel like there's a lot of stuff in focus though (rollers, ink tanks, fireworks) and I'm wondering if some blur effects could benefit this. It would take away from the details though so it might not be ideal on a commission. However as an illustration, it might benefit from blurring so that it can help emphasize the focus on the characters. I also think the cast shadows might be inconsistent with how the rim lighting is but we will call that aesthetic and to help the items "pop out" of the background more.
And if you have any questions let me know! This was I think the first time I took screenshots and made notes on the screen. Hopefully you can read the handwriting. I thought that would be more helpful to know what I'm referring to as well as my real-time thoughts while working on this. This is like literally every decision I made. I think some things could be cut down to include two steps, but I guess that will come with time and ability to judge that.
Now that I have fiber, I'm also recording my live process on Behance, which has been described as Adobe's Twitch. I'll try to tweet out when I'm live. But what's also nice is that you can do private livestreams as well, so if I'm taking a lot of breaks or something, I'll do a private livestream instead so my process is still recorded but it's not public.