#was all up n zooming in on the layers of trees n foliage

[alaraxia]

Process Breakdown: Starfall

Since I got some positive responses to my question on process stuff I’m gonna do a behind the scenes breakdown for my most recent piece to help people see the process I use and how I problem solve. I didn’t plan to do this initially so I won’t have a ton of process shots but I did save a handful. There’s a few scattered hyperlinks to other pieces I reference too. Just a warning this is mostly train of thought so it’s super verbose.  

So base sketches were mostly focused around me defining the shape of the girl since she was the focal point and building the environment around her. Going in the things I knew I wanted were a girl precariously balanced on top of a massive capybara catching a falling star, while surrounded by smaller sleeping capybaras on rocks. I layered out a general forest scene surrounding it but didn’t really commit to much in the sketches. Messed with the angles of the large capybara a few times to make it feel less flat and more 3D in the space, used a lot of reference photos of capybaras and sorta simplified them to what I thought was cute/ what stood out to me as their defining features.

Skipping ahead a solid amount is midway through the initial lineart, with some areas just colored in to define them as separate. Initially this piece was supposed to be in a similar style as my “Stratosphere Dreaming” art, with a single uniform line thickness, bright colors, and no gradient shading at all, but I realized pretty soon after I finished the lineart and started coloring that I had done what I tend to do a lot and made it too complex to pull off successfully in that style so I had to pivot to using gradient shading and other non-cell style techniques (though you can see a lot of those methods still in the coloring of the girl). This caused an even bigger challenge as I was drawing on a large canvas with high DPI in Procreate which resulted in me having a cumulative 50 layers to work with at any given time (hell).

Now once I made that rendering style pivot is when the really hard part began, and why on top of my persistent arm injuries this took me about two months to finally finish.

1.) I had an extremely difficult time trying to figure out the color pallet for the piece. I had an idea of the values and general colors I wanted (you can see some pallets and random base color tests in the image above) but I just couldn’t get them to look right and I became extremely more aggravated as I kept trying new and different things. My biggest mental block was feeling like I was stuck trying to make the initial pallet idea work, but eventually I was able to bump it to a slightly adjacent pallet and it worked far better. Essentially a lot of angry experimenting and testing.

2.) I made the piece too complex for its own good when it came to the foliage and scene. After finding success with a very specific way to render foliage in one of my favorite pieces I started to use it as my standard, but that standard started to show cracks when I had foliage heavy scenes like in my Hollow Knight piece from last year. The rendering style became insanely too time consuming, and incredibly distracting when used in abundance, taking away from the focal point. I knew this but I still attempted to use the same style to render the foreground foliage MULTIPLE times in increasing states of frustration until I stepped back, evaluated it wasn’t working, and tested out a very similar style with the same effect but that I could throw together twice as fast without the aggressive distraction and minuscule details that were irrelevant in the scheme of the art. This frustration in the rendering not working was only exacerbated by the color pallet indecision making a lot of the attempts just look bad both color and style wise.

Due to the limited layers I had to finish rendering out the girl very early and merge her together to free up layer space, and couldn’t keep my lineart layers as separate as I would have liked to allow for quick line color swaps. She ended up being a key point in defining the rest of the color pallet of the piece. The dress shape was indeed inspired by the Lirika Matoshi strawberry dress, but with my own twist.  

Once I got a more solid color pallet down the rest started to come a lot easier and I was able to begin filling stuff in and doing general color adjustments to make the backgrounds darker and give it more depth. I don’t have any more process shots beyond the initial color pallet exploration unfortunately, but the last hurdle I hit was at the very end once I was doing final touch ups. I found that with the only light source/ lighter color being the falling star that it washed out a lot of the rest of the pieces and made the details I spend so much time on feel unnoticed. I found though that adding the bright orange stardust specks into the trees, the girls hair, and falling from the star itself gave the last bit of color I think it needed without completely destroying the atmosphere. Originally (you may see it in some of the process shots) there were going to be jars with stars already in them illuminating the bottom of the piece, but after multiple trial and error iterations it just didn’t work out and ended up taking the focal point away from the girl and the star too much so I scrapped it.

Finally once I got everything done I made a copy of the entire art file to save as a backup, then with one of the copies merged all the layers together. Once all merged I made a copy of the fully merged layer, and went and adjusted the entire layer copy using a Gaussian Blur, reduced the opacity of the blurred layer to a super low percent, and put it on top of the original merged layer. This gave it that ethereal sort of feel that is difficult to notice unless you zoom in but really helps soften the piece and make it more dreamlike overall. Then I merged that blur layer down, and turned on about a 3% noise layer on it all to give it a bit of texture.

But that’s enough rambling from me, hope this helps give a bit of background to my process and decision making and it wasn’t just a wall of random musings. 

My last piece of advice is if you’re looking to do art professionally, do commissions, or make a lot of pieces in a short period of time I would highly advise against directly copying techniques I use. Because while I’m always working to improve I do only do this as a hobby rn so I have the luxury of being able to invest a lot of time, energy, and details into higher complexity pieces that would take way too long in a professional environment. I can put a lot of time into making a single piece exactly as I want it since I’m not reliant on art as my sole income. As I improve I can make things faster, but it’s still an overall slow process and I just end up moving my quality standards up with any level of improvement anyway. Use stuff I do as inspiration but I cannot stress enough to learn as many shortcuts as possible (I’m still struggling with this myself).

If y’all have any questions about bits feel free to dm, if I do something like this again I’ll try to get better screenshots during the process n try to be less verbose.