It's been around 3.5 years since I really got into spriting, and since I've gotten to the point where people ask me for tips/tutorials sometimes I decided to write something up about my experience learning/improving upon my spriting abilities. I'm writing this partially for myself, and partially as a sort of guide/inspiration post for beginning spriters. Throughout this entry I'll be linking a lot of my spriting work, old and new.
When beginning spriters ask me for help and I take a look at what they have, I'm reminded of some of my first real (or attempted) spriting projects, back in like late 2010/early 2011. Stuff like
the Ancient Minister and Konata (
JUS edit attempt,
semi-custom attempt). Even somewhat later stuff from early 2012, like my JUS edited
Kanade and
Kat, neither of which I'm particularly proud of despite the (relatively) obscene amount of views/favorites they have. Very simplistic, small, busy palettes and detailing, and either edited from commonly used sprites or badly customed. Around this time I was telling myself "If I have a base to work with, I can pretty much do whatever." I had very little faith in my ability to make sprites from scratch, and my largest projects to date (R.O.B. and Pichu, Smash Bros sheets for Super Smash Bros Crusade) both had existing bases that I edited and worked from.
After finishing said projects, I decided to expand on my talents a bit while having fun, and make a bunch of sprites of anime characters I liked. I found a couple good bases and just edited for hours on end. This resulted in some decent sprites (
Nichijou,
Angel Beats!), which were definitely leagues better than the JUS edits I made just a year before, but still not particularly impressive; anyone can edit a sprite into something else. It just takes references, colors, and patience.
Around this point I had the idea of
Miyazaki Mash (or rather, decided to expand on someone else's idea), which turned into my first real attempt at custom spriting. It was very challenging at first; Totoro, the first one I made out of the bunch, was remade around 3 times throughout the process of making all the bases, and there's a definite disparity in quality between some of the earlier ones and the later ones. It did force me to gain experience with a lot of different anatomies and palettes, something I hadn't really dealt with before (and before this I always told myself I'd never be able to sprite humans from scratch). The most important thing it did though was prove that I could do something other than edit, which was a very important step forward.
By this point I decided to try out some more styles, and ended up making some
Cucumber Quest sprites (it being one of my
favorite webcomics), as well as
revisiting my Ancient Minister sprite. Eventually I stayed up with a friend one night while she was writing an essay and I decided to sprite
Ryuko from Kill la Kill, which inadvertently ended up defining my process and what I now call my particular "style". As I mentioned before, there's a 3.5 year gap between when I started, when everything I made was a messy jumble of pixels, and now, when I actually look at the sprites I make and can say I'm proud of what I made. It wasn't an easy process and a lot of people from several communities helped me along the way.
On a slightly different note, for the longest time I've had difficulty getting started on spriting projects because, until a sprite is 95% done, I don't like how it looks at all and criticize literally every part of it, even at the skeleton/outline stages when you can't tell how the finished product will really look.
Here's a sort of collection of in-progress stages of my most recent sprite; until I hit some point between the 9th and 10th step (just as I started filling in colors), I was convinced it was going to come out terribly, the anatomy would be off, the pose wouldn't look right, etc. This is basically what keeps me from spriting more often - hating how it looks until it's basically finished, constantly nitpicking until there's nothing I have a problem with anymore. My original fix was to just sprite late at night (like 1-3am), when i wasn't awake enough to criticize it while I worked, but now I can just look at the last couple things I made and say "Yeah, I think it'll turn out fine", even though I still hate how it looks during the entire process. Basically the point of this paragraph was, don't quit in the middle of something just because it's not quite right yet, because when you're done with it it'll have to look good to you.