「シンプル・イズ・ベスト」
So, I actually drew a bit for Project Homuncupunk while I was thinking about the identifying/differentiating characters post. The main objective was originally the female face, but I think I've gotten that reasonably well now. Due to the very small volume of drawing I do, I have serious problems with consistency-in fact, although I've been trying to draw straight front views almost exclusively, there have been times when the eyes were so crooked that I ended up turning them into three-quarter views. I've experimented some with different hairstyles as well.
Considering the volume of work I need to do and the more and more obvious fact that I'm far too lazy, I really need to try to make things as simple as possible. I can identify several things other than facial features that I need to get right: anatomy, clothing, weapons and another homunculi, and scenery. The main problem with clothing is that it looks too neat and unrealistic; not all characters should have perfectly smooth, unwrinkled clothes, especially when they're moving around and taking action. I may end up ignoring that fact, though. With the scenery, it may actually end up being too simple: I haven't really had much of a variety of ideas of what kinds architecture have been grown into the City. I really can't match Hirano's talent for writing crude graffiti, either. The anatomy and homunculi just need some practice (and development in the latter case), I guess.
For ways of simplifying things, I can to some degree rely on common Japanese methods. I find it somewhat surprising that Western artists use hand-lettering (sometimes requiring a completely separate person) and -shading, even when they aim to make them as neat and even as possible. Japanese artists use mostly computer-printed text, even in some extreme situations (one of the more interesting being Kujiranami's "Battōsaiiiiiiii" with the backwards characters). It seems like this would also help to judge how big your word-bubbles need to be, as you can print the text out beforehand and hold it in the spot where it's going to be even before you draw the panel. I still haven't gotten around to checking out the site that arifyn suggested, but there must be good fonts there or elsewhere for pretty much any situation.
In terms of shading, the Japanese method is to use screen tone for pretty much everything other than showing emotion (blushing, blue in the face, et cetera). It has been pointed out by Ms. dream_bender, of course, that screen tone can be expensive. Since my work is intended to be a Webcomic, I thought I could do the shading with the fill tool in the GIMP (and may as well have done color too, while I was at it). However, it later occured to me that I would have to use the mouse to draw in the borders for those areas of shading that don't have black-line boundaries (such as shadows on a character's face or other rounded object), and there's little chance I would be able to get properly shaped areas of shading that way. I may just end up using as little in the way of light effects, etc. as possible, which is somewhat unfortunate given the largely lamp-lit setting.
Incidentally, I originally intended to try not to spend money on this project, as I can't receive any significant income from it, but obviously I'm going to need at least a new scanner, as I don't have access to any of reasonable quality. For any reasonable results, I'll also need to spring for actual artists' pens and possibly paper as well. On the topic of paper, there is the question of what size is good. Again, since this will supposedly be a Webcomic, there's no need to worry about the size of the printed page or things getting lost in the bindings, but I do have to worry about convenience of storage (a significant point, since I don't actually have a home of my own). I originally planned to do each frame individually and paste them together in the image editor, like Piro, but I was inspired by Ms. Hayashiya's use of the pictures popping out of the frames in places in HAYATE CROSS BLADE. I imagine this gives the page a more dynamic and organic look, but in any case it's certainly more interesting that way. However, the fact that images can overlap, or even not be in frames in the first place, makes it much more convenient if you draw the entire page together and not on separate sheets. (To make this a bit easier, this book suggests that you can paste on new frames to replace ones that are really messed up, but this would obviously require thicker paper than I use.) I am leaning toward using standard 8½×11 paper (cursèd SI units) so that it will be easy to store them in ordinary file folders without creasing them. However, this may be a problem if the pages are too small to make proper action scenes and panoramic views. A simplistic art style may work all right when drawn small, but there is also the matter of legibility of text. A lot depends on the dpi rating of the hypothetical scanner.
Well, enough waffling about future plans. In the here and now, I need to get drawing. But first, I need breakfast and a shower. . . . [This post was typed by 11:40 today and uploaded at the timestamp above.]