Why is pixel art so HAAAARD?

Why is pixel art so HAAAARD?

...I mean, apart from the fact that you're trying to cram a whole person into a 16x32 space.

I don't know whether this was hard or not, but you did a great job.

This looks OK. A bit on the safe side, since the "black outlines with solid fills" is one of the most straightforward techniques in pixel art, but it's a valid style that's been used in several commercial games.

The shading on the skin looks fairly pointless though... it's too subtle to make any real difference, so it's basically wasting a sprite color (assuming your working with a limited color count).

If you do pixel art long enough you end up learning other techniques that'll allow to create more unique graphics, but what you have now is adequate if you're going for a cartoonish look.

Are you sharing your work, asking for critique, or searching for commiseration? What kind of game is this character in? Is this for a JRPG battle screen, a side scrolling platformer, a top-down dungeon crawler?

This image doesn't appear to observe the NES pallet or limitations.

I find it very usefully to sketch out ideas in full screen mock-ups so you can see how all the different elements relate to each other. It also helps generate more constructive feedback.

Good luck.

It's a shit ton harder when you have to start dealing with 48x96 and whatnot spaces...

True. Less space to fill = fewer possible ways to do it.

Like any craft, it takes years of practice getting good at drawing; pixels or not. Best tip i've got is to roll back to assuming no prior knowledge, go to an arts & crafts store, buy a good looking text book on how to draw, and practice through all the examples, whether they seem interesting/fun/relevant or not.

True. I revised the colors a bit and came up with this.





Kind of all of it?
Mostly I wanted to talk about the challenge of making pixel art within such tight restrictions.
I'll admit, I probably shouldn't have been posting this late at night. I tend make poor decisions when I'm tired.



True, it isn't the correct NES palette. And I was trying to go for six colors, which of course you can do on the NES, it's just wasteful.
I'm trying to break beyond the NES limits a little bit to see what I can do. The last time I was trying to make NES character art, bumping up to two palettes helped me figure out better what I was trying to do before condensing it down to one.
I also might change gears to a GBC game. Still undecided on doing that, but right now I'm experimenting.



Good call.



In my experience, I've found that's not quite true. Drawing on paper and drawing in pixels are different talents. There's some carryover, but really only to beginners.

One thing I believe is that there is really only one skill in making art, and that is being able to properly visualize what you are trying to create. And a possible secondary skill (depending on how you phrase it) is being able to identify how your work is failing to meet that intended goal. Every other skill is only a matter of being able to produce your work faster and/or in fewer iterations.
Now practicing drawing (like on paper) can help you tap into that base skill of understanding what you are trying to make, and certainly it would be important for a beginner. But it only can get a person so far. It can help you understand what you are trying to draw; help you understand what kind of style you prefer. But once you've reach that point, there is nothing else that traditional drawing is going to help you with in creating pixel art. The art of conveying the small details is completely and utterly different between the two. You have to approach how you are thinking about your art from a completely different angle; you need to understand pixels, not drawing. Effectively, drawing helps you understand yourself, but beyond that, it becomes a skill of understanding how the pixel art works with itself. The art you are visualizing as a drawing is different from the art you are visualizing as a sprite.


I completely disagree with that. With more space you have more ability to express yourself. With larger canvases, the argument I made above becomes less true because the medium begins to get closer to a traditional medium. In fact, all sorts of different mediums have their skills more pertinent as you have larger canvases. (Some games made their sprites from clay models or miniatures.) You can explore more styles and more techniques, and small errors can become more concealed and swallowed up in the rest of the work.

But under these tighter restrictions, getting something that looks good and skillful is much harder; there is FAR less room for errors.

pixels and pens are just different media expressing the same ideal concept: capture something (from percieved reality, imagination, or both) on a two-dimensional, graphic plane, with the help of manual labour and creative thinking. Both are very uniquely formative, of course, but still.

Besides, there's nothing stopping anyone from grabbing a book of exercises (or theoretical art studies) and apply that to pixels directly. I'm not saying "before you can master the pixel, you must master the brush", i'm rather saying that courses and tutorials on pixel art tends to be very specific, narrow, and convention/genre-dependent, and that it's easy to miss out on great tips, tricks, techniques and studies in neighboring media. Sometimes using pen, brush or some other media will still inform what you do with pixels (i guarantee it), but that wasn't the really my. It was more to do universally useful studies. Like depicting a draped piece of cloth in directional spotlight to get better at doing shadows, to observing video captures in slow motion how different people walk differently. I promise it's not just good for beginners. The key to getting good is doing a lot and observe anew. Also helps to try and vary what you're doing, and how, which helps inform your main techniques and push propel them forward. At least that's my experience.

Honestly it's not too far fetched to practice wholly different artistic practices either. Writing a few short stories or attending amateur theatre will probably improve embedding narrative tensions when drawing, too. That's a bit on the holistic side, but i think it has helped me.

Anyway it's just advise. You asked why pixel art is hard. It's kind of hard to answer since difficulty is very personal, but a hopefully comforting aspect of it is that it (pixel art) can also be quite forgiving in its expression. Because of the ambigous nature of large pixels (or woodcuts, or pointillized pictures, or minimalist paintings) the mind fills in the blanks and we tend to see what we're guided to see (helped by semiotic systems)*. Meaning the raster is not just a limitation, it's also an opportunity.

In contrast to naturalist art, pixel art is then more reliant on symbolism. Learning how to manipulate all the minute signs embedded within the picture is a key skill to practice, i think.

Right, that is very true. I guess I misinterpreted what you were describing.
I've been seeing this commercial play before all my YouTube videos that's been talking some drawing course and talking about how important drawing is. I've been growing sick of that mentality and of a lot of tutorials in general, because over the past few years I've learned so much about the limits of art and skill and of what can be taught. When I saw you talking about drawing I kind of glazed over and thought you were barking up that tree and saying that just by drawing you're going to be good at all art.

Yes, there is a lot that you gain from studying everything you can. Study painting and sculpting and it's going to help with all your other art. Study pixel art; it's going to help with your drawing.
But there are two factors above all that dictate the quality of your art. One is the time and practice (experience) that you put in, and the other is an ineffable skill to properly visualize what you are trying to accomplish. And they work out as an equation, one multiplied by the other.

And I'm just getting really frustrated at my second part of that equation when it comes to pixel art.
This is something I love; this is something I want to do. I want to make games, and I love pixel art so very much. I can make some decent pixel art environments. And I've managed to draw some good looking characters. But it seems that I have abysmal talent for making pixel art characters.
I spent a couple hours today drawing out different versions for the character I'm trying to create; I drew them by hand onto graph paper. And you know what those drawings look pretty nice. (I mean they're just sketches, but it's like you talked about: your brain fills in the details.) But then I tried to transfer these over to pixel art. I knew they wouldn't come out exactly right, and so they would need to be cleaned up a bit. But these don't need to be cleaned up, they need to be taken out back and shot.

And anything multiplied by zero, is zero.

Being a coder rather than an artist, I tried to learn pixel art from the other direction. I looked at sprites by other people, identified shapes and patterns, and mashed the results together by counting out pixels. It's a matter of combinatorial logic: at a low enough resolution and color count, there are only so many ways you can represent any given subject. Look at other people's solutions, maybe brute-force a few permutations of them, and build your own library of patterns.

Probably not the sort of mentality an artist would appreciate, but it might help...?

Pixel art at such low resolutions differ from hand drawn art in the sense that you don't need dexterity to draw it. What you need the most is art direction skills: you need to know what you want and you need to know what makes the art look bad when the art looks bad. Then you can iterate moving some pixels around until you get the results.

But don't be fooled, the fundamentals of art are there: you need shape, color, balance and clarity. Visual arts are about communication. That means it needs to be direct, concise and confident. No room for undecided elements. Especially on the NES with such limited resources, achieving good results with few colors is not only elegant, it frees those resources for use elsewhere on the picture, vastly improving the overall sense of beauty and optimization.

Exactly.
You could have the most natural talent in the world, but if you put in literally no time to create something, you will have literally nothing to show for it.
Moreover, if you have a skill that is say 0.03, you're going to have to put in a LOT more time to produce something than someone with more natural talent.


That's largely how I've been trying to create my art. I study a lot of other examples and try to extrapolate tricks and techniques. But as you also said, there are fewer permutations that can correctly demonstrate a character. And I struggle to develop a character that doesn't look too obviously like I just stole it form something else. By which I mean, the legs might be exactly copied over from another sprite. That's not acceptable; that's theft. (Okay, it's fine for certain kinds of projects. Plenty of people make ROM hacks and tributes, and those are fine. But that's not what I'm going for; I want something that can stand on its own.)

Nesrocks is absolutely right. That said, if you're losing something transitioning from gralh paper to screen, you could try some on my custom graph paper. It is divided lightly down to the pixel. I often use it, especially when sketching out keyframes for the first time.