I don’t understand why modern pixel art (outside the demo scene) almost never includes dithering and anti-aliasing. A lot of the creations are artistically impressive, but lack fundamental techniques.
Employing such techniques would make for a wider range of aesthetics and the ability to cover more subjects and a broader emotional range.
I've thought about this before, and the conclusion that I've come to are that dithering and anti-aliasing are both techniques from the wrong eras of pixel art.
Dithering is something that you need the most when you are working with either a very limited or a fixed color palette. On the NES backgrounds were dithered because you had so few colors to work with per tile. EGA era PC games used dithering because the standard 16 color choices weren't very useful. Modern pixel artist don't have these technical limitations to work with. And I think that for most, dithering just doesn't look very appealing vs adding an intermediate color.
Anti-aliasing is on the other side of technology window, from an era where there was enough resolution and color choices to support those techniques, but before 3d really took over. By the time you have that level of rendering capability, your are starting to push into the realm of photo-realistic or cartoon techniques. Late in the Super NES and VGA era the trend was away from hand drawn pixel art, and towards high resolution renders scaled down. And when you're working in that space today, you're starting to lose the appeal that the limitations of pixel art offer.
This is all conjecture though. And I'm sure there are pixel artists out there doing amazing things with dithering and anti-aliasing.
There's a lot of discussion about that in the "classic" pixel art circles. More generally, if you look at it from this angle, pixel art itself is "from the wrong era" of digital art since it is literally self-imposing constraints that maybe used to be there by external imposition but don't need to be anymore.
I've come to the conclusion that whatever we call "pixel art" is just a range of subjective preferences that may be more biased towards aesthetics, or techniques, or tradition, or even "pseudo-tradition". Nowadays the trend seems to be aesthetics with disregard to technique, but there are still many people who consider it "cheating" if you use the "wrong" tools, there are those who want to emulate the exact specs of certain machines, and there are those still who have a mashup of rules that aren't fully justified if not historically (the best example I know is Pixel Joint, but in the end their restrictions have created a community that has flourished and produced awesome art anyway).
To me it’s probably just the multiple meanings that are confusing.
Reading the replies in this thread I realize that to a lot of people it seems to mean a certain aesthetic (I.E. blocky pseudo-retro game graphics). Nothing wrong with that.
I guess I fall into the traditionalist camp. Pixel art to me is about working against fixed limitations to create an illusion of more colours, higher resolution etc.
> almost never includes dithering an anti-aliasing
When you follow the contest link, all three images there on the landing page have dithering, as do the backgrounds and a bunch of examples in the contest rules. Googling pixel art, it’s not a majority that use dithering, but there are lots of examples of dithering. And I’ve just seen a lot of dithering over the years in games and art. So I don’t think your claim about dithering being rare is true.
It is true that anti-aliasing is rare, that’s because it is somewhat antithetical to the intent of pixel art. Anti-aliasing is trying to reduce the effects of visible resolution, where pixel art is intentionally embracing low-res and trying to emphasize the pixels. Anti-aliasing automatically means you’re using a higher resolution and downsampling to a low resolution, where the intent of pixel art is to work at the low resolution in the first place. (Not to mention that anti-aliasing needs continuous color values, where pixel art is often intentionally limited to a small color palette.)
The best pixel art is very creative with the use of pixels and makes different decisions than you would in a higher resolution with anti-aliasing. I think of it a little bit like font ligatures - they exist because they read better and look sharper than antialiasing a higher res version.
For me, anti-aliasing is a fundamental of pixel art precisely for the reasons you say it's not :) The idea of pixel art and the techniques employed is to work _against_ the low resolution, using clever tricks to simulate more detail, higher resolution and more colours.
An example just off the top of my head is this 1988 C64 image by The Sarge[0], using both dithering and anti-aliasing combined with other pixelling techniques to convey depth, texture, extra values and to smooth the edges of a sphere to lessen the garishness of having pixels twice as wide as they are tall.
Your preference is legitimate, and art will always be subjective. I think a big difference is whether you’re trying to show a photo in low res versus do art manually and use pixels as your brush and embracing the limitations. Anti-aliasing is typically a post-process. There are examples out there of careful and manual use of intermediate values in pixel art to make an edge appear smoother, that’s a type of antialiasing. It’s just rare to use more than that and call it pixel art today, because general anti-aliasing makes the image blurry or muddy, and it doesn’t really leverage the pixels, it just works around them.
Good pixel art reminds me of the lego TV families: https://imgur.com/gallery/Cz3zR. It’s such good concept art precisely because it’s as blocky as possible, yet somehow still completely recognizable.
I don’t know if that Sarge cats image is manual anti-aliasing or a more automated process, but for me it doesn’t draw me in or demonstrate a broader emotional range, as you say, compared to some of these examples:
That's the thing, good pixel art should be _as unblocky as possible_. By working against hardware limitations and with the particular properties of a CRT screen, artists are able to create amazing-looking pictures with extremely limited means.
> use pixels as your brush and embracing the limitations
> don’t know if that Sarge cats image is manual
The Sarge's cats are a great example of pixel technique, not composition or emotional depth. It's a sketch, painstakingly drawn by hand (most likely using the keyboard or a joystick), each pixel truly placed and selected individually. It's made using precisely the care for each pixel you talk about, but the care has been taken to hide the inherent blockiness and low colour count.
I suppose it's a generational thing. Having grown up with CRTs and low-res, low-colour machines perhaps creates a different kind of admiration for the immense skill needed to make a lowly C64 or Amiga 500 appear capable of something it's not. Here are a few examples of such graphics from the game Agony from 1992:
> good pixel art should be _as unblocky as possible_. [...] I suppose it’s a generational thing.
Yes, exactly, that might be part of the disconnect here. It used to be the aim to make art unblocky, in the 80s, when there was no choice, when the highest res was 320x200 with 16 colors. Today, there is zero reason for blockiness, ever, you never have to see pixels again. It’s an aesthetic choice to make blocky looking art that is reminiscent of the 80s. And it’s also part of the contest rules for the link you’re commenting on, making it blocky looking is required.
The term pixel art today, unlike 40 years ago, is referring to choosing to be blockier than necessary. If you make something as unblocky as possible, that is now called “art”, not “pixel art”, since it’s trivially easy to get rid of visible pixels and avoid a blocky look. Unblocky is the default, and now it’s harder to be blocky than not.
I do hear what you’re saying - that manual anti-aliasing - is one of the old techniques that made good computer art back in the day. You’re right, it’s true, it just works against today’s pixel art goals.
I may not understand what you mean by dithering. The 2nd picture https://www.lexaloffle.com/media/1/44_m7kenji_2019.jpg has plenty of dithering to give shading effects, so it seems to me. There's lots of 'chessboarding'. Look at the tulips for example, isn't that dithering? Also it's done to soften some edges using white/light grey, though there's less of that that I can see.
That's dithering, absolutely. It fits well with the overall theme of the picture as well, which seems to be some kind of stylized Game Boy vibe. In a lot of new pixel pieces I see, that's about where it stops.
In the hands of a skilled pixel artist it can be used to extend color ramps, add texture, build depth, etc.
In this picture[0], exocet not only uses dithering to really stretch a limited palette of six values for the skin tones, it's even used for little easter eggs (find the heart!). An extreme example, of course, but still highly enjoyable IMO.
Where are you basing these claims? I do pixel art, and dithering + anti-aliasing are one of the fundamental techniques in this medium. It’s widely-employed that there are so many beginner tutorials that teach these methods (I highly recommend MortMort’s channel, for example).
Also, I invite you to look at the Pixel Dailies twitter handle (and the #PixelDailies hashtag), you’ll see the works of several pixel artists and you’ll notice that dithering and anti-aliasing are common patterns.
To me, the main appeal of pixel art is to be able to clearly see the individual pixel boundaries. To a lesser degree, I also like the look created by using a limited palette. Antialiasing makes it harder to tell where one pixel ends and the next begins, and dithering makes it harder to identify the individual colors in an image. While technically impressive, both of those strategies are attempts to imitate traditional mediums instead of embracing pixel art on its own merits.
On the contrary, I'd say they're core concepts of pixel art as a medium. It's not about emulating something else any more than crosshatching with ink is emulating woodcuts. It's about achieving the best results with what you have.
In fact, certain kinds of dithering in combination with PAL colorblending on the C64 will produce completely new color values not available in the fixed palette. Similar tricks were employed with composite CGA output.
Dithering adds texture and makes surfaces look rough. It's fine when applied to a rock or patch of dirt, but it's an amateur mistake to use it everywhere.
Anti-aliasing is used a fair bit in pixel art, but less so in artwork designed to look 8-bit. There weren't enough colors and palette space to anti-alias in older games, so it looks out of place when replicating the look.
When used correctly, dithering can be a work of art itself. Pretty much all the artists I know who work on actual 8-bit platforms use both dithering and AA extensively, hardware permitting.
To clarify by example, consider this picture by ptoing[0]. It employs various dithering techniques and AA in the right places to achieve amazing results with a fixed palette on an 8-bit system. The individual pixels can still be discerned but the dithering adds both depth and texture and the AA adds a sense of smoothness.
I see dithering used plenty. Keep in mind that pixel art was displayed at native or close to native resolution and blurred on CRT displays back in the day. Now it’s displayed at 10X resolution on sharp LCD or OLED screens. As another comment mentioned, it adds texture.
Pixelart means that ever pixel is chosen individually and specifically, over any other pixel. If you palette is 16 million colours, this becomes impractical. In a given situation, which of #e6fc04 and #e7fc04 is the better choice? Is there even a difference? So you have to have a constrained palette, which more-or-less precludes any sort of antialiasing.
For my part, I have only ever seen one piece of antialised pixelart: the chain holding the elevator in the video game celeste. It looked awful.
> you have to have a constrained palette, which more-or-less precludes any sort of antialiasing
> I have only ever seen one piece of antialised pixelart
You haven't seen a lot of pixel art then. Just visit Pixel Joint [1] (a well-known community) and on the homepage (which is updated all the time) you'll already see a few. One of the most distinguished members even has a popular "tutorial" [2] that discusses this and other techniques.
If you're interested, there's also Pixelation [3], a community where people discuss techniques in depth.
In short, anti-aliasing can be done by hand and using a low color count. There's also "index painting", something that many "hardcore" pixel artists consider cheating, but which creates the style without manual work [4].
The point others make is true, though: The sharpness of newer screens ultimately reduces their visual appeal (although there are people who actually prefer them).
I'll buck the apologist trend here and mention that artists by and large are not technical people and are only concerned with the superficial pleasant appearance of pixel art, and not a faithful reproduction even assuming they were capable of understanding all of the optimization tricks that made pixel art technically beautiful.
Another example of laymen latching onto and ruining a "nice thing".
These techniques greatly reduce the sharpness of the resulting art. Anti-aliasing also depends on the resolution, making it impossible to scale the result without degrading quality. For example, many old mobile video games had anti-aliased fonts which looked fine on small screens but became almost unreadable when scaled to modern screen resolutions.
It's fascinating that some images borrow the style of character displays. I imagine the addition of PETSCII, ATASCII, and other mosaic characters Unicode 13 will help to inspire more of their usage.
I can't wait for the MZ-80 characters we are trying to include in Unicode 14.
It's an art contest overall, but TFA is about a video game category of that category. The game category is new this year and sponsored by TFA/the submitter; the contest (pixel-art.jp) is a larger, pre-existing thing.
Japan highly encourages children, starting in middle school, to join clubs; illustration clubs are common, and kids who like to draw will tend to join those. Children are not, on average, very creative (surprisingly!), so when told to create a comic book for such a club, they'll usually create a story with characters resembling themselves/their friends, and a setting resembling a school.
Thus, most adults with illustrative talent in Japan, have usually coincidentally had a lot of formative practice drawing children in school uniforms, starting back when they were children in school uniforms.
I think there might be something deeper to it. A lot of manga and anime take place in school as well, and I'd be surprised if it's just because all the artists know how to draw is schoolchildren. This is even more surprising because it seems like to me that the majority of western comics and tv shows don't take place in school settings.
I expect that this is the sort of thing that only seems weird to outsiders like me. Maybe people in Japan look at the US and wonder why we have so much media about superheros, and so little about highschool, even though highschool is something almost everyone has experienced and has powerful memories of.
> I think there might be something deeper to it. A lot of manga and anime take place in school as well, and I'd be surprised if it's just because all the artists know how to draw is schoolchildren.
That's for the simple reason that this is their main target-group. Anime is for kids, because adults don't have time to waste for them. Adults read manga, on train or break-time. And many manga have mature characters and topics.
Of course are there are also manga aiming for kids and anime targeting adults, and in the last decades things have changed, because technology as also society changing. Also today there are strong tropes, sophisticates nerds, etc.
> A lot of manga and anime take place in school as well, and I'd be surprised if it's just because all the artists know how to draw is schoolchildren.
Besides what the sibling poster said about target audience—these days, a lot of it's down to the fact that many manga/anime are being adapted from commercially-published(!) novels written by high-schoolers.
Yes, Japanese book publishers are signing deals with teenagers these days. I think with the justification that they can better write things their peers want to read. I've got to believe they have sales numbers to back that up.
But high-schoolers are still uncreative fiction-writers, writing mostly high-schooler characters in high-school-equivalent settings. I guess the publishers don't complain as long as the books sell.
---
Though, honestly, what's really going on here is that everyone is mostly uncreative when writing voluntarily under no particular artistic constraints. Tangential Sturgeon's Law rant about fiction-writing (from the perspective of an editor reading through a slush-pile) follows:
People mistake the advice of "write what you know"[1] for a recommendation to just "write your own life-story, gussied up a bit."
For every one amateur writer who constructs a novel fictional setting, there will be nine writers who will just pastiche together the last three books they read; and 90 more who will won't bother with a novel setting at all.
For every amateur writer who constructs an interesting character "from whole cloth", there will be nine that just re-skin characters from the stories they like, and 90 that just attempt to recapitulate the mostly-mundane people in their own lives.
For each amateur writer who starts with the intent to write a novel conflict/scenario, letting the work define its own genre; there are 999 writers who will instead choose a genre to work in, and then regurgitate exactly the same scenario tropes they've already seen from works in the genre. (Many authors seem to think that scenario tropes are what genre is.)
Professional authors aren't any better on their own, but usually they'll either have a publisher or editor who will push them to work outside their comfort zone; or they'll have written the same thing enough times that they'll finally get bored of it, and enter a more experimental phase.
[1] If you're curious, "write what you know" is supposed to mean "if you want to include X in a story, then go research+experience X. Embed yourself in the world of X, like a gonzo journalist. If you want to write a story focused on a real-world place, you've got to live there for a while; if you want to write a story focused on a certain industry, you've got to have actually worked in that industry for a few years." It's not supposed to hold you back from writing what you don't know; it's supposed to push you to go know it!
Children are young, innocent and widely appealing. School uniforms invoke a sense of nostalgia and make people living in Japan's harsh working culture think of their much happier high school days. Plus, much of a child's life will be spent wearing one so if you're depicting a child it's a realistic choice of clothing.
As for why girls over boys, I think it's that cute and innocent things are appealling and more people find girls cute and innocent than find boys cute and innocent, even the artists themselves, regardless of gender. Many artists have mentioned that they find girls easier to draw than boys as well.
There's also the sexual aspect but though I think such media is more common in Japan than elsewhere, I don't believe it's anywhere near mainstream.
I'm not sure what you mean by "highlight", but I remember reading somewhere that Japanese people are usually very nostalgic of their school years, though I don't remember if there was an explanation given for it.
My personal opinion is that the Japanese society is very demanding for adults, and people recall these years as less burdensome. Perhaps this is the same everywhere?
Employing such techniques would make for a wider range of aesthetics and the ability to cover more subjects and a broader emotional range.