GR#015 - Death Walks - Realistic Face

[NaCl]

Pixel art is kind of weird to me, it's somewhere between a style and a medium. You wouldn't ask "is this an oil painting?", but you might ask, "is this cubist?". So the fact that the question, "Is this pixel art?" exists says to me that it is more of a style, though it is tightly coupled with the medium, and gets its name from the medium.

As for this piece of art, I'd say it's a CG drawing that was carefully given a highly efficient and small palette. However I don't think palette size is the best indicator of whether or not something is pixel art. Oekaki is often drawn with very few colors, but I wouldn't consider it pixel art. Pixel art has seemed to always try and transcend whatever hardware limitations existed, and this piece succeeds at transcending the approximate restrictions computer art used to face. So what does that make it?

For inclusive purposes, like posting here or on PJ, I would easily say this is pixel art, but if you want a rigorous semantic discussion about what is pixel art, then I'd say this was probably not pixel art.

[Helm]

Mike: the difference between working from a sketch is that usually in pixel art we use the sketch as reference and then retrace strong pixel lines that simplify the sketchwork back to the familiarspace of how pixel art works. Here I opten to keep the computer-reduction sketch and work directly on it, fix its AA and colors and *then* do traditional pixelling on top.

Accident: what is then the usual charm of pixel art? Because I'm starting to believe that really it has very little to do with *form* and more to do with *content* like we expect from pixel art very period-defined SNES style fantasy Rococo creature sprites and then we extrapolate the technique from there to other contents?

This isn't exactly tracing as I said to Mike above though, that's the interesting part about it. Pixel artists have been tracing from penic (edit, I mean 'pencil' but I won't remove 'penic' because it's a hilarious freudian slip) linearts since the dawn of the form, this isn't exactly that.

Also, fix your Firefox!

NaCl: That's an interesting thought about pixel art as a style and a medium, it's like I said above people think pixel art and expect a certain type of content and aesthetics, they don't just think 'has this been handworked on the pixel level?' and those are concepts I'd like to challenge if not for any other reason than that creating such 'holy laws' especially for people who are inexperienced to the medium will lead to stagnancy. As to 'a CG piece with a efficient palette' you underestimate the amount of straight-up hardcore pixelling that went into it. And I don't say this to say you shouldn't, it's actually interesting that you look at it and you don't immediately notice the pixelling. I'm interested in why this is and how it works.

[Rosse]

Very very interesting topic! There are many things discussed here, not sure if I collected every thought, but here some comments.

First, thanks to NaCl for posing that questions. I think PixelArt can defined as many things, a medium, a style but maybe as a philosophy too. I've read what wikipedia states about media in arts and I've got this:

In the arts, media  (plural of medium) are the materials and techniques used by an artist to produce a work.

Then the example for a drawing:

In drawing, "media" refers to the type of held dry tool used and the base onto which it is transferred. The "held dry tool" normally means a pencil, or stick medium, referred to as a "crayon". Small particles of broken-off stick medium are transferred to a base or plane of production on which the artwork is produced. A typical base is paper, but canvas and other surfaces can also be used.

If I want to do something similar to what PixelArt is to me I would post something like:

In PixelArt, the artist works on a canvas with an ordered raster built upon a specific quantum (undividable part, similar like the old idea of atoms). These quanta (plural of quantum) might have different forms, sizes or colors, but are never allowed to break the ordered raster. The artist doing PixelArt is concious about the ordered raster and the available quanta and tries either to obscure or highlight the ordered raster or a combination of it.

I just made this definition up, so it might have flaws, but let me make some examples:

• First, what we consider normal PixelArt in Computers, for example a Sprite. We have a ordered raster (in modern displays squared pixels [consisting of three subpixels, not important because the subpixels are not conciously altered, except for real subpixel work]) and quantum (palette, for example fixed 16 colors or 8 colors out of (128x128x128) colors) - valid PixelArt concerning my definition
• If we look at indexpainting, we are more on a gray level, but it's not PixelArt in my definition. We have a ordered raster and a given quantum (mostly straight ramps with different hues). But some tools ignore the grid completely and while painting the artist pays mostly no attention to the grid. - NO valid PixelArt concerning my definition
  As you can see, I think the raster is more important in pixelart than the choosing of the specific quantum (in computer pixelart the pixelcolor). I don't have a problem when you use brushes which cycles the index of the colors, because it doesn't break the grid. Other tools like doge&burn, smudge break the grid. Therefore IndexPainting is in most cases no valid PixelArt. Please notice, I only look at the part of IndexPainting, the cleanup part with nice tools is PixelArt!
• HighRes Painting is not PixelArt, even when it has indexed colors / fixed palette, because the painter pays no attention to the grid.
• Making a picture with mosaics is considered pixelart by me, because you have a fixed grid and a specific quantum (the tiles. They might even have different forms like squares, triangles or spheres).
• Making pictures with multicolored stones on the ground is mostly not valid pixelart, because you have no grid you could pay attention you. If you make yourself a grid it becomes like mosaic, a valid pixelart
• Sand mandala has a quantum (grain of sand), but it has no raster.

This is only as far as pixelart is concerned as a medium. In Helm's case, the part about the highres painting and downscaling is obviously not pixelart (I know, nobody stated that) and the last part about the cleanup is valid pixelart. The outcome is therefore valid pixelart (at least at the cleanuped parts). A blank canvas or a random canvas or a downscaled painting is for the definition the same. Every quantum which has been intentionally dealed with is pixelart. In a blank canvas you maybe leave some pixels intentionally white, therefore they are "altered by your mind" - they have content. In a random canvas or downscaled painting it's the same, it's maybe just harder to spot which pixels have "content" or not. Some pixels might even be correct even if they were never altered.

That's about the boring part.



One very interesting thing Helm said was "I'd never have pixelled hair strands like these if I started with a pixel doodle base for example". I claim that if Helm wouldn't have painted this image first in highres, the outcome would be certainly different!
Please let me elaborate: To me, making pictures should be a form of communication, you want so say something. Helm wants to say something, something which has a "infinite resolution" and that picture contains a head with a very faint tilt. In highres he is able to draw that. I just looked through Helm's pixeljoint gallery and I'm sure, if he had pixeled that image from scratch in the native resolution, the wouldn't have that faint tilt, it would be more extreme (only my opinion!).
Well, I can't say if you should have a fixed idea and then choose a medium or workflow or if you should be inspired by the tools you use or a mixture of it and I think in the end it's not even important (as far as you are honestly expressing yourself - bruce lee ).

Please let me make a very unpopular statement: There are not enough original downscaled paintings which then are cleaned up! Why is that so? There's always that talk about Styles, Final Fantasy Sprite, Street Fighter Sprites - there's no content which is communicated, only shadows. I think it's easy to adapt to a style on a pixel level and create nice looking pixelart. That's how the pixelation forum started and how many artists got their job in the gaming biz. There's a reason why there are not enough downscaled and cleanuped paintings and so many "not very good" pixelartists. They lack in a higher resolution idea/picture. That's why it's also important to have strong traditional drawing and painting skills. It's way harder to draw something highres - pixeling is only a handful of techniques and experience. Many might think that downscaling an original painting is like cheating, maybe because it's so rarely seen, maybe because they don't understand it (it looks too CG-like for example) or because they simply are envious (as I said, cheating). But I do think that the workflow of Helm, painting big- downscaling and refining would make more good pixelartists which even satisfies the pixelpurists.

In the end I don't really care. I think it's fair if you have an idea and you use the tools to communicate it the best way possible. Maybe it's a highres painting or a pixeldoodle. In the end, pixelart is more a philosophical thing to me. Creating the most out of the restrictions, limit yourself. I think Helm somewhere made the example of a GO-game. Pixelart can sometimes even be something like meditation.
Just ask yourself (while critisizing), are you ranting about some arbitrary constructs or are you really seeing what some fellow human wanted to express?

[Chis]

Sure it looks like scaled-down cg at first glance, but even without zooming in, if you look carefully you can discern pixels in her hair, in the background, under her chin, etc. And for some (not all) pixel artists that is the intended result - leaving it smooth and polished except for a few hints here and there that it's more than your ordinary drawn picture. I think that although Helm's piece takes this farther than most, the quality is still there. Maybe in that sense it's better that you left it almost-but-not-quite finished.

[Tuna Unleashed]

fuck you helm i love this so much

anyway it took me like 3 seconds to see the pixelartyness in it

[ptoing]

Awesome to see this finally finished. I agree it is probably not the level you could have taken it to, but it is nice nonetheless and I think the pixelart aspect of it is obvious. There is lots of really nice pixel level detail in the face and tasteful dither all around.

Can't find anything I would crit about it really (apart from sloppy aa here and there, but it's a moot point)

[Chris2balls]

I think you've turned a CG sketch into pixel art. I won't reiterate what's been said, but this is to me like any base you'd use for a pixel piece, except that it's an already "finished" piece that's going to be reinterpreted in pixel art format, and not simply some sketches lines, blobs or volumes: you're reinterpreting it, it's a redraft so yes it's pixel art. This work pattern creates a dual vision of the piece, like in digital art, but married with the crispness and "solidity" (I mean by that the pixels are mostly "controlled" in terms of technique) of pixel art.
Thank you for sharing this, and what has been said was very interesting. Bookmarked, and I think I'll try this process myself!

[Helm]

Thank you all for your thoughts. Especially Rosse, you've given me a lot of food for thought that I'm sure will reach productive ends. I will submit this to pixeljoint now.

[Helm]

Actually I want to tackle some of these thoughts if that's alright:

In PixelArt, the artist works on a canvas with an ordered raster built upon a specific quantum (undividable part, similar like the old idea of atoms). These quanta (plural of quantum) might have different forms, sizes or colors, but are never allowed to break the ordered raster. The artist doing PixelArt is conscious about the ordered raster and the available quanta and tries either to obscure or highlight the ordered raster or a combination of it.

I think this is on the right path, though I'd keep talking of pixels as atoms instead of quanta just for readability (the building block of the universe being the atom is still more visible in public consciousness than quantums). The thing we hold from this is that pixel art is concerned with the raster grid and the pixel clusters they place on it and how they interact.

My problems however with your examples are that you take a spite and say de facto 'this is pixel art' although the actual pixel placements on this little 16x16 thing can be pretty senseless. Let's say I take the freehand tool in mspaint and just doodle a few jaggies and then clean them up minimally and slap some colors on and call it pixel art. By your definition is it inescapably pixel art because the grid is so highly visible the artist *had* to deal with it at least minimally. It might not be good pixel art but the principles are in action there. I don't necessarily disagree, I just think it's not a particularily brave definition. It's serviceable as a description of method but not very suggestive of ambition and high-end result. But perhaps that's besides the point.

If we look at indexpainting, we are more on a gray level, but it's not PixelArt in my definition. We have a ordered raster and a given quantum (mostly straight ramps with different hues). But some tools ignore the grid completely and while painting the artist pays mostly no attention to the grid. - NO valid PixelArt concerning my definition

What happens in your rulespace when the artist uses the 'dirty' tool and then spends anywhere from a token to a considerable amount of time going over the choices the automated tool made over the raster grid and fixing according to a micro-cosmic atomic-correlation preference?   This is the matter at hand.

This is only as far as pixelart is concerned as a medium. In Helm's case, the part about the highres painting and downscaling is obviously not pixelart (I know, nobody stated that) and the last part about the cleanup is valid pixelart. The outcome is therefore valid pixelart (at least at the cleanuped parts).

You answer this but:

A blank canvas or a random canvas or a downscaled painting is for the definition the same. Every quantum which has been intentionally dealed with is pixelart. In a blank canvas you maybe leave some pixels intentionally white, therefore they are "altered by your mind" - they have content. In a random canvas or downscaled painting it's the same, it's maybe just harder to spot which pixels have "content" or not. Some pixels might even be correct even if they were never altered.

Yes there were combinations of pixel clusters that I looked at and they seemed fine as they were so I didn't touch them much (though I recolored them) and this will rub pixel artists wrong because I believe the psychoanalytical part of any definition of our medium involves a pride in the ownership of *every* artistic choice made. The avarice of 'complete control'. This is what I believe deserves to be tackled directly.

Do pixel artists think they're more entitled to feeling pride on their piece over say generic CG artists because they've paid more micro-cosmic attention to the building blocks of their image?

Might it be the case that when a pixel artist does that, they're trading in a set of freedoms to gain a different one and the end net result of intentionality and control of the image remains the same across most laboured-on mediums, given that the artists working on them are of equivallent skill and determination?

One very interesting thing Helm said was "I'd never have pixelled hair strands like these if I started with a pixel doodle base for example". I claim that if Helm wouldn't have painted this image first in highres, the outcome would be certainly different!

Exactly. But the pixel artist looking at this would say I (Helm) perhaps do not "own" the headtilt choice in the pixel art piece because it was conceived outside of the psychostructure of what pixel art is. That is to say: once you enter into the pixel art realm, you will fight only using pixels, whatever you did that brought you here before is worthless and your status will only be judged from what you can do with pixels. I find that a very telling characteristic of our cliques and I'm interested in provoking some discussion on it.

f he had pixeled that image from scratch in the native resolution, the wouldn't have that faint tilt, it would be more extreme (only my opinion!).

I'm very certain.

Please let me make a very unpopular statement: There are not enough original downscaled paintings which then are cleaned up! Why is that so? There's always that talk about Styles, Final Fantasy Sprite, Street Fighter Sprites - there's no content which is communicated, only shadows.

I wouldn't say that all those stylistic snes variations of fighter sprites etc are contentless, they are certainly derivative, but why do you say shadows?

I think it's easy to adapt to a style on a pixel level and create nice looking pixelart. That's how the pixelation forum started and how many artists got their job in the gaming biz. There's a reason why there are not enough downscaled and cleanuped paintings and so many "not very good" pixelartists. They lack in a higher resolution idea/picture. That's why it's also important to have strong traditional drawing and painting skills. It's way harder to draw something highres - pixeling is only a handful of techniques and experience.

That's interesting, and what I'm trying to get to as a viable concept: pixel art technique applied to something that comes from a 'longer road' of methodology, not completely unlike how the Demoscene worked but without the ailments of that scene (or any 'scene') namely: dishonesty and misdirection to gain short-term recognition. If there is respect paid to how pixel clusters work and on their relative strengths, shouldn't we be open to work where pixel art was only the final stage of its evolution?

In the end I don't really care. I think it's fair if you have an idea and you use the tools to communicate it the best way possible. Maybe it's a highres painting or a pixeldoodle. In the end, pixelart is more a philosophical thing to me. Creating the most out of the restrictions, limit yourself.

What's fascinating for me for the average pixel artist regurgitating those philosophical bullet points (many of which I've endorsed in Pixelation for what will soon be a decade) is how self-delusional they are. They say 'strive for greatness through restrictions' but they are only talking about the technical restrictions inherent in pushing pixels. When it comes to conceptualizing a larger work and trying to make it fit in those restrictions, that's 'cheating'. This is why I talk about avarice:  most pixel artists are content to make startlingly simple pictures, no, most of the time fragments of pictures and at the same time hold the pride of working on them so intentionally on the pixel level that they think they've elevated the content due to labored upon form.

But perhaps they are right!

I'm very interested in continued discussion on this. I'm sure the pixeljoint submission will gather fascinating anthropology too.

[Ryumaru]

Most likely more than half of these words go straight over my head, but pixel art is a medium, and as such- the process can never dictate whether it is or is not that medium. If you painted over a CG image created by you in oil paint- It would be an oil painting. If you pixeled over a CG image created by you in pixels, it would be pixel art. Trying to involve the process into dictating the medium of pixel art has brought much debate before. One could not say that a piece entirely done with a single pixel brush in an ancient art program is more pixel art than say something I did in microsoft paint, or what you did here. By reducing the amount of colors in your original CG image- you imposed a restriction- and from there you re-worked the image on the single pixel level- even if you did not change every pixel. With your current ability I'm sure you put more thought into every pixel you did or did not place in this image than the average pixel artist. if this piece was done on a metaphorical layer above the reduced CG image, and the sloppier pixel artist did his image on a metaphorical layer of a blank canvas( but really had a mental image imposed onto it) would you say it's not pixel art just because of their lack of ability to use pixel clusters to their full potential?

I do believe you have entered a sketchy realm in pixel art- where the amount of single pixel attention would be the main factor in determining whether it is pixel art or not. Just as a very rough work in progress piece of pixel art may be considered to have parts that are merely oekaki, I would say that if you had done about half work that you did on this piece- it would be the equivalent of a "WIP".

edit: some clarification of above- obviously I mean as long as you are " painting" with that particular medium. if you worked in charcoal and graphite and watercolor over a CG image it would obviously not be an oil painting. An important part to this is that the image was originally created by you, and so you are making the artistic decisions in both of the " layers".