AuthorTopic: Pixel Art 3.0  (Read 9081 times)

Offline RAV

  • 0010
  • *
  • Posts: 293
  • Karma: +0/-0
    • View Profile
    • Blackbox Voxel Tool

Pixel Art 3.0

on: March 28, 2015, 01:26:40 pm
"provocative". "controversial". Let's roll the ball and keep discussion productive.
As some of you know, I have been talking about this in the CB on PJ for quite some time now. And this is an attempt to move the topic over here.



First, take the time to observe. You might still remember my other videos from the past.

The last video, called Basement, I kinda forgot posting here. It is to be considered largely obsolete now; what you saw doing there would look very different now in this newest version:

The first few minutes showcase the improved Smart Select. It gets more meaty afterwards. it's half an hour long, try to give it the time to understand what we're talking about.


However, unfortunately, I could no other than produce this video with an excruciatingly bad video capture setup; this video begins with my earlier setup, but from mark 6:45 turns into my current bad, so you notice a terrible drop from there. This tool by itself still runs smooth as the earlier video. I had a similar problem way back in the past, and the difference between this bad video and the good reality of the tool running on even lower-end hardware, is the same as what you saw between the old "Blackbox" and "Colorado" videos in my youtube gallery. Painful as that is, since I am not sure when I can get better video capture again, I thought I might as well just jump in the cold water now, and start discussing the ideas behind the project. For the sake of discussion, try to concentrate on the meat of the proposed ideas for now, and just assume that this technically very well works actually.




We try to maximize distinctive differences of pixel art to other digital art, regardless of past limitations.
And finding solutions that are within this spirit to the art, that help define identity of the art, instead of confusing it.



There is no difference between a pixel or sprite or tile or map.
It can be considered all the same on a technical level.
Does this sound like confusing the matter? I believe these distinctions confuse the matter.
And this has consequences to the workflow; and the question of what is pixel art.


And as such, you could see here, there is no difference between colour-picking and tile-picking, it just depends what your current selection encompasses. And neither makes it a difference what size your brush has or what form it has, like that of a tile for example. And that's also how the 45 blocks will make a comeback as basis of an advanced tileset logic, as much as pixel logic--if you so please. Because it all makes no longer a difference, it's unified, while keeping the spirit of pixelart intact, and even pronounce the core distinctiveness of this artform, regardless technical limitations of a time. This is pixel art in its most distinctive form as a process of art compared to other digital art, and arbitrary limitations are not necessary to force a defensive definition of this art form.


Pixel art is about combinatory logic in a grid space.
limitations beyond that definition are just funny context to it. modes of the game. preferences.
each interesting and fulfilling in its own right. Even the question of 2d or 3d becomes just another context to it.
CG workflow that is less about combinatory problem in grid space, is less about pixel art;
-- the more strictly it is about combinatory problem in grid space, it transcendents contextual limits.



Snapping the brush of any size into a resolution grid, reinforces the identity of what is pixel art. That does not mean doing it different is not pixel art or bad pixel art. We must go away from taking it the wrong way. It means we have a strong point of orientation, a northern star, within the identity space of what is pixel art. And by that you are able to navigate your own way. By that we are able to conceptually grasp and develop pixel art, we are able to translate into other territory, be it platform or style. We are even better able to defy and spite it. And whether you accomplish making it look like painterly or realism or whatever, that's great, that's a praiseworthy accomplishment on its own, especially if you manage to do that within this strict logic of the art, especially though it might not be the best suited for it.



Tiled based workflow is not necessarily meant for just producing tiled looking works. First and foremost it helps you build substance in the starting phase. you quickly build critical mass that you can customize, mold and tune into whatever unique vision you have. The dynamic fractal-tile approach turns pixel art into its own creative advantage. And there is a large potential of skill and creativity in fully utilizing this.



It seems to be a common verdict here that mixed resolution is bad. Many of these sort of statements look at the issue from a stylistic/artistic point of view on a given work, as whether it is obviously mixed resolution. However, if you look at it from a technical perspective on the fundamental logic of the medium, you come to another conclusion: Almost every pixel art actually already is mixed resolution, whether it is intended or not. Whenever there are bigger clumps of clusters, there is a good chance this area could be technically described the same with bigger pixels of lower resolution. It's just non-obvious at first sight without dynamic adaptive grid visualization highlighting this abstractive fact.

Realizing this, with proper technical support, you have a much better orientation in canvas space. The dimensions and proportions of areas are much easier to eyeball on the fly in the planning phase, since instead of judging big confusing clusters of countless little pixels at same size, you intuitively compare their simplified definition as mixed resolution behind their seemingly continuous space: instead of 194 pixels of one area confusingly compared to the 295 pixels of the other, you may look at this same area as comprised of simpler cluster forms with 3 very big pixels, 4 smaller pixels and 5 very small pixels, of the same colour, attached to each other, clearly comparable to the other area made of 5 very big pixels, 2 smaller pixels and 4 very small pixels.

In the result this doesn't matter, without the grid overlay it's just continuous space in either case. but in the process it is quite convenient more than might be expected.


That voxels can have unique sides as much as a wholesome colour. makes for an interesting mechanic.
You can shade an object by wholesome voxels, or shade it by-side of voxel. You can and will even mix the two techniques within the same image. wholesome voxel colourization helps hide the cubic form, it becomes a blob of colour, instead of a volume-object by itself. This keeps the viewers attention on the overall form of the object that this voxel is part of; it suppresses corner noise that distracts from the actual object that the cubes combined are meant to describe. However, by-side colourization is just as useful, as there are just as many situations in which the emphasized "edgy-ness" is important part of overall definition of object. You want both. you have both.

This project is no "2d versus 3d", it merges both realms seamless, and what perspective is chosen, is decided on the fly; it is possible to just ignore 3d, make classic 2d sprites on the wall and convert it directly into bitmap, voxel for pixels on assumed full frontal. There really is no downside to this, and it can operate in a mode such, you wouldn't even realize you are not using classic "MS paint" instead of staring at a wall straight, within 3d space. You can easily ignore that, it's just an option that 3d is always just around the corner. Whether it has all the features you expect from mature 2d tools, is just a matter of effort and time.

However, I consider the development of this tool as something else than piling up all kinds of features and options. The goal is to produce a tool that is very directed in its design; that the usability design itself helps strongly define the core identity of the art, and serves as basis, as a point of orientation, like pixel art is in itself. A dedicated design towards the lean spirit of the basic combinatory problems in grid space. and that most of the effective workflows can boil down to surprisingly little but versatile functionality, that is playfully re-purposed towards the tactile situation, and thus highlights the spirit of the art instead of detracting from it.


« Last Edit: March 29, 2015, 02:43:25 am by RAV »

Offline astraldata

  • 0010
  • *
  • Posts: 391
  • Karma: +1/-0
    • View Profile
    • MUGEN ZERO

Re: Pixel Art 3.0

Reply #1 on: March 30, 2015, 08:15:52 pm
I think it's a neat tool.

Unfortunately, I also think you might be greatly exaggerating it as a new form of pixel art. Why call it "Pixel Art 3.0" when this is clearly its own thing?

Why not just call it resolution-independent voxels? Or just something else? It's clear you don't understand what "pixel art" really is, otherwise you wouldn't feel the need to justify your system as being called such. Pixel art is apples and oranges in the case of comparing it to your system.

Pixel art utilizes a lot of techniques that rely on a single-sized grid square to create the illusion of a higher-resolution image than is actually the case. Look at any NES or SNES sprites for good examples of this concept.

Pixel art is based around the concept of a "pixel" -- the fundamental unit of measurement (and also the most atomic building block) in pixel art.

Your system, on the other hand, enforces no sense of that fundamental unit of measurement. At best, it offers a way to place many of these down at once (and in 3d), which /could/ still be considered a form of "pixel art" (in my opinion at least, going on the fact that 'pixel art' can generally be created using other mediums such as beads, thread, legos, or any other differently-colored same-sized units) if it weren't for just the one simple fact that, in your system, these fundamental rules are designed to be broken (unlike when using lego cubes for pixel art, for example).

That being said, even if we were to stretch the definition of what we consider to be pixel art to include your system, none of the rules of how to handle color, intensity, shapes, clusters, detail, or depth apply -- which is, fundamentally, what any skilled pixel-artist sets out to master.

On the other hand, with your system, these things are handled /for/ you by the tool, taking away the need to learn any of these skills to render your image/scene/whatever and are, ultimately, no different from using a photoshop brush to automatically AA your lines (rather than learn how to do it manually). In your system's case, I'd let its 3d and automatic lighting and shadowing to manage my depth for me (rather than color and value selections done by hand), and differently-sized grids to manage my need to explicitly define detail (rather than subpixel 'blending' of colors in my mind to arrive at the correct color and contrast combinations to hint at details rather than explicitly define them, which is impossible in all forms of pixel art except for very high-resolutions in which case you should just be using digital painting tools instead of individual pixels anyway, since per-pixel techniques apply less and less at higher resolutions, thus reducing the ability to even call it 'pixel-art', which, in your case, is why I suggested calling your tool something else.)

All in all, it's a nice resolution-independent almost-voxel/cube-editing tool, and I would suggest working on it more -- after all, people play minecraft for the same reason one might want to use your tool -- but, in the end, I wouldn't call it pixel-art at all, nor would I consider it any sort of an 'evolution' of pixel art (or even voxel art for that matter) either.

It's its own thing. Celebrate its uniqueness rather than trying to label it as something it's not.
I'm offering free pixel-art mentorship for promising pixel artists. For details, click here.

     http://mugenzero.userboard.net/

Offline NowvaB

  • 0001
  • *
  • Posts: 93
  • Karma: +0/-0
    • View Profile
    • Tumblr

Re: Pixel Art 3.0

Reply #2 on: March 30, 2015, 09:11:12 pm
I'm sorry if I screw something up here because I only read two paragraphs before my ADD took advantage of me. A TL;DR perhaps?

Pixel Art 3.0? When was 2.0? :o

I there some efficiency between being able to have multiple resolutions of voxels? because I can kinda see how 1 voxel is better than 8 if they take up the same area.

but on the other hand saying that this is pixel art means that you are saying this is okay vvv



I guess that this is still considered pixel art but it's bad practice.

Offline RAV

  • 0010
  • *
  • Posts: 293
  • Karma: +0/-0
    • View Profile
    • Blackbox Voxel Tool

Re: Pixel Art 3.0

Reply #3 on: March 30, 2015, 10:09:01 pm
I hear that a lot, that I don't understand things. :)

Offline PixelPiledriver

  • 0011
  • **
  • Posts: 997
  • Karma: +6/-0
  • Yo!
    • View Profile
    • My Blog

Re: Pixel Art 3.0

Reply #4 on: March 30, 2015, 11:01:52 pm
RAV has sent me a few versions of this tool.
Haven't had much time to play with it.
But my thoughts were:
- In general the tool is super neat and has a lot of potential  :y:
- Being able to break down and rebuild blocks feels great
- I really like the controls
- Each version he sent had good small improvements, some based directly on notes I gave
- Not being able to save is a bummer and stops it from being something that I want to mess with on a daily/weekly basis
- Widespread testing and a public log of bugs and wishlist features would do it a lot of good.

I love the stuff towards the end of the video with the stamping.
That feature wasn't available when I tried it last.
Feel free to continue sending me new versions.
Testing is fun and I'll keep doing small bits of it if you find it valuable.

As for the theory and stuff.
It's not bad practice to mix resolution, it just needs to be organized in some way.
3d textures have plenty of difference in size of shared unwrap space.
Using it to imply scale, perspective and or focus can also work well.
And knowing that it is, we seek what it is... ~ Aristotle, Posterior Analytics, Chapter 1

Offline cels

  • 0010
  • *
  • Posts: 379
  • Karma: +1/-0
    • http://pixeljoint.com/p/32715.htm
    • View Profile

Re: Pixel Art 3.0

Reply #5 on: March 30, 2015, 11:50:50 pm
Have to agree with astraldata. I'm more inclined to call this Lego 3.0 instead of Pixel Art 3.0.

Seeing is believing though. So far I've only heard a lot of eloquent explanations and deep thoughts, without actually seeing something that would warrant the description of pixel art. I know RAV doesn't like to share his actual artwork, but maybe someone else can use this tool to make something.

So in the name of being constructive... let's see some art.

Offline RAV

  • 0010
  • *
  • Posts: 293
  • Karma: +0/-0
    • View Profile
    • Blackbox Voxel Tool

Re: Pixel Art 3.0

Reply #6 on: March 31, 2015, 12:10:57 am
Yeah, yeah, you suckered me into this on Valentine's day, PPD. Took advantage of my drunk and labile emotions! :p

- Each version he sent had good small improvements, some based directly on notes I gave
- Widespread testing and a public log of bugs and wishlist features would do it a lot of good.
Those hotfixes to controls were temporary, since you got unfinished versions with critical features missing. For example, now that I got Smart Select done, the game has changed a whole lot, and walk/run toggle or hide-gui toggle are much less needed, and may be scrapped for more useful functionality. It's why I was so apprehensive at first releasing any version at all, since then I'd be pressured to hotfix little things that should be handled in much more fundamental manner as planned. It takes away dev time from the real meat that I think I should rather work on still at this point. And as long as I feel major game-changing features of my list are not done, the intermediate testing, feedback and fixing are not very effective development, we're doctoring around on surface problems. Even just preparing temporary releases for fun takes out time. It also makes development inefficient in so far, as the most reasonable order of coding things, is not the most desirable order for the user. An early Save isn't good while I'm still messing with the baseline, and I don't want a mess of file versions to unnecessarily confuse development further this early, yet of course that frustrates users.

As for the critiques in this topic so far, I must admit it's frustrating me a little. I thought, with all we've been through and the videos, we were beyond that level of debate. It's kinda why I avoided further discussions since a year ago, and rather worked on creating facts that speak for themselves. Yet discussion might just be still too inefficient even at this point in time. I'm not sure the misunderstandings can be resolved in just talk, since I'm not sure how to explain more clearly. Needless to say: any pixel art that can be done now, can be done in this tool. Pixel art as you know it, is a subset in it by choice. It should be obvious why. Problem is that I've been working on it alone for so long, it has been such a dramatic journey, so much in it happened, and I got so used to it all now, while not giving you the opportunity to getting used to it, and now I'd like to speak about it on a level that naturally no one else can really follow, because too much of my perspective I'm taking for granted. *sigh*



« Last Edit: March 31, 2015, 01:00:12 am by RAV »

Offline Indigo

  • Administrator
  • 0011
  • *
  • Posts: 944
  • Karma: +0/-0
  • Artist, Indie Game Dev
    • DanFessler
    • DanFessler
    • http://pixeljoint.com/p/849.htm
    • DanFessler
    • DanFessler
    • View Profile
    • Portfolio

Re: Pixel Art 3.0

Reply #7 on: March 31, 2015, 01:16:36 am
Can we go back to the question of what Pixel Art 2.0 is again?

Offline RAV

  • 0010
  • *
  • Posts: 293
  • Karma: +0/-0
    • View Profile
    • Blackbox Voxel Tool

Re: Pixel Art 3.0

Reply #8 on: March 31, 2015, 01:30:42 am
Well I suspected this would very much irk you guys, that's half the fun. ;p

Aside from that it is more or less a pun, it is a very rough segmentation at my convenience:

Pixel Art 1.0:
pre PixelJoint & Pixelation era of pixel art, CRT and non-PC and early PC based.

Pixel Art 2.0:
the PJ and Pix era itself, with definitions to pixel art to artificially confine it within computer graphics, in ways noone cared before, because everything was considered pixel art, if it was even consciously considered as such. entering LCD and modern computer architecture. with according changes in preferred pixel art techniques.

Pixel Art 3.0:
Full virtualization and abstraction of pixel art. seamless merge of workflows for both 2d and 3d. platform and hardware independent rendering. meaning of workflow over technical specifications.


Sure we can now obsess about the merit of this wording, and I brought it upon myself I guess. It is a categorization that for the context of my own technical work makes convenient sense to me, even though you might not like it as best and full description of what happened to pixel art in all this time as it concerns you. I am not really interested in making this little thing a major point of contention overshadowing the main menu.


« Last Edit: March 31, 2015, 01:51:12 am by RAV »

Offline CelioHogane

  • 0010
  • *
  • Posts: 108
  • Karma: +0/-0
    • View Profile

Re: Pixel Art 3.0

Reply #9 on: March 31, 2015, 03:01:05 am
So basically we are calling the voxel art pixel art 3.0 because we are hipster? A little lame.