AuthorTopic: 8-bit Lexicon  (Read 1956 times)

Offline lonlaz

  • 0001
  • *
  • Posts: 18
  • Karma: +0/-0
    • View Profile

8-bit Lexicon

on: August 26, 2013, 05:10:02 am
Hello, I'm not a pixel artist, but I've always loved the idea of low-spec pixel art.  I am interested in the conveyance of ideas in the form of symbols, especially with a very minimum of information.  I just so happen to be in a place where I need to tackle a programming project, something interesting and interactive using these ideas.

I'm envisioning creating a crowd sourced lexicon of game tiles.  At first I figured I would start by allowing people to draw and generate noise using a single color from a 16 color palette on an 8x8 grid.  Then that would enter a database and there would be some sort of gameified incentive for people to label those tiles with an idea of what it is they are looking at.

My idea for the end result would be a database containing a dictionary with lexical (idea) relationships between tiles, all based on a lexical dictionary that exists already free to use.  Then I would love to slowly up the resolution and colors and be able to build off of those low spec tiles and just see what develops.

I'm posting because I can have some quite off-the-wall ideas, and I was curious if there would actually be any interest in using such a thing, or accessing a database like that.  Or if anybody had any other angles on it that I might want to consider.

Offline tcaud

  • 0001
  • *
  • Posts: 60
  • Karma: +0/-0
    • View Profile

Re: 8-bit Lexicon

Reply #1 on: August 30, 2013, 10:23:35 pm
Yeah I've got a website I could host it on. Although I think that a general tile/sprite archive would be more useful. A number of sites already do this, but they aren't designed with inspiration in mind. I can see where you're interested in this kinda thing -- how the art "emerges" (the word for it is "clustering") -- ... but what is your particular goal?

Offline YellowLime

  • 0010
  • *
  • Posts: 227
  • Karma: +0/-0
    • View Profile
    • Sour Pixels

Re: 8-bit Lexicon

Reply #2 on: August 31, 2013, 01:38:46 pm
I'm all for the idea! Although I don't think too many people would enjoy tagging 8x8 tiles for the fun of it... (I'm part of the minority :crazy:)

You'd also have to be prepared for differences in opinion. The smaller the resolution, the bigger the abstraction. Maybe a polling system...

After spending some time refreshing my combinatorial mathematics :blind:, I deduced that the number of possible variations is 2 to the 64th (8x8), for binary images (so, times 15 if you consider the rest of the palette).

That number is BIG, so have in mind that your database will not come close to containing all possible combinations (in case that was going to be a milestone before upgrading the resolution) and users will not come across "old" entries by chance.

Since I think peer-review is important (different opinions!), I believe you should set up a "review-mode" (that directs users to old entries) as well if you ever go through with this project ;)

Offline Ai

  • 0100
  • ***
  • Posts: 1057
  • Karma: +2/-0
  • finti
    • finticemo
    • View Profile

Re: 8-bit Lexicon

Reply #3 on: September 01, 2013, 02:34:06 am
^ actually it's a little less than that if you factor out trivial transformations (horizontal / vertical flip, rotate +90 or -90 degrees). Roughly divides it by 8, so, 2^61 for binary images. You can probably also factor out inversions, so 2^60 (roughly 1.1 billion billion (1152921504606846976)). Also, factoring out simple offsets of the same pattern divides by either 49 or 64 (don't remember which calculation is correct), leading to either 2^54.385 or 2^54.

Then much of the remainder is still just bitnoise. A rough test based on requiring the image to have at least 4 2-bit runs (that is, 11 or 00), 3 3-bit runs, 2 4-bit runs, and 1 5bit run, over 10000 random input images, classifies roughly 1864 of them (18%) as interesting.

This would lead to an estimated number of potentially-interesting 1bit 8x8 images equalling ~2^51.576 (3,357,883,882,167,442 or 3.3 million billion.)

For 16-color images, the base number would be 256bit rather than 64bit, leading to an estimated number of potentially-interesting images equalling ~2^244.576
(a 74-digit number). A more sophisticated classifier (based on clusters rather than scanlines) might get that down to 2^240.

If you insist on being pessimistic about your own abilities, consider also being pessimistic about the accuracy of that pessimistic judgement.