The Pixel Aesthetic

[AlexHW]

The Pixel Aesthetic




I took some time to put up some writing I did a while back based on a personal study of the pixel. The intent was to explore pixel aesthetics in a black and white sort of way- making clear the differences between specific aspects related to the pixel and how they might interact with each other.

I'm not sure what others will think, which is why I've put it up. It might help me going forward if I get some feedback on it.

---
Working on some visualizers to help illustrate and provide more context.
Select an image to see the linking/binding of the pixels. Requires webgl enabled in your browser.
http://alexhw.com/pav/index.html

---

[Atnas]

Repeating a few thoughts on terminology. Please consider:

If you're going to make gem analogies, Face and Facet feel reversed. Face is a flat shape with sides. Facet is "one side of something many-sided" by definition.

Less important, but Laced does not convey as much meaning as Interlaced could. Interlaced is more descriptive because it implies there is more than one influence on the pixel - being bound and linked.

A lot of thought provoking perspectives, regardless.

[AlexHW]

I'll spend more time thinking about those, Atnas.
There are various reasons why I chose those- I need to make those reasons more clear I suppose.
I've started working on a visualization tool to help generate more context and bring a sense of application to these concepts.
http://alexhw.com/pav/index.html
Select an image to see the linking/binding of the pixels. Requires webgl enabled in your browser.

It's interesting seeing the different dynamics of various sized images/details, etc.
http://i.imgur.com/dCfv929.png

[0xDB]

I am wondering about the practical usefulness of those concepts of "linking/binding". It seems to me the output of the visualization of those things is not recognizable as anything, so why even think about it? Or would you say the lack of the visualization looking like anything is a hint that the arrangement of pixels in the source image is flawed and does not do a good job at making use of the expressive possibilities in the arrangement of pixels / hints towards poor aesthetic choices?

[ptoing]

Looks to me like some really grainy edge detection filter.

[AlexHW]

It's possible the technique involved with visualizing the linking/binding could be improved, 0xDB. That is something I will be experimenting with. I also plan to add other data and controls to the visualizer.
You could be right in that certain patterns may correspond to certain expressive qualities. Where it looks jumbled may hint towards more focused connections, whereas less jumbled areas hint at less focused connections (or maybe the other way around?). It's open to investigation as to the applications, etc. Perhaps such things will be clearer as the visualizer becomes more developed.

[AlexHW]

Had a thought about banding, and wanted to analyze it a bit.
Left side of image illustrates the issue with banding pretty well.

[Ai]

Just took the time to read through this. Some very interesting thoughts. I suspect that the terminology could be made a bit more approachable, though I'd like to note specifically that the face/facet distinction seems straightforward and uncomplicated to me. I suspect that both you and I have a more technical bent than the average pixeller, though

Especially liked the coverage of Value: this shows a simple, quick technique that was used in demoscene graphics to good effect (altering line color in relation to the line's angle, to maintain correct looking thickness), but doesn't appear to be used much in modern pixel art.

I'm not sure I understand your earlier diagrams very well (ie. I know what they are supposed to depict but they don't 'read'), but it seems to work very well for detecting banding.

Does the potential for a thickened line to 'band' with itself need to be mentioned in Weight section? Here is what I mean:

Good example - each edge pixel(e) aligns with exactly one other edge pixel.

Code: [Select]

***e
e****e
..e****e
....e****e

Bad examples: each/most edge pixel(s) aligns with 2 other edge pixels

Code: [Select]

e*e
..e*e
....e*e
.....e*e
.......e*e


[AlexHW]

Thanks for the feedback, Ai.
The way of visualizing the concepts could be improved I think. At the time, it was all I could think of to help illustrate things.
The banding occurs with thin lines rather than lines that are of different thickness. This is due to the looser connections of linked pixels. So adding weight would be one way to prevent it, or by using less linked pixels.

I'm experimenting with another way of visualizing the connections/directions involved with pixels:

With this, it is easier to see the laced pixels- how they are bound/linked, creating a latticed pattern.
The bound and linked pixels also read better (horizontal/vertical lines for the bound connections, and diagonal for the linked connections).
I also took note of spots where the directions appear to cross (represented by the dots in the image). These occur when two linked connections compete with each other. I want to add this detail into the document (linked pixels that cross), since it is something I think is important.

[Indigo]

this latest representation is superb.  definitely suggests your on to something.  It detects edge flow really really well.