summarized from Discord conversation 2017-05-18
technical work by extra; ideas + transcription by unfavorablist
Can we extract data from the shifting “histogram” images in Stabilitory newing videos?
The images have fifty “pillars” or columns that can have a height from 0-35.
extra assigned the heights in the “pillars” alphanumerically, A-Z and 0-9. (36 unique values)
Here’s each unique frame from BELT 1 with each line being one frame: https://pastebin.com/raw/X82VLNA9
This data can be saved (in a text editor) as a .data file, and that file can be opened in GIMP (or other image editor).
The image rendered by GIMP reveals a certain pattern:
This pattern exists in a dataset derived from RATE as well.
Adjusting the width to 572 gives this:
To compare, I grabbed a random book (in plain txt format) from the Gutenberg Project, and saved that to a .data file. When I opened that in GIMP, it looks like this:
Doing this with all of SN’s BROTHER seems like a reasonable step.