onfocus

Back in 2015 I ruined good art by Audubon, Rothko, and Albers with a Python script that creates a low poly version of an image. I put the script on GitHub and forgot about it.

Until today! Someone submitted a pull request that updates the script for python 3 and some newer versions of the dependencies. I got it running again to test it out:



This is a low poly version of Van Gogh's Wheatfield with Crows.

In conclusion, coding in public is good and maybe there's a hidden demand for generating triangles with Python.

How about three Eames chairs barely rendered as various sized dots?

PAW, 1952. DSR, 1954. And LCW, 1946.

Something about the curve of that not pipe reminded me of the curves in Audubon's birds. This ruined image started out as Audubon's American Crow, 1861.

After my experience ruining squares, I wondered if images of color field paintings could be ruined with triangles even with no sharply defined structure. So I tried the same thing with some images of paintings by Mark Rothko.

Here are the three I picked: No. 61 (Rust and Blue), 1953. White Over Red, 1957. And No. 46 (Black, Ochre, Red Over Red), 1957.



And here they are ruined with Python:

I've been having fun lately generating triangles with Python, as you do. And the other day I stumbled across one of Josef Albers' Homage to the Square. My first thought was, "Wow, the colors." My second thought was, "Those should be triangles."

I went to Google Images and searched for Josef Albers Homage to the Square and picked three at random: Ascending, 1953. Glow, 1966. And Soft Spoken, 1969.



Then I ruined them with computers by turning everything into triangles:







You might want to take a look at some non-ruined versions of Homage to the Square, like Impossibles from 1931 or this one from 1967.