iiinova — where your digital self keeps evolving

Michael Bach

@bachmichael.bsky.social

michaelbach.de. I'm interested in #VisionScience, #illusions, #psychophysics, visual #acuity, #neuroscience, #electrophysiology, #music, #programming, #jekyll, cappuccino (the framework), #rstats. Democracy👍, Hate👎

Posts Replies Media

Michael Bach @bachmichael.bsky.social · 31/10/2025

A new addition to my “Visual Phenomena and Optical Illusions” collection: “Mainz-Linez”. Here's the live demonstration and explanation of the somewhat mysterious title: michaelbach.de/ot/mot-Mainz.... Enjoy!

A static image of the Mainz-Linez illusion: 4 vertical wiggly lines, in the middle of each a red disk. A small black cross in the center.

0147

Michael Bach @bachmichael.bsky.social · 02/05/2025

michaelbach.de/ot/aud-Shepa... My demo of the ever rising scale, invented by Roger Shepard, had failed due to new security restrictions in browsers. I rewrote it and added plot showing amplitude vs. log(frequency), hopefully making it easier to understand this amazing effect. Enjoy! #illusion

Screenshot of the ever-rising Shepard scale with frequency spectrum.

040

Michael Bach @bachmichael.bsky.social · 05/01/2025

Quitting the Washington Post: anntelnaes.substack.com/p/why-im-qui.... Impressive, my admiration! Good read. Hope not harbinger of worse things to come. I learned of it via daringfireball.net.

100

Michael Bach @bachmichael.bsky.social · 31/12/2024

Bicycling through the nearby woods, I happenchanced upon this frosty gem.

050

Michael Bach @bachmichael.bsky.social · 10/12/2024

🧪In our journal club: Just had a paper where the authors needlessly made their figures rectangular, although the content was quadratic (same range on x- and y-axis). It really should be quadratic to understand the content better, also the identity line is then at 45°. Old gripe of mine, sorry 😎.

Two data plots. On the left needlessly landscape, on the right quadratic like the data themselves.

020

Michael Bach @bachmichael.bsky.social · 27/11/2024

A new paper on FrACT michaelbach.de/fract/: “Freiburg Vision Test (FrACT): Optimal number of trials?” (open access) dx.doi.org/10.1007/s004.... Mysterious “kink” at 18 trials (figure): 18 trials a good compromise between fatigue and precision for Sloan letters or 8-alternative Landolt-Cs.

Graph showing LoA steadily decrease with the number of trials, above 18 slower. Inset is the FrACT icon.

042

Michael Bach @bachmichael.bsky.social · 26/11/2024

I just finished an experiment: For three weeks I responded to every "invitation" from a predatory journal with "STOP PESTERING ME". Results: I now receive about twice as many per day than before… Time to end the experiment and conclude that I'm just validating my email for them 😎. #science

193