Various topics of interest... to me, at least!
A belief propagation network is one (popular) algorithm used to process digital messages containing low-density parity check (LDPC) forward error correction (FEC). While not being "embarassingly" parallelizable, the algorithm can benefit from parallel execution using simple instructions and is therefore a candidate to run on a GPU. Some proof-of-concept code is contained here.
Turn Your Raspberry Pi into an IR/Infrared Remote Control
UPDATE: For lower cost, I've also implemented this on the Raspberry Pi Pico W using uasyncio. See bottom of post for details.
I had run into issues with infrared remote control routines available on the internet for the Raspberry Pi. The main issues: being forced to use some specialized hardware that cost as much as a Raspberry Pi or getting interrupted by the Linux kernel all the time when controlling pins myself! Check out the details for how I overcame those issues.
Adding a Variable-Volume Line Out to my Sony Bravia TV
My Sony Bravia HDTV has built-in speakers and a fixed-level line out. I had a Logitech Z323 2-way + subwoofer speaker system with a volume knob. This is how I merged the two and kept my remote volume control!
Calculating (instead of PSPICE'ing) the value of the Resistor Mass in XKCD #730
I couldn't find anyone who had simply calculated the value of the resistor mass, vs finding it via simulation in PSPICE, so I thought I'd go ahead and do that. Hint: there are 27 unknowns.
Trimming Silence in Audio Files
FFMPEG's silenceremove functionality is a bit cryptic and unwieldy. This substitute was done as a quick project just to see how simply I could code it using Python, while still maintaining performance. It ended up being able to process 2 hours of audio in 5.7 seconds on my midrange, 2020 machine! Not bad for Python! I also put in a lot of nice features like auto-thresholding and blip ignoring.
Making Movies from tinySA Captures
The tinySA is a small spectrum analyzer that lets you send commands over serial/USB, one of which is to retrieve a 'capture' (screenshot). This script uses moviepy to periodically capture images and combine those images into movies every frame-count frames. This is useful when using the spectrum analyzer to monitor activity over a long period of time. The max-hold functionality can be activated and reset at the beginning of every movie.