12 𓅃 @12.shsbs.xyz · 16/10/2026ok we were finally able to confirm that the same frequency was also used in the ensoniq SD-1 and SQ-1 (the "overclocked" 32-voice variants; the originals used 20MHz), one of which we know the F3 is based on. 220 2
12 𓅃 @12.shsbs.xyz · 15/10/2026still very common to get sample rate jitter in my recordings though, which is a separate issue: 110 1
12 𓅃 @12.shsbs.xyz · 15/10/2026"bit errors are extremely rare" OH YEAH? here's 2 of em (and these aren't just due to sample rate variation, where you'd see 1 missing or duplicate sample) 110 1
12 𓅃 @12.shsbs.xyz · 13/10/2026so like look at this shit. this is totally valid afaik. the F3 outputs 16 bits for the left channel, then 112 bits for the right channel. of course, most of those are just repeating junk data, and the DAC only looks at the first 16. their place values are so small you wouldn't hear them anyway. 100 1
12 𓅃 @12.shsbs.xyz · 01/10/2026for some reason CO2 seems to rise a lot faster at night. it can go from 400ppm to unhealthy levels from just 1 night with the window closed. 130 3
12 𓅃 @12.shsbs.xyz · 01/10/2026the oven is crazy though. CO2 goes from 600ppm to 1500 in SECONDS 240 4
12 𓅃 @12.shsbs.xyz · 30/09/2026anyway yeah it was BAD like CO2 was probably over 2000ppm most of the time (it should be 400-800) 140 4
12 𓅃 @12.shsbs.xyz · 30/09/2026i opened the window (this time not the one next to the sensor) and it dropped to 1500 for a few minutes, but then we turned on the oven and it's back up to 3000... 000 0
12 𓅃 @12.shsbs.xyz · 30/09/2026opened window then closed it again (presumably the level increased so fast bc it was only measuring the pocket of fresh air around the window which went away quickly after closing it) 210 2
12 𓅃 @12.shsbs.xyz · 26/09/2026if i extract the font from the pdf, the points are like this: (with the original shown behind it in green) specifically the curve on the inside of the right hump of the m, was represented by `5 12 14 4 8 0 16 0 0 -9 0 flex1` (which is shorthand for 2 curves) got interpreted incorrectly 100 1
12 𓅃 @12.shsbs.xyz · 26/09/2026somehow when i export a PDF from libreoffice, it corrupts the letter "m" glyph in the font... i figured out what feature is causing it, but i don't understand HOW it could break in this way 381 8
12 𓅃 @12.shsbs.xyz · 25/09/2026remember the knockoff version of harry potter that existed in the background of Arthur on pbs kids? we should replace harry potter with that. write your fanfics about henry skreever 0100 10
12 𓅃 @12.shsbs.xyz · 18/09/2026circuit construction technique: i'm gonna mount these IC's upside down with their legs sticking up, and then use wire wrap to connect everything. should be very reliable, and with no soldering and no expensive sockets 260 6
12 𓅃 @12.shsbs.xyz · 14/09/2026but no please explain to me how this guy is just some poor misunderstood autistic man who doesn't deserve all of our ire. like even if these 2 had some past beef, this is unacceptable behavior esp for such a large account. 2100 10
12 𓅃 @12.shsbs.xyz · 11/09/2026so apparentlyyy i wasn't changing the clock carefully at all... the "glitchless" multiplexer i guess just means you don't get any tiny short pulses, not that the phase is actually matched. 100 1
12 𓅃 @12.shsbs.xyz · 10/09/2026like ok what is the point here? you can just turn the former into the latter by holding it 1 inch closer to your face lol maybe i'll never understand like i don't even bother turning my phone sideways to watch videos, so i'm definitely not going to Transform it just to make the screen 20% larger 230 3
12 𓅃 @12.shsbs.xyz · 08/09/2026usb type b (the big kind) is by far my favorite usb connector (especially the extended SuperSpeed one) like they always feel so satisfying and substantial (and are the easiest to tell which way to plug in) i will never be annoyed when a device uses THIS instead of usb-c. they arent compareable 171 7
12 𓅃 @12.shsbs.xyz · 07/09/2026and here's a captured frame of video with the partial scanlines (source: github.com/oyvindln/vhs.... i un-deinterlaced it and scaled it horizontally to the same width) 000 0
12 𓅃 @12.shsbs.xyz · 07/09/2026anyway i finally made my own diagram of ntsc timings because i could never find a good one. 1px x = 1µs, 2px y = 1 scanline 240 4
12 𓅃 @12.shsbs.xyz · 07/09/2026here is a to-scale diagram of NTSC visible scanlines, and you can see how the 2 partial scanlines fill in the space left over in the corners since the raster would be a parallelogram otherwise 251 5
12 𓅃 @12.shsbs.xyz · 06/09/2026positioning your cursor over the cursor in a screenshot/video so they touch,,.. 2468 46
12 𓅃 @12.shsbs.xyz · 26/08/2026btw it really is like this that is a Waveshare Core2350B in the middle 130 3
12 𓅃 @12.shsbs.xyz · 26/08/2026video output is working, currently rendered from the same data format as it would be if read from the F3 (doing the alpha blending etc) 130 3
12 𓅃 @12.shsbs.xyz · 14/08/2026these blocklists are so comical lol (and of course 100% of the people on it are like trans girls who like to play toys with their friends) 2211 21
12 𓅃 @12.shsbs.xyz · 04/08/2026still thinking about how a real trans person actually said this here.. being a puritan rots your brain 10486 48
12 𓅃 @12.shsbs.xyz · 01/08/2026i don't know if these resistors are necessary but other designs included them so i built this resistor array and connected it to the DVI connector with wire wrap (they're barely large enough to fit the tool between. another thing HDMI could never do) 130 3
12 𓅃 @12.shsbs.xyz · 29/07/2026lolllll thats rigjt thats me lil piss baby cry boy rotate that in your mind 3121 12
12 𓅃 @12.shsbs.xyz · 27/07/2026state tree, if this makes sense umm green arrows are when the low bit changes (so we decrement y, then negate it) blue arrows are when the high bit changes (so we negate y, then decrement it) 100 1