I went through two battery charges today with the OneWheel. I've been working on trying to intentionally defeat the balance and power to spin around on the edge of the wheel to pirouette like a figure skater.
I can pretty consistently spin 180 degrees and then ride off the way I came, but am hoping to either get all the way around or even have a stable continuous spin.
I can spin 3 or 4 times in one motion, but it's more like riding around a bunch of 1 foot diameter circles.
I had a couple low-speed spinning falls, resulting in the biggest contusion I've ever seen. I have two elbows on my right arm, and the actual elbow is barely visible because of the giant false elbow contusion next to it.
I've been trying to describe to people what being high on ketamine does to auditory processing. It's a really hard thing to describe, but last night I cracked it.
You know that "video feedback" effect in early music videos, where copies of the video kind of zoom around each other?
Well, imagine that, but with your ears instead of your eyes. That's what ketamine sounds like. That video effect, but also with the worst tinnitus in the world. (But in a good way.)
@onelson if it didn't have positive non-audio effects, I wouldn't want it either. I don't find that aspect pleasant the way I do the extreme peace/calm, and the post-high mental health improvements.
I don't think the ketamine causes tinnitus, it just makes my existing tinnitus way, way worse.
My hypothesis is that ketamine affects the "perceptual filters" (a concept I invented to simply describe the way your brain decides what's noise and what's signal), turning way down the noise removal function.
This overlaps quite a bit with previous observations about the effect THC has on both auditory and visual noise processing.
I kinda wish I got to have drug experience in high school. If I had, maybe I'd be a neuroscientist now instead of a fed up / washed up web developer.
What is that old-school video effect that was used in a lot of early music videos, where something is shot in front of a dark background and then sort of repeated forever kind of like the "infinity hallway" effect you get by standing between two mirrors?
I'd be happy even with just an example of one, but would love to know how it worked.
@onelson the terrible Back to the Future II/III for the NES misspells DeLorean. As long as you spell Victorian correctly I don't think you'll get any grief
As a teen in America, I used to pray for a PBS pledge drive so I could see a bunch of episodes of Red Green in a row. Kind of curious how well that show held up.
Microsoft: takes screenshots of the screen, OCRs them and places the data in an unprotected sqlite file for later recall.
Apple: creates new NPU architecture designed to save PDFs of the screen and OCR them, storing the data in CoreData for later recall by authenticated users.
Ubuntu: every time the computer sleeps, it wakes up slightly slower than the last time, until you reboot it.
To clarify here, when I said "wakes up slightly slower," I mean "it runs slower and slower after each sleep," not "it takes longer and longer to wake up."
is anyone familiar with a method of extracting a stillframe image from a particular timestamp of a video that considers temporal information from neighbouring frames, in order to produce a still image that looks more like what we perceive during playback? non gen-ML based solutions strongly preferred.
regular frame extraction tends to have blocking artifacts and other quality issues that are harder to perceive in realtime.
I assume something along these lines exists for video forensics.
@gsuberland mencoder/mplayer and ffmpeg both have somewhat tunable de-interlacing and deblocking as well a reversing a 3:2 pulldown, but I'm not sure that reverse pulldown really counts as "temporal"
@gsuberland I suspect it's also highly dependent on the framerate history of the content. Like if it's a 24fps video that's been interlaced into 29.97fps fields, it'd be pretty tough to get back to 24fps cleanly. But this could be a fun experiment.
Kevin Beaumont suggested that Recall history can be "edited" with Photoshop via the stored jpegs, but it doesn't seem to me like that would result in the contents of the database changing. I think sqlite queries would be needed to truly edit history...
Any tips for getting Windows 11 to shut down Bluetooth when my Surface Laptop sleeps? My headphones keep remaining connected to the device, and often get "stolen" from my phone. I'll be listening to something on my phone and all of a sudden, it'll switch to the (sleeping) PC and output the nothing that the PC is outputting instead of the actual music my phone is playing.
Seems like the consensus for how to do this is "fuck off because that's how it works now. Feel free to switch to 'hibernate' instead of sleep."
So I switched to hibernate and it took a full minute to un-hibernate on wake. The first 45 seconds of which was just a black screen with no indication that anything at all is happening. I was laughing out loud at the prospect of Windows having the same "fails to wake up" problem that Linux laptops often do. Was gonna say WSL2 has gotten really good.
Got another comment about my Holo Taco hat. It blows my mind that some girl in Canada can put too much nail polish on her nails ("polish mountain"), go viral, become a successful YouTuber, start a nail polish business and have random strangers all over the world recognize her logo.
One person joked (I assume they were joking anyway) about how it's impossible for an Agile project to fail, so since these projects failed, they can't have been Agile projects.
This whole thing does suggest to me that this is also a hotdog/Pluto/airborne issue: "fail" is not adequately defined in this context.