--- Log opened Thu Mar 24 00:00:34 2022
00:49 -!- Jay_Dugger [~jwd@47-185-200-188.dlls.tx.frontiernet.net] has quit [Ping timeout: 256 seconds]
01:03 -!- Jay_Dugger [~jwd@47-185-249-104.dlls.tx.frontiernet.net] has joined #hplusroadmap
01:11 < fenn> there is a lot of 1700s vacuum tech that still works. low vapor pressure waxes, glass blowing, the sort of thing that requires a fair bit of fiddling and manual labor to make it work, which is why modern companies don't use those techniques
01:12 < fenn> why bother with con-flats when you can melt solder to form a perfect vacuum seal
01:12 < fenn> a reusable seal even. better than new!
01:24 < fenn> lol "Full disclosure: I designed the PCB and burned up a motor driver pushing current through the back-side coils before ever sitting down to do this writeup and the theoretical calculations above"
01:25 < fenn> good job
01:25 < fenn> lsneff how did you find this kevin lynagh's blog in the first place?
01:36 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
01:50 -!- Malvolio is now known as Guest9810
01:50 -!- Guest9810 [~Malvolio@user/malvolio] has quit [Killed (silver.libera.chat (Nickname regained by services))]
01:52 -!- Malvolio [~Malvolio@user/malvolio] has joined #hplusroadmap
05:26 < lsneff> He’s an acquaintance of mine
05:59 -!- bababb [~baracurda@user/parserblabla] has joined #hplusroadmap
06:33 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::bb3b] has joined #hplusroadmap
06:44 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 272 seconds]
07:10 < muurkha> you can't build a working vacuum tube with 01700s vacuum tech.  01800s yes
07:11 < muurkha> not sure why you'd need a specific pattern on the phone screen, but 10s of nanometers sounds unreasonably precise
07:14 < muurkha> before I read the paper let's see if it's plausible: a phone screen is maybe 4 megapixels over 70×120mm
07:15 < muurkha> like maybe 1500×2700
07:17 < muurkha> which gives maybe 45μm dot pitch, or 15μm in the X direction if it has vertical RGB subpixels as most do.  so probably with a perfect camera you could detect the position of a single pixel with an error of about 1 μm
07:18 < muurkha> or 3 μm in the Y direction
07:18 < muurkha> averaging normally distributed measurement errors of about 2 μm over 4 million independent samples should give you about 1 nm error
07:18 < muurkha> so 10s of nanometers is not at all unreasonably precise
07:20 < muurkha> though actually the abstract of the paper says "a precision below one nanometer" but it doesn't say it's using a cellphone screen
07:39 < lsneff> There’s a couple different papers. The one on the phone screen has an rms stdev of 10-30 nm
07:39 < lsneff> The one on a high precision substrate had an rms std of less than a nm
07:40 < lsneff> *rms error
07:46 < muurkha> there's a note in Dercuano describing (untested!) approaches to increasing the precision and DoF of such measurements: https://dercuano.github.io/notes/starfield-servo.html
08:01 < fenn> "Since announcing one’s wealth attracts a lot noise and toxic behavior, people keep it quiet, there aren’t Schelling points like “everyone knows that once you have $X you’re good to drop out of the rat race and pursue your creative dreams”, so it’s easy to keeping running on the treadmill (possibly indefinitely)"
08:13 < muurkha> lsneff: the principle in https://charted.space/papers/phase-encoded-pattern-for-visual-x-y-orientation-positioning.pdf is the same as the principle in the Anoto/FlyPen, though I think with a much less efficient encoding (requiring a much larger field of view for absolute positioning)
08:18 < lsneff> Which is less efficient?
08:21 < lsneff> Oh, the anoto system uses a de Brujin sequence
08:23 < lsneff> It looks like it’s probably not significantly more accurate than the grid of the paper
08:53 < muurkha> the Anoto system is more efficient
08:56 < muurkha> but if you have a small field of view you can't average a large number of dots to measure the fine detail of your position (what André et al. call the phase of the periodic frame)
09:29 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has quit [Ping timeout: 256 seconds]
09:30 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
09:40 < muurkha> they've gone to a lot of trouble to make the phase information easy to decode (that's part of where their low bit efficiency comes from) but I feel like once you have the right rotation you could just use a standard Fourier convolution coregistration algorithm to compute the displacement to high precision as long as there's a lot of high-frequency information
09:44 < muurkha> they're doing all that Fourier work *anyway*
10:00 -!- bababb [~baracurda@user/parserblabla] has quit [Remote host closed the connection]
10:43 < lsneff> Okay, interesting
10:43 < lsneff> I'll definitely think more about that
10:43 < lsneff> Thoughts? https://github.com/nffm/roadmap
10:58 < nmz787_> in my experience, you are more interested in doing high-value open-source projects when you're young (or at least willing to keep living like a poor college kid)
10:59 < nmz787_> my point being, I wonder if lsneff will eventuall tire from this pursuit (high value micro/nano positioner/litho machine) or start slapping on price tags that he'd scoff at today
10:59 < nmz787_> $20-28k for an AFM isn't really crazy if you're a shop charging $300/hr (and have a full work schedule) for imaging work
11:04 < lsneff> Well, I plan on being young for quite a while longer.
11:08 < lsneff> I find that I'm capable of doing multiple projects at once, so I'm planning on working in a normal engineering job and doing the open-source nanofabrication stuff in my free time.
11:26 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has joined #hplusroadmap
11:27 -!- Codaraxis [~Codaraxis@user/codaraxis] has joined #hplusroadmap
11:28 -!- Codaraxis__ [~Codaraxis@user/codaraxis] has joined #hplusroadmap
11:31 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has quit [Ping timeout: 256 seconds]
11:32 -!- Codaraxis [~Codaraxis@user/codaraxis] has quit [Ping timeout: 256 seconds]
12:11 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has joined #hplusroadmap
12:13 < kanzure> .title https://news.ycombinator.com/item?id=30789046
12:13 < saxo> YouTuber builds his own x-ray machine after $69k hospital bill (2021) | Hacker News
14:11 < jrayhawk> "i'm going to build my own x-ray machine for way cheaper!" *buys old x-ray machine off of ebay*
15:10 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has quit [Remote host closed the connection]
16:24 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
17:34 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 240 seconds]
18:10 -!- Codaraxis__ [~Codaraxis@user/codaraxis] has quit [Ping timeout: 256 seconds]
18:47 -!- Codaraxis [~Codaraxis@user/codaraxis] has joined #hplusroadmap
19:07 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::bb3b] has quit [Quit: Leaving]
20:15 -!- Moon11 [~Moon11@d-216-36-20-130.md.cpe.atlanticbb.net] has joined #hplusroadmap
20:15 < Moon11> hello?
20:15 < Moon11> anyone there?
20:16 -!- Moon11 [~Moon11@d-216-36-20-130.md.cpe.atlanticbb.net] has quit [Client Quit]
20:16 -!- Moon11 [~Moon11@d-216-36-20-130.md.cpe.atlanticbb.net] has joined #hplusroadmap
20:17 < Moon11> Im looking for Bryan Bishop (Kanzure)
20:36 < superkuh> Hello. Some people are here. Sometimes.
20:39 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has joined #hplusroadmap
20:42 < Moon11> what is this place
20:43 -!- Codaraxis [~Codaraxis@user/codaraxis] has quit [Ping timeout: 260 seconds]
20:54 < lsneff> Ah, must be someone trying to hire him
20:54 -!- Moon11 [~Moon11@d-216-36-20-130.md.cpe.atlanticbb.net] has quit [Ping timeout: 256 seconds]
23:22 < nmz787_> https://www.microscopy.org/images/posters/Rempfer.pdf
23:22 < nmz787_> "Demonstration of aberration correction, from Rempfer et al. 1997."
23:22 < nmz787_> "The correction could easily be appreciated by imaging a fine-mesh screen."
23:34 < nmz787_> https://vintagetek.org/tek-electron-microscope-donated-to-museum/
--- Log closed Fri Mar 25 00:00:36 2022