| qi-bot | [commit] Maarten ter Huurne: fbcon: Tweaks to non-ASCII characters in 6x10 font (jz-3.3) http://qi-hw.com/p/qi-kernel/2284faa | 07:34 |
|---|---|---|
| qi-bot | [commit] Maarten ter Huurne: Use title case for Dingoo link titles. (master) http://qi-hw.com/p/gmenu2x/e5a9227 | 11:01 |
| Action: DocScrutinizer idly wonders what might be the largest photosensor array you can get for low price -MP count irrelevant | 15:19 | |
| DocScrutinizer | pondering a DIY Roentgen device, with electronic photo platter rather than chemical film. So any slaughtered VGA webcam module with a sensor area of less than 3mm diagonal is not suited for that purpose | 15:21 |
| DocScrutinizer | I'd need at very least 3/4" sensor | 15:22 |
| DocScrutinizer | better 2" | 15:22 |
| DocScrutinizer | or rather combine some smaller sensors in an array - I only wonder what to do for the frames | 15:23 |
| DocScrutinizer | AAAH, I could just overlap sensors, and subtract the "artifacts" from final combined readout | 15:24 |
| DocScrutinizer | :-D | 15:25 |
| DocScrutinizer | pretty good advantage when your sensors are "transparent" | 15:25 |
| DocScrutinizer | or do two shots in sequence, and offset the sensor array between them, so the borders in one snapshot are covered by good areas in second | 15:28 |
| DocScrutinizer | :-/ for a complete coverage, I'd need THREE shots | 15:29 |
| whitequark | DocScrutinizer: won't xrays destroy the sensor? | 16:28 |
| whitequark | ion migration and stuff | 16:28 |
| DocScrutinizer | nah | 16:29 |
| DocScrutinizer | don't think so | 16:29 |
| DocScrutinizer | not this level of x-ray at least | 16:29 |
| DocScrutinizer | it's more like "will the sensor be sensible enough for this low x-ray exposure?" | 16:30 |
| whitequark | ahh. | 16:34 |
| whitequark | artifacts | 16:35 |
| whitequark | what's with artifacts? at least I believe that CMOS noise pattern is always same for each sensor and always differs between ones | 16:35 |
| DocScrutinizer | stuff like http://www.celnav.de/hv/hv5.htm and http://www.fingers-welt.de/gallerie/eigen/elektro/roentgen/roentgen.htm work with exposure times of like hours, on chemical film | 16:36 |
| whitequark | what do you want to measure? | 16:37 |
| DocScrutinizer | well, technical things. Like e.g. balls / solderpoints of BGA chips | 16:38 |
| whitequark | ah, then it's kinda simple | 16:39 |
| whitequark | I think you need to get an exposure of CMOS noise to get the pattern | 16:39 |
| whitequark | and then the same for the signal | 16:40 |
| DocScrutinizer | sure, on electronics you can just subtract the blackpicture, then divide by whitepicture | 16:41 |
| DocScrutinizer | would even eliminate the artifacts of overlapping sensors' frames | 16:42 |
| whitequark | huh, I know a guy who got an xray tube and exposured his hand through it | 16:45 |
| whitequark | got him into local news and to a radiation therapy unit | 16:45 |
| whitequark | ... which he was quite happy about because he got to know a few guys related to nuclear power station maintenance | 16:46 |
| whitequark | yeah, "know" not "knew". he also had his scope plug broken, so he just slided the contacts in the socket with his bare fingers | 16:46 |
| whitequark | I'm somewhat surprised he is still alive | 16:46 |
| wpwrak | russian engineering wouldn't be half as entertaining without all the vodka ;-) | 17:20 |
| DocScrutinizer | hehe | 17:35 |
| whitequark | funny thing is, he doesn't drink due to high ICP | 17:52 |
| wpwrak | Insane Clown Posse ? | 17:53 |
| wpwrak | ah, Intracranial Pressure. hmm. sounds fitting ;-) | 17:54 |
| whitequark | so you use duckduckgo too, hmm | 17:54 |
| wpwrak | naw, chromium | 17:55 |
| wpwrak | so google by default | 17:55 |
| whitequark | these two statements do not contradict one another | 17:55 |
| whitequark | ah | 17:55 |
| whitequark | huh, someone from Spain wrote me about a decentralized VPN I made several years ago while in high school | 18:25 |
| whitequark | no idea how did he manage to ever find it | 18:26 |
| rz2k | someone, please give me an explanation about why my new tablet works only when I press flex-pcb connector from lcd to mainboard. | 18:27 |
| Action: rz2k cant even imagine what the f. is the reason | 18:28 | |
| rz2k | shielding, flex-pcb cracks, RF stuff? :/ | 18:29 |
| whitequark | obviously, poor soldering of flex-pcb | 18:30 |
| whitequark | or cracks in it | 18:30 |
| whitequark | I'm not sure what do you mean by "not works", but maybe it requests something via I2C from it and hangs without a reply | 18:31 |
| whitequark | or something like that | 18:31 |
| rz2k | nope | 18:31 |
| rz2k | it just freezes random | 18:31 |
| rz2k | on ipl/spl/android logo | 18:31 |
| rz2k | just random as hell. | 18:31 |
| rz2k | thanks to blue soviet duct tape, its fixed for now. | 18:33 |
| --- Sun Apr 29 2012 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!