Re: Camera Resolution Test
Posted: 18 Apr 2017, 16:55
One more thing that can effect the FPS throughput. Make sure the "prevent processor idle state" setting in the advanced settings is on.
the board for everyone interested in 3d scanning
https://www.3dscanforum.org/
It's on, one of the first thing I did when my camera kept hanging up. Theres another thing that makes me wonder, does the app make use of all the cores?
The Tech at ZWO informed me it detected multiple cameras using the directdraw drivers, so it's all good. Wasn't your Point grey a firewire?Curiousjeff wrote: ↑18 Apr 2017, 09:37 Since I had a bad experience with the PointGrey driver (you cannot use two cameras within Daivd 5), you should check the software compatibility with the manufacturer if you intend you use two cameras.
My 2ct.
Jeff
Something is not right. I run two cameras at 1920x1080 54fps and the scan processing is virtually instantaneous. I'm getting a solid 54fps (the max frame rate of the cameras) all the time and a single 2 camera scan at default (33 patterns) takes ~3.5sec - 4sec. Last night I completed a mesh that had 109 scans (extended mode). Global fine registration took about 9 minutes and fusion about 4.5min. My machine is 2 years old and not crazy.Oteck wrote: ↑21 Jun 2017, 06:19 Back for some more testing. It's now apparent that at very high resolution it will cause some serious cpu processing to run. I don't believe any upgrade will do as this is pure processing power to get gigs of scans in stereo.
A single stereo scan would take literally 5 minutes to get past the processing, this is not good but shows higher res isn't really better productivity wise.
I can get around this by using 2x binning the pixels giving me reduced the resolution of 1544x1020. Doing this will also reduce the required exposure as 4 pixels are gathering light to become 1 and increase frame rates.