SharpCap Forums

Hello Robin,

I checked the fps performance with and without binning with QHY174M-GPS. In the color space MONO16, the fps does not change with or without binning, but strangely, in the color space MONO8, the performance deteriorates with binning x2.
Is this a QHY driver issue?
SharpCapPro 4.0.8047, QHYdriver is ver.2021.03.13
PC is HP notebook i7-CPU, 16GB memory, Win10 pro 20H2

Best Regards,
Hiroyuki Watanabe

Hi,

Just tested with:
SharpCapPro 4.0.8060 32bit (latest), QHY driver is 20.6.3.0,
PC is Lenovo IdeaPad L340 AMD Ryzen3 3200 2.6GHz, 8GB memory, Win10 Home 21H1,
Original QHY USB3 cable, with GPS-antenna, locked state,
The fps listed is the SC shown one, no test regarding dropped frames

The result you see in the picture attached.

BTW: In only one setting a "Bad calibration - reduce end pos" was shown, in all other cases the autocal gave a lock immediately.

Cheers, Christian

Hi,

binning rarely increases frame rate on CMOS cameras, as it is usually done in software (and most often on the PC after transferring the full frame from the camera anyway). I wouldn't expect a reduction in rate though.

As a comparison, I tested with a QHY178C and the 174GPS (not enabling GPS) and I got the same frame rate in bin2x2 as bin1x1 for both. Could it be something to do with having the GPS on?

cheers,

Robin

Hello Robin,

Turning on GPS seems to reduce binning performance.
However, turning off GPS does not seem to eliminate the tendency.

Best regards,
Hiroyuki Watanabe

Hi,

I just tested again with GPS on and I still get maximum FPS in 2x2 binning.
I did notice that the GPS info window is updated every frame, which is just silly at high frame rates, so I will tweak it to only update at a maximum of about 10 times per second.

cheers,

Robin

Hello,

I test with my camera ... here the following results (I do only with 1920x1200).

Mono16 without GPS connected (1x1 and 2x2):
5 ms : 76.8 fps, -
10 ms: 75.7 fps, -
15 ms: 65.4 fps, -
20 ms: 49.6 fps, 49.2 fps

Mono8 without GPS connected (1x1 and 2x2):
5 ms : 132.6 fps, 130.8 fps
10 ms: 97.7 fps, 98.1 fps
15 ms: 66.4 fps, 65.4 fps
20 ms: 49.6 fps, 49.4 fps

Mono8 with GPS connected and locked (1x1 and 2x2):
5 ms : 131.4 fps, 134.1 fps
10 ms: 98.2 fps, 99.2 fps
15 ms: 65.6 fps, 65.6 fps
20 ms: 49.7 fps, 49.6 fps

No test with Mono16 + GPS.

SharpCapPro 4.0.8042, QHYdriver is ver. 21.02.25
AMD Ryzen 9 3950X, 128GB memory, Win10 Pro 21H1

=> no drop with 2x2 binning (in some cases, the rate is a little bit higher in 2x2)

Jean-Francois

Hi,

With the same settings/setup as above I repeated the test with GPS OFF (blue), but only for full frame. I confirm Robin's and Jean-Francois's result: no framerate drops.
In addition, I repeated the test also with GPS ON. Surprisingly I could not reproduce the drops found (red) in my first test. So all the green values are now identical for 1x1 respectively 2x2.

so I will tweak it to only update at a maximum of about 10 times per second

Good idea, Robin - but I hope: not for the GPS-log.

Cheers, Christian

Hello,

It looks like Binx1 and Binx2 use CPU threads differently. Binx1 uses 4 threads, but Binx2 seems to use only 1 thread.
This seems to require a CPU with a high capacity of one thread in Binx2.
Can Binx2 also be multi-threaded?

Best regards,
Hiroyuki Watanabe

Hi,

Hiroyuki, I cannot confirm your observation regarding threading. With my system described above I do not see a difference, always all 4/4 threads seem to be used.
On a second system under W7SP1 i7 I observe 4/8 threads are used regardless the SC settings.

Cheers, Christian

Hi,

SharpCap always uses multiple threads - one to grab from the camera, one to perform any processing, another to display the images, another to save, etc. The SDK may also use threads, but I don't have any control over that.

cheers,

Robin