You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I'm having a hard time getting the camera to use a different sensor mode. The camera seems to be switching mode but always returning image with resolution 1920x1080, regardless of what I pass to the sensor-mode argument.
If I run the following pipeline, I get the correct image
Now, if I mix things up, so choose the sensor mode 1296x972 but don't specify the resolution, I get the following image, which is (visibly) correct, has a wider field-of-view (as expected given that the mode 1296x972 has a wider field-of-view compared to 1920x1080) but the size of the image is still 1920x1080.
I haven't tried the raw image output mode for a long time - I normally use h264 or jpeg capture, because the transfer from the VPU to CPU is inefficient for raw frames and limits you to just a couple of FPS. If your goal is to record jpeg files, you should let the VPU do that encoding:
I think the output size choice is something to improve, but not necessarily a bug. The output resolution is negotiated independently of the capture resolution. If you add the resolution to the output capsfilter it should do what you want:
Hi, I'm having a hard time getting the camera to use a different sensor mode. The camera seems to be switching mode but always returning image with resolution 1920x1080, regardless of what I pass to the
sensor-mode
argument.If I run the following pipeline, I get the correct image
Though, if I try to select the sensor mode
1296x972
with the following pipeline, this is what I getNow, if I mix things up, so choose the sensor mode
1296x972
but don't specify the resolution, I get the following image, which is (visibly) correct, has a wider field-of-view (as expected given that the mode1296x972
has a wider field-of-view compared to1920x1080
) but the size of the image is still1920x1080
.The artifacts look like they come from a rowbyte/stride problem.
Any idea what the cause could be?
My environment:
The text was updated successfully, but these errors were encountered: