ph5 / ouvrt Goto Github PK

Tried on a Samsung Odyssey plus. Build with gstreamer and pipewire disabled.

udev: Found HoloLens Sensors: /dev/bus/usb/008/003
HoloLens Sensors: Serial e146fa9d-fd47-4b81-97cd-11745667db51
HoloLens Sensors: USB3
HoloLens Sensors: acquired new id 0 for serial e146fa9d-fd47-4b81-97cd-11745667db51
ouvrtd: Acquired session bus, name: 'de.phfuenf.ouvrt.Ouvrtd'
ouvrtd: Acquired name "de.phfuenf.ouvrt.Ouvrtd"
HoloLens Sensors: Missing frame: 0 -> 25
debug: connected
debug: disconnected

I recompilled with gstreamer disabled and pipewire enabled and it works.

udev: Found HoloLens Sensors: /dev/bus/usb/008/003
HoloLens Sensors: Serial e146fa9d-fd47-4b81-97cd-11745667db51
HoloLens Sensors: USB3
HoloLens Sensors: acquired new id 0 for serial e146fa9d-fd47-4b81-97cd-11745667db51
ouvrtd: Acquired session bus, name: 'de.phfuenf.ouvrt.Ouvrtd'
ouvrtd: Acquired name "de.phfuenf.ouvrt.Ouvrtd"
HoloLens Sensors: Missing frame: 0 -> 25

There seems to be always missing frames 0->25.

Edit:
The first stream is 30 fps and works fine for gray video.
The second stream is 60 fps and is for tracking the controllers as it will only show bright led.
Seems like second stream could track PS move controllers !

Thanks for your work !
But still can't get O+ to display with openhmd.

Hi friend,
I got a problem during debug mt9v034 snapshot mode in another project . And see you code have used snapshot mode, I'd like to ask if you have the same problem.

Firstly I initialize the sensor, set the sensor to snapshot mode (r0x07set to 0x0398), 
and turn off  aec / agc function (r0xaf set to 0x0000).
Then give trigger pulse to ex-posure pin and read out an image.

The problem is:
When the trigger cycle less than about 0.5s, the read out image looks more normal,
When the trigger interval is growing, the image appears more and more white(color) noise, (See attachment)
When the trigger interval time reaches a few minutes, the image is basically a white paper.

In addition, through observation and experiment also found:
1, the location of the noise completely fixed.
2, in the idle time period standby sensor can not solve this problem.
3, from the point of view, like in the idle time, sensor has been exposed,
but by observing led_out pin we found exposure time is correct.
4, I tried similar initialization code in your “int mt9v034_sensor_enable_sync (int fd)” function , get the same result.

So I submit the issue to ask for your help.

Looking forward to your reply, thx!~

0.5s

1s

2s

5s

10s

30s

60s

600s

Related video
20170825154727.zip