Hi, I have some question with OAK-D Pro Wide(OV9782). 1. We know like this. the current default setting of Color camera is 30 frames 1920 1080. In this case, Based on our test, the lag occurs such as the depth camera shows a real-time image, but the color camera gradually slows down and the image comes out 1 second earlier. So it seems the captured time of the images is different between stereo camera and color camera. And one more thing, When I set it to 10frame, I didn't test for a long time, but there was no slowdown problem. So we want to know if I can create an additional function to still receive the synced latest video with stereo camera and color camera. 2. We would like to know when can we get the C C++ CODE of the WIDE camera. 3. AI Network file is *.yml, but we hope to know whether it is possible to convert it to Hef or not. Best regards, Ryan.

Hello @"RyanLee"#p4312 , 1. What script did you use to get 1sec delay? We have done some tests with OAK-D (so not the same sensor), you can see [delay results here](https://docs.luxonis.com/projects/api/en/latest/tutorials/low-latency/). 1.1 **Syncing**: There are a few approaches to sync (eg. color frame + depth frame, as in your case), [see tutorials here](https://github.com/luxonis/depthai-experiments/tree/master/gen2-syncing). 2. Which code are we talking about? See [depthai-core repo](https://github.com/luxonis/depthai-core) for the C++ API code. 3. I am not sure what hef is, could you elaborate? Thanks, Erik

Questions about OAK-D Pro Wide

RyanLee

Hi,

I have some question with OAK-D Pro Wide(OV9782).

We know like this. the current default setting of Color camera is 30 frames 1920 1080.
In this case, Based on our test, the lag occurs such as the depth camera shows a real-time image, but the color camera gradually slows down and the image comes out 1 second earlier. So it seems the captured time of the images is different between stereo camera and color camera.

And one more thing, When I set it to 10frame, I didn't test for a long time, but there was no slowdown problem.
So we want to know if I can create an additional function to still receive the synced latest video with stereo camera and color camera.

We would like to know when can we get the C C++ CODE of the WIDE camera.
AI Network file is *.yml, but we hope to know whether it is possible to convert it to Hef or not.

Best regards,
Ryan.

erik

Hello RyanLee ,

What script did you use to get 1sec delay? We have done some tests with OAK-D (so not the same sensor), you can see delay results here.
1.1 Syncing: There are a few approaches to sync (eg. color frame + depth frame, as in your case), see tutorials here.
Which code are we talking about? See depthai-core repo for the C++ API code.
I am not sure what hef is, could you elaborate?

Thanks, Erik

RyanLee

Hi,

Could we make C or C++ application to use its luxonis/depthai-core ??

And is the below link for that?? Could we have c or cpp version of depthai-python like example source code??
https://docs.luxonis.com/projects/api/en/latest/references/cpp/

Best regards,
Ryan.

erik

Hi RyanLee ,
depthai-core is cpp version of the depthai-python. So yes, you can make c++ apps with depthai-core, example here.
Thanks, Erik

RyanLee

Hi EriK,

Thank you for your feedback. I got it and i have additional inquiry about excuting multiple models such as face detection, emotion detection and hand tracking at the same time in C or C++. are there any example application to run multiple model with c or c++??

Best regards,
Ryan.

erik

Hi RyanLee ,
Python and C++ depthai APIs are 1:1 (as depthai-python are just bindings on top of depthai-core). So it should be easy to translate eg. gen2-emotion-recognition to C++.
Thanks, Erik