[mmapi] how many input buffer is supported by function "NvBufferComposite"

I use R28.2.
I gave 9 input buffers, but I could only stitch 8 buffers.
The last buffer cannot be stitched into the output image.
the parameters were correct.

#define MAX_COMPOSITE_FRAME 16
From the macro definition, I guess that up to 16 buffer can be supported

1 Like

Hi,
Please share test code so that we can build and run to reproduce the issue. Thanks.

Did you ever find an answer to this? I am having the same problem. It only works with 8 buffers.

I am using r32.3.1.

Would be nice if you guys could share the source for these routines. I have posted about another problem with NvBufferComposite with width and height problems.

https://devtalk.nvidia.com/default/topic/1070855/jetson-nano/nvbuffercomposite-width-amp-height-limitations/

Thanks.

Hi,
The limitation on TX1/Nano is 8 buffers. Please check CHAPTER 15: VIDEO IMAGE COMPOSITOR(VIC) in techinical reference manual:
https://developer.nvidia.com/embedded/downloads#?search=trm

The compositor implements much of the DirectX Video Acceleration 2.0 Enhanced Video Processor specification, including deinterlacing,
scaling, color conversion, proc-amp, and compositing for up to 8 input surfaces.

Sorry I thought I was on the Nano section and my question relates to Nano.

Never mind I found the same limitation in the TRM for Nano.

Thanks for pointing me in the right direction.

Hi,
TX1 and Jetson Nano are with same constraints.