Use csi without i2c to get the streaming from FPGA. format is 800*800 60fps, RGB888, 4lanes

My SDK version
cat /etc/nv_tegra_release

R32 (release), REVISION: 6.1, GCID: 27863751, BOARD: t186ref, EABI: aarch64, DATE: Mon Jul 26 19:36:31 UTC 2021

Refer to this URL

see dts
pixel_t = “rgb_rgb88824”;

but check item "pixel_t "
https://docs.nvidia.com/jetson/l4t/index.html#page/Tegra%20Linux%20Driver%20Package%20Development%20Guide/camera_sensor_prog.47.1.html

pixel_t Deprecated. Use the following properties instead:
•mode_type
•csi_pixel_bit_depth
•pixel_phase

Does this setting work properly?
Does the SDK officially support the RGB888 format?

hello channinglan,

that may still works since the code still keep the flow to parse pixel_t property.
however, you should try to avoid using pixel_t since it’s already deprecated.

VI did not support RGB888 memory formats due to it’ll also need luminance formats.
please also refer to similar discussion thread, Topic 1067424 as see-also.
thanks

In SDK32.6.1
Can these items be used to set RGB888?

•mode_type
•csi_pixel_bit_depth
•pixel_phase

"
VI did not support RGB888 memory formats due to it’ll also need luminance formats."
This means gst-launch-1.0 only use v4l2src ,not use nvcamerasrc (Xavier ISP )??

hello channinglan,

you can use them to device sensor properties, but RGB888 still cannot supported.
while you’re having gst-launch-1.0 you can have different plugin to enable different pipelines. for example, v4l2src, and nvarguscamerasrc.

I mean whether it can be modified to accept FPGA CSI output RGB888 and then use the gst-launch-1.0 v4l2src command to display?

I don’t understand "RGB888 still cannot supported. "

hello channinglan,

please check [RGB Memory Formats] session via Xavier TRM,
the memory buffer will add luminance format, for example, it will use T_A8R8G8B8 instead of RGB-8-8-8 to store the frames.