Right now the software-conversion is in libcamera (but not in the convert utility), and the scaling-based hack is in convert (but not in libcamera).
The hack only works when the X component is the highest byte in address order, and when no other scaling is required.
So if you are using C1 this won't be of immediate use.
FWIW, the software conversion is done in situ in the DMA buffer (which is allocated as 32bpp for this purpose).
The hack only works when the X component is the highest byte in address order, and when no other scaling is required.
So if you are using C1 this won't be of immediate use.
FWIW, the software conversion is done in situ in the DMA buffer (which is allocated as 32bpp for this purpose).
Statistics: Posted by njh — Mon Dec 01, 2025 11:00 am