On 10/08/2020 09:16, Tomi Valkeinen wrote: > To be honest, I don't even quite know how subsampled formats are supposed to be handled in DRM. > Above we pass width as it is, but divide height by ysub. And then we tune the bpp adjust to the fact > that we didn't divide the width. > > For e.g. YUYV, the bpp tells the container size. But for NV12's second plane, bpp is not the > container size. Ah, no it isn't... Not enough coffee yet. So the bpp for YUYV and NV12 is the "effective bpp", how many bits are consumed on that plane for each real pixel. Then, shouldn't bpp for YUV420's second and third plane be 4? Tomi -- Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki
