Of course, I meant while rendering. However I would think that keeping it in YUV, with an upsampling scheme like Canopus does, would probably be better, since every time you are going from one color space to another you have the opportunity for quantization errors.
>> However I would think that keeping it in YUV, with an
>> upsampling scheme like Canopus does, would probably
>> be better, since every time you are going from one
>> color space to another you have the opportunity for
>> quantization errors.
True, but comparisons between Canopus-DV codec and SoFo-DV codec states SoFo DV has no disadvantages there.