Or putting it another way, at what frequency does the filter (linear phase, minimum phase etc) operate at.
By going from 16bit to 24bit, SOX doesnt do dither so thats one avenue of potential degradation eliminated.
So the next avenue is the filter and due to my ignorance, I have to assume that when going from 44.1k to 48k that the filter works at 48k (or 24k in the audio band).
If I look at an upsampled WAV file using SPEC, there is no audio content between 22.05k and 24k so this means that if the filter operates at 48k, whatever "damage" it does is way above the original CD filtering/dithering point at 22.05k (and is therefore inaudible)
But if the filter works at 44.1k, then there is the potential for compounding any filter "damage" at the same point as the original CD filtering/dithering point.
So just need a simple answer from minds greater than mine: Does the filter work at 44.1k or 48k in my use case.
Thanks,
Peter
By going from 16bit to 24bit, SOX doesnt do dither so thats one avenue of potential degradation eliminated.
So the next avenue is the filter and due to my ignorance, I have to assume that when going from 44.1k to 48k that the filter works at 48k (or 24k in the audio band).
If I look at an upsampled WAV file using SPEC, there is no audio content between 22.05k and 24k so this means that if the filter operates at 48k, whatever "damage" it does is way above the original CD filtering/dithering point at 22.05k (and is therefore inaudible)
But if the filter works at 44.1k, then there is the potential for compounding any filter "damage" at the same point as the original CD filtering/dithering point.
So just need a simple answer from minds greater than mine: Does the filter work at 44.1k or 48k in my use case.
Thanks,
Peter