format support

[Manoa1911]

48 khz stereo 16 bit wave runs for 2 hours and then outputs garbage instead of telling the user the input is not supported at the start :(

44.1 same thing, something is wrong
running now mono 32 khz...
it worked after updating python stack :)

no documentation on which inputs are and are not supported, but the code looks like the program only works on 44100 hz
so next question which bit depths are supported ? if input depth is 24 and model depth is 16 does it give any higher quality ?
and if sampling rate of input is 96 and output 96 but the model is 44.1, does it give any higher quality ?
does model sampling rate affect output sampling rate ? can I inference using a 44.1 model to output 96 khz ?

[JusperLee]

Our pre-trained model only supports a 44.1 kHz input sample rate. For other sample rates, you would need to train the model yourself.