Mnyb;597221 Wrote: > The trick is to the record the stuff at 24/192 or higher and digitally > re-sample to a consumer format , theese filter does way more damage at > ADC end than the DAC end. And as most DAC's oversample 128 times and > use a very benign filter in the ultrasonic region. So this is actually > a solved problem. >
Well, let's be clear, the problem of *how* to do the filtering efficiently and accurately has been solved, but the fact of doing the filtering is in itself a problem: you are losing something if you do and losing something else if you don't. The plain facts are as follows: Our Red Book CDs are sampled at 44.1kHz. Whatever process was used to get to 44.1kHz sampling rate must have made an attempt to first satisfy Nyquist's sampling theorem by removing *all* spectral content above 22.05kHz. I don't care if that process was done digitally at very high sample rates or otherwise; the fact remains that a filter has to have been used that provides very good attenuation above 22kHz. The closer this filter is to a brick wall in shape the longer the impulse response tails (and pre-echo) will be. But there is very little latitude for making this filter 'relaxed' in the frequency domain because we want to have the passband extend to ~20kHz, and we want to be ~90dB down by 22.05kHz. This implies a high order filter, and an extended (some would argue unnatural) impulse response. There are still some choices to be made, but I think all of them are non-ideal, and yet at the same time indispensable. A NOS DAC attempts not to make the problem worse, but you must still live with the choice of decimation filter the mastering engineer chose to use. -- dsdreamer ---------------------- "Dreamer, easy in the chair that really fits you..." ------------------------------------------------------------------------ dsdreamer's Profile: http://forums.slimdevices.com/member.php?userid=12588 View this thread: http://forums.slimdevices.com/showthread.php?t=83976 _______________________________________________ Touch mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/touch
