The r2r thread got me thinking, so if we feed a DAC a bit perfect FLAC/WAV file but that DAC uses filters is the sound still bit perfect or a true representation of what was in the original source file?
Don;t want to be picky- but this question is a bit confusing. The bits are in effect the sample values, which should strictly be the sample values of the original signal after it has first been band-limited. Even then they are probably not actually the values recorded by the original adc (which will almost certainly nowadays be some form of delta sigma design not running at the distribution format sample rate or bit depth). It might be the same set of sample values as were spat out when the ADC first converted to PCM.
So bit perfect probably means that the bits are not altered relative to the distribution format (16/44, 24/96 etc).
This could be true up to the point that the bits are converted to analogue, but it has no meaning when applied to "the sound".
As
@John Phillips points out, the filter is necessary to convert the sample values back into the original signal. If done properly the whole adc/dac loop can be accurate to an arbitrary degree ie the output can be made to be about as similar as you can be bothered to the orginal band-limited signal.
A dac could convert to analogue without any digital filter using only an analogue filter. or no filter. If it uses a digital filter this will by definition involve changing the bits such that the bitstream will not be "bit perfect". However that is not the same as saying that the filtered bitstream is not an accurate representation of the original (pre-sampling sigal) or even the original sampled bitstream,. It will almost certainly be a more accurate representation than any analogue filtered output let alone an unfiltered dac output. Ultimately that's all that matters.
Also, find it strange that ASR get all excited about SNR when anything better than 90dB is moot. A much better measurement or something I'd be more interested in is the settling time of the DAC or measurements for its slew rate, delay and settling time
Hmmm, I'm not really following. The first bit -yup dacs can basically be made accurate to an extent which exceeds the limits of human hearing.
The next bit is a non-sequitur. I'm not sure what exactly you want. if you look at a J test, it measures what a dac does when producing a sgnal of minimum level and near- maximum level at the same time. If you look at what a competently designed dac does when faced with such a signal, the answer is essentially- do exactly what the bits say and nothing else. On recent designs you can resolve one bit toggling on and off in a 24 bit system. ie at -144dB. You can also look at what the dac does when faced with 32 different tones, and again the answer is pretty much nothing down to -120dB or lower in the audible range. (You will always have some amount of distortion in the real world). of curse one could always measure for something else, but do not be taken in by the flagrant and oft-repeated lie that only one measurement is taken, or by the fantasy that there is one measurement around the corner which will show that an LP12 has a more accurate output than a topping dac.