judmarc
100+ Head-Fier
- Joined
- Jul 11, 2011
- Posts
- 372
- Likes
- 271
No need. Benchmark already ran this test as part of their DAC1/ADC1 development. They ran the DAC1 into an ADC1 looped over and over 20? times I believe and in the end... absolutely no change to the data. Meaning yes the DAC1 is literally bit perfect. Which is not to say that it is pleasant. Also not to suggest that bit perfect relates to stage focus, smoothness, dynamics or any other none numerical, measured performance.
I don't think this can be literally true. The DAC1 was set, if I recall correctly, to do something called asynchronous sample rate conversion (ASRC) as an anti-jitter measure. Sample rate conversion (plus the accompanying necessary anti-alias filtering) is mathematically not a bit-perfect process. Nor, for that matter, is conversion from digital to analog or the reverse. These processes are also technically "lossy," meaning once the data have been converted, there is no mathematical process for perfectly reconstructing them.
So parts from the ADC1 and DAC1 may have been tested, or a testing configuration different than the consumer unit (without ASRC) may have been used, or a test may have been run where the results were audibly indistinguishable from analog input, but I do not think the circumstances as presented - a full ADC and full DAC, each doing sample rate and digital-analog/analog-digital conversion - could mathematically obtain the same data stream as input.