As I have already said, I am not disclosing the method and the hardware used, unless we have some test results. I think you might be surprised. I am aiming at the usual suspect against ABX test with digital files. I can only make you sure that the files are not identical and that DAC and ADC are used in the test.
If you look at these in DeltaWave, the PK metric comparing @pma 's files is about -62dB. So that is possibly audible. Certainly difficult for me to find a difference, though. In his update, Josh reports PK metric of better than -100dB (and again, too bad he won't give up the files!). If we can't hear a difference between PMA's recordings, it is hard to believe that differences Josh heard were in fact differences in the DACs where they match 100X better.
I have similar results.
Very good. If you get 10/16 three times, then you have statistically valid positive result (I need to check). But it must not be a cherry-picking, like neglecting all the bad results and showing only the good ones. The testing may be done per partes. You sum up all the positive results vs total number of trials.
The P value on my report is 22%. Does that mean by guessing I would expect to do as well 22 out of 100 trials? If so I would say much more likely than not I was just lucky.
10/16 is still possible only by chance, easily. p<0.05 (5%) is requested at least, based on probability theory and statistics.
But only for the unattended path through the comparator. For the attenuated comparison of one DAC to itself, it's around - 70 dB (median) / - 58 dB (max). So much, much worse and also surprisingly bad. For some reason he did not report the PKmetric RMS value and I think it was also discussed that he measured the comparator incorrectly using the high voltage 2.5 mm input on the Cosmos, but that was all a bit unclear to me.
DeltaWave allows for clock synchronization by adjusting for constant clock drift and by removing the initial delay offset with accuracy of a tiny fraction of a sample.
Yes, I get that. I just thought that the drift compensation can never be as good as having synchronization. Have you seen -100dB matching without synchronization?
PKMetric as a single number is an RMS (average) over the whole recording, so there could be much larger components to it than -100dB - a graph of PKMetric is a better way to look at it. In fact, -100dB PKM speaks pretty poorly of the A/B channel matching in the ABX box.
Leaving out the comparator, what would you expect for PK Metric for two consecutive recordings, same track, the same DAC and same ADC. But not with sychronized DAC and ADC, just having timing and clock drift corrected by DeltaWave? And I should apologize for asking you instead of just trying it myself!
Depends.
.
Two consecutive recordings with the same circuit and a consistent clock drift between DAC and ADC should produce PKMetric in the -120 to -130dB range assuming the noise level isn't high and doesn't fall in the more "sensitive" frequencies where it would affect the result more.
I did not even dare to compete in ABX; my brain stops all analysis when confronted with this music choice; swapped some twenty times between the files for seconds, but no difference audible for me.I did 2/16 with speakers
To the point I enjoyed my spectacular fail .
If only it was music I can enjoy, once at an ABX!
On a serious note, I know nothing about the work or artists.
But unless some (even deliberate) very bad (by artist's side) recording is going on, the DACs (or whatever pma used) sound REALLY poor, reminds me of Low's b-sides (but those are actually fun)
.