As an addendum to the post from @DDPS , RME has a bunch of WAV files in order to test Bit perfect playing. This (short) WAV files are configured with different bit depth (16, 24 and 32 bits) and different sample rates (44,1 kHz, 48 kHz, 96 kHz and 192 kHz).
If you play these sample test files on your player right to an RME ADI-2 DAC, this unit shows a message on the display when it identifies the (bit perfect played) sample.
If you play any of these samples with Roon, with NO PROCESSING, the bit perfect is shown (obviously 32 bit samples does not play on SPDIF). So, Roon plays bit perfect: What is played with Roon arrives bit perfect to the DAC.
Again,If you play any of these samples with Audirvana, with NO PROCESSING, the bit perfect is shown too. So, Audirvana plays bit perfect: What is played with Audirvana arrives bit perfect to the DAC.
Conclusion, with NO PROCESSING on any of both players the same info is transmitted end to end. So there is no difference to be perceived.
The metadata is to a great extent out of Roon’s control , its 3rd party contracted out to Tivo and MusicBrainz . Every time someone reports an error I understand it’s reported back for correction.
Given the number of albums involved it is a mammoth task
What type of music are you listening , maybe some of the more obscure genres are worse than the main stream rock and classical . I listen a lot of classical but don’t find many examples of incorrect metadata.
Advanced filtering I use mainly in the area of classical music and I use Roon primarily for this purpose - unfortunately, it often fails me and I have to laboriously google manually…
That’s an admirable goal, but really, something that needs to be learned from engineers, not from salespeople peddling wwhatever it is that they have to sell today.
Of course they would have you believe that, how would they differentiate their product otherwise?
The only way to rule out the psychoacoustic reason for perceived sound difference is to remove the “psycho” part fron the equation. In absence of a blind test showing that the difference can be heard, it is perfectly safe, and really completely reasonable to assume that the science and engineering that keeps satellites from falling down on our heads, nuclear power station from exploding (more often than they do, anyway) and everything else working is not wrong, and there is no real difference in SQ whatsoever.
Perhaps I’m going to fast here but can I conclude from your writing that when a path is bit perfect then it’s not possible to hear any differences in sound beside psychoacoustic differences ?
Nothing else matters besides better besides bit perfect ?
I don’t know, I’m not knowledgeable enough on it and some of the equipment I’ve seen that supposedly remedies this, whilst affordable to me, is not something I’d want to invest in.
I have seen videos from some of the people you named regarding timing and jitter.
I’m mainly a headphone listener, an owner of several headphones from the likes of Meze, Focal, Fiio and a couple of headphone DAC setups plus a number of dongle DACs for hotel use etc.
So room acoustics do not matter to me.
2 of the people you mentioned are not headphone listeners and have speaker setups, with physical room treatments, DSP room treatments, DSP bass traps in one case. Clearly, what they hear is not bit perfect, it’s treated according to their room and personal taste.
This is the rabbit hole I alluded to earlier. I’m lucky to be satisfied with what I have.
With timing I don’t mean playing with DSP or special setting but the timing accuracy of data moving from a to B. ( timing includes jitter and smearing)
To be honest at first ( 5 years ago) i was also skeptical about timing.
Keeping an open mind and listening to equipment vendors like Grimm audio who presented a few times on the importance of timing for digital audio helped me to gain knowledge that I later implemented in my audio system As the brits say “ the proof is in the eating of the pudding “
To make a long story short. Timing is not just important, it’s crucial
This happens in the DAC for any DAC worth discussing and hence is decoupled from the timing in the source or on the wire, be it Ethernet or USB. The next sample is either in the DAC input buffer when the DAC clock fetches it or it isn’t (in which case there is an error)
Hi Hans, I have an open mind, maybe one day if I am lucky to demo an MU2 I will experience it myself.
I wanted to point out and, I’m sure you know, that there are many things in the chain to influence the sound, including the room, furniture in a speaker setup.
Back to the original post, now Audirvana have a Linux install I may try them out, I’ve been listening to LMS on a headphone setup this weekend rather than Roon and been very happy with what I hear.
It’s a wonderful hobby and good luck with your journey.
This makes technically zero sense. The unpacked raw samples are clocked out of the buffer, how else?! but OK you „learned“ from the usual suspects, so I’m not surprised
