MQA claim they apply corrections for the decoding hardware. Certainly you could unfold completely in software but the theory is you need to know what the DAC is for the full effect. MQA suits mass production and some uniformity of architecture but a good DAC will always be just that. Partially unfolded at 96k still has the potential to sound good, MQA DAC or not.
Ah, thank you,
Now I understand the analogy, made in another thread I think, with photographs, and correcting for the specific printer-model
He also made the analogy of correcting for camera and lens. Does that mean that MQA files try to correct for recording hardware/equipment in the case where they have detailed information about that?
This makes MQA different from, and much more than, hi-res files then, more interesting I would say.
(Presumably the correction for the recording equipment is dac-independent, and that should be possible to do completely on the software side?)
(And it also makes the discussion of lossy vs lossless kind of moot, doesnât? We are actually talking about correcting the file, i.e. altering it.)
1 Like
If a loudspeaker is incapable of reproducing 25 kHz due to high frequency roll off, then impulse response rise time will be affected.
AJ
High frequency roll off where?
It is not so much on the âringingâ (The ear is less sensitive to this area) but the amplitude of the impulse response, at 48kHz, it losses more than 70% of the available amplitude (energy)!. Transient response will suffered as form of âsmearingâ and âblurringâ. It takes PCM 384kHz (DXD) to come close to analogue and DSD. âRingingâ is a time domain and it can happen anywhere in the audio range 20Hz-20kHz. So your theory of having a speakers roll off above 20kHz to attenuate the âringingâ is incorrect.
That is not due to 48 kHz sampling, per se. Both the reduction in amplitude and widening in time of the impulse are products of low pass digital filtering. Take away the oversampling digital filter, and even at 44.1 kHz, the impulse response looks a lot like that of DSD.
(The DAC in question inverts polarity, by the way).
AJ
That is not my âtheory.â It is a fact that the ringing is from digital filtering and occurs at the Nyquist frequency. If 44.1 kHz sampling, the ringing is at 22.05 kHz.
AJ
Yes I agreed that, most DACs in the market employ 8x over sampling digital filter at 44.1/48kHz. This is a problem unless you can bypass it like my Holo Spring DAC. But by doing that, you got unfiltered sampling noise which will degrade the SNR. So for most DAC with over sampling digital filter this issue will lessen when the sampling rate increases.
I will be interesting to see how MQA is able to lessen the âringingâ effect at the same time maintaining most its energy at its peak amplitude. Could someone show me that?
AJ
I came across this type of ringing before, it shifts the pre ringing to post ringing. The result is all the energy in the pre stage is now transferred to post stage, resulted in even higher ringing.
Iâm interested in the original impulse peak amplitude input with relative to the output from this device and see how much has been attenuated. Does this response correspond when this device decoding MQA or just a normal PCM response from the DAC?
In case there are BluOS owners new to Roon reading this thread, it is a known problem that Roon is NOT passing MQA data correctly to BluOS devices - at least not end-to-end. BluOS-supplied playback software lights the blue MQA light (figuratively - as there is no physical LED to light on BlueSound devices), but Roon does not. This is contrary to the way Roon behaves with all other MQA DACs.
The silence around this problem is deafening from both BlueSound and Roon. Neither company will say a thing, other than that they are both âworking on itâ - which could mean anything - as it has been quite a number of weeks since they acknowledged the problem. My guess is that this is a low priority for BlueSound. Roon has a history of quickly responding to software problems, so at this point I canât help but surmise that their hands are tied, as far as actually solving this problem.
Actually I would suggest that the BluOs devices decode MQA from within their own software. As Roon circumvents this software the BluOs device simply becomes a dumb end point. That to me doesnât suggest a big conspiracy of silence, just the matter of commercial confidentiality and a pause until Roon does its own decoding or a Roon/BluOs workaround is approved by MQA. Speculation on my part again but more plausible.
Yes, I agree wholeheartedly about the importance of time alignment â I will never go back to standard loudspeakers (my Alexiaâs are certainly time-aligned). Would we be having this discussion 10 years ago (about the whole digital time resolution domain) when the majority of devices out there just didnât have the horsepower to measure such things?
And isnât jitter a cousin to what weâre talking about here? The degree to which clocks can accurately sample seems to be just as much to blame.
PS Audio is adding software unfolding to their Bridge 2 (Roon Ready optional Ethernet port for DirectStream Dac). It will do at least the first unfold, and maybe add some custom filters for MQA playback. PS Audio is not charging for the firmware upgrade. They expect to have it ready soon (weeks, not months).
1 Like
This shows a Minimum Phase filter being deployed. Is my personal preference. Pushes the pre-ringing to post-ringing. Pre-ringing, or sounds before the sound donât exist in nature, so post ringing in audio is much more natural sounding. Less smearing, better transient attacks, much easier to listen to IMHO.
1 Like
Yes I know. I still want a hardware MQA solution. So I bought a Mytek Brooklyn and will be selling the DirectStream Dac and my Explorer2.
I also bought the Mytek Brooklyn about 6 weeks ago. It does sound fantastic with Tidal MQA streaming, but I wouldnât want to give up my DirectStream. The Brooklyn is a great value and a very versatile Dac.
SGB,
I too have a DirectStream DAC I bought about 2 1/2 years ago and it does sound awesome with CD quality (and above) AIFF & FLAC files.
But I must say, streaming MQA files from Tidal in âpassthruâ mode using the Meridian Explorer2 DAC does sound better than streaming MQA files from Tidal with the âpassthruâ box unchecked.
From what Iâm reading, there are two benefits to MQA, 1) embedding into 48k/24 bit files higher resolutions than that, and 2) the âdeblurringâ that is done, if your DAC is an MQA-capable DAC.
What PS Audio has in mind, adapting the software in the Bridge2 to do the unfolding will take care of benefit #1 (just like unchecking the âpassthruâ mode on Tidal). In my experience, âunfoldingâ the MQA files and sending them into the DirectStream DAC sounds âworseâ than playing the âregularâ 44.1k/16 bit FLAC files.
I think you may find that if you canât enjoy âbenefit #2â you wonât find MQA will sound âbetterâ.
Minimum phase digital filtering has become more prevalent over the last several years. The idea behind shifting all pre ringing to post ringing, while not embraced by all in the know, has become popular.
As a caveat, unless there are any exceptions, minimum phase filters are not linear phase. That means they exhibit differential group delay â all frequencies enter the filter simultaneous but not all frequencies exit the filter simultaneously. So, if pre ringing is a linear phase time domain tradeoff, then differential group delay is a minimum phase time domain tradeoff. As for the superiority of one tradeoff or the other, that will continue to be debated.
I posted this article in another MQA thread. Anyone who is interested in technical and/or audible differences between digital filters should read it. It is a well balanced article.
AJ
1 Like
I think it will be hard to judge until PS Audio has implemented the solution. The Bridge 2 is a very good Ethernet input, and IMO better than any USB interface. Ted Smith has also suggested he will be developing filters for MQA playback. Ted is very talented in this area, and it may be possible for him to provide a better âDac profilingâ for âdeblurringâ on the Directstream than MQA is capable of. But as always, there is no way to judge other than by listening. For the time being I am enjoying Tidal MQA on the Brooklyn, If the PS Audio solution works well Iâll redeploy the Brooklyn to another system.