advertisement


MQA

Status
Not open for further replies.
tuga said:
It wouldn't work.
You'd need to compare de decoded file with the original file which was used to create the MQA file.
You'd need trained listeners.
And you'd need to set up an iron-clad and fair methodology for the listening test (level-matched, double-blind).

Once you go subjective it stops making sense...
Click to expand...
It's an easy excuse to avoid comparison.

2L has made carefully handled and custody maintained files EXACTLY for purposes of SQ comparison. They are well respected and there is no reason to mistrust them.

Besides, if MQA files, once decoded, analyzed and listened to, still show obvious numerical flaws and SQ problems, so much the worst for MQA.
 
Heckyman said:
... listening tests would be pointless. It's like DSD vs PCM. Or DAC filter A vs filter B. Some people would prefer one and some the other. It would depend on the music too.
Click to expand...
In this case I am proposing a hybrid methodology that combines both subjective and objective evaluations.

I would find it very interesting to look over MANSRs analysis that may show an increased noise floor at a certain time stamp due to MQA processing, I would go and listen to that location at high volume to see how audible it is.

If the opinions broke half way, that would be valuable as well - it would validate my assertion that MQA competes with LPCM within "range of audiophiles" personal and system preferences."
 
Last edited:
DimitryZ said:
It's an easy excuse to avoid comparison.

2L has made carefully handled and custody maintained files EXACTLY for purposes of SQ comparison. They are well respected and there is no reason to mistrust them.

Besides, if MQA files, once decoded, analyzed and listened to, still show obvious numerical flaws and SQ problems, so much the worst for MQA.
Click to expand...

That's your opinion.
My opinion is that MQA is technically flawed and thus not worth listening to if I am able to opt for high-res or 16/44.1 instead.

I could make a similar case for mp3. If a majority of test listeners were unable to identify obvious numerical flaws and SQ problems would you accept and adopt mp3?
 
tuga said:
That's your opinion.
My opinion is that MQA is technically flawed and thus not worth listening to if I am able to opt for high-res or 16/44.1 instead.

I could make a similar case for mp3. If a majority of test listeners were unable to identify obvious numerical flaws and SQ problems would you accept and adopt mp3?
Click to expand...
So you can only pay attention to the numerics that MANSR will provide. And you can substitute fact for opinion, if evidence is supportive.

I am no majority (rarely have been). I will accept what I like for SQ, though I do listen to obviously flawed formats (like 8-track and VHS) because I believe they provide important historical insight into music production and listening.

But I am as curious as the next guy to see what exactly MQA does with a waveform in an actual music piece. So if that is now available, it would be fun to see.

I will look at the charts and go listen as well.
 
Last edited:
Zombie said:
Doesn't Dimka like that you reproduce copyrighted silence?
Click to expand...
Are you trying to be friends with me again, by using a familiar version of my name?

What is *Zombie" in Russian, anyway? I think it's this:
оборотень
I think I should call you that. A English synonym is "changeling."
 
Last edited:
FWIW I'm currently doing an analysis of various MQA files. One preliminary result is currently here
http://jcgl.orpheusweb.co.uk/temp/MQAspectra.png
This shows a time averaged FFT spectrum of a section of one of the GO test files. Note this if what you'd get when playing them with a *NON* MQA system. That's because one of my concerns is the impact of MQA encoding for people who don't have an MQA decoder and may find something is only available as MQA.

Mentioning this now this seems in accord with Mansr's examination of the same file. I'm using totally different software to do my analysis. Currently doing more examinations of other files, inc the 2L examples. Hope to make a fuller account public soon. Thus far, regard the above as 'preliminary'.
 
Jim Audiomisc said:
FWIW I'm currently doing an analysis of various MQA files. One preliminary result is currently here
http://jcgl.orpheusweb.co.uk/temp/MQAspectra.png
This shows a time averaged FFT spectrum of a section of one of the GO test files. Note this if what you'd get when playing them with a *NON* MQA system. That's because one of my concerns is the impact of MQA encoding for people who don't have an MQA decoder and may find something is only available as MQA.

Mentioning this now this seems in accord with Mansr's examination of the same file. I'm using totally different software to do my analysis. Currently doing more examinations of other files, inc the 2L examples. Hope to make a fuller account public soon. Thus far, regard the above as 'preliminary'.
Click to expand...

Well well well, here we have independent corroboration of Mansr's results... Preliminary as you explain. The similarity is remarkable!

MQA adds HF audible noise to any silence in a music file. If confirmed as real, this result is (should be) devastating for MQA.

If MQA are monitoring this thread, you need now address this.... Head on, no double talk will be accepted.

DimitryZ will of course try his hardest, but we need Bob to step up here now.
 
mansr, please share more. We need to document all the data you may have. Perhaps if you have a nicely prepared paper/document?
 
Jim Audiomisc said:
FWIW I'm currently doing an analysis of various MQA files. One preliminary result is currently here
http://jcgl.orpheusweb.co.uk/temp/MQAspectra.png
This shows a time averaged FFT spectrum of a section of one of the GO test files. Note this if what you'd get when playing them with a *NON* MQA system. That's because one of my concerns is the impact of MQA encoding for people who don't have an MQA decoder and may find something is only available as MQA.

Mentioning this now this seems in accord with Mansr's examination of the same file. I'm using totally different software to do my analysis. Currently doing more examinations of other files, inc the 2L examples. Hope to make a fuller account public soon. Thus far, regard the above as 'preliminary'.
Click to expand...

Jim, or @mansr , are you able to show the spectra of a source file converted without upsampling/filter (NOS)?
 
mansr said:
I suggest you start by reading this: https://audiophilestyle.com/ca/reviews/mqa-a-review-of-controversies-concerns-and-cautions-r701/
Click to expand...
I've just read this and noticed you were acknowledged very prominently. So I would like to congratulate you and Archimago for maintaining a professional tone in this paper and bringing these major issues with MQA to the uninitiated audiophile like me. I haven't followed up as yet to see what McGill University, Montreal, QC, Canada found on MQA, but if anybody has a lowdown I'd be interested to read it.

I think Archimago has clearly shown that it's the filter that MQA must impose that creates the issues and rather than deblurring anything MQA adds delays. Which leads me to conclude the end-game for MQA must be DRM (in the future for sure).

I am even more adamant now ....actually worried...that as an audiophile music loving community 'we' may have inadvertently helped MQA establish itself (probably because of Bob's reputation with Meridian). But 'we' must now do all we can to stop MQA gaining any further ground. We must actively appose it and cancel our Tidal subscriptions. I personally don't have Tidal, but a good friend and my daughter do, I will be persuading them to ditch it at their earliest convenience as soon as tonight!

I now see more clearly why Rosewind brought this thread to PFM and kept it alive for so long and I'm so glad he did. It seems there is still time to prevent MQA establishing itself as the end-to-end DRM that it will be at the expense of HiFi, which is our passion, raison d'être.

Archimago's article was written in Feb 2018, kudos to Audiophile Style for publishing it. Fellow PFM members who have similar fears of our music being locked into an inferior format, we must act and announce this widely to stop MQA before 75% of our music is mutilated forever.

We must keep the pressure up. GoldenOne, you are a hero along with mansr, Jim audiomisc, Rosewind and Archimago and co.

Those on the fence, especially with an independent platform like Darko need to take a side now and do something. Inaction and dithering (not digitally speaking) will not be forgotten (by me at least)...we need to know if you're actually on the consumers' side and whether you care about HiFi or not.
 
nostromo said:
Well well well, here we have independent corroboration of Mansr's results... Preliminary as you explain. The similarity is remarkable!

MQA adds HF audible noise to any silence in a music file. If confirmed as real, this result is (should be) devastating for MQA.

If MQA are monitoring this thread, you need now address this.... Head on, no double talk will be accepted.

DimitryZ will of course try his hardest, but we need Bob to step up here now.
Click to expand...
You have clearly been well out of the loop.

I have long confirmed, years ago now, that MQA is a poor universal delivery system, through listening tests alone. That this is confirmed through technical analysis is of academic interest.

My actual technical interest is how MQA compares to LPCM, when properly decoded - i.e. unfolded and rendered.

Now that MANSR has cracked the MQA playback, I hope that people with relevant technical abilities will perform actual analyses of music files with proper pedigrees and chain of custody - 2L files. They should also publish final decoded MQA versions as LPCM, so those without MQA hardware can judge for themselves.

When "cost of MQA" has fallen to $35/DAC, focusing on undecoded MQA is interesting but practically irrelevant.
 
Status
Not open for further replies.


advertisement


Back
Top