MQA Controversy

Audio Components DACs
1143

Here is a comparison I would be interested in hearing about from owners equipped with the DSD (Huron), DMP and Bridge II:

Take a good-sounding Redbook CD and compare the sound of the CD played with the DMP to the same HIFI and MASTER files through the Bridge II. Given all of the variables that have been aired here ad inifinitum, it will probably take a number of different comparisons to reach a conclusion.

I can’t control anything in the production chain, but I can decide to buy a CD or not. With the DMP, I’ve found that the the SQ of a decent CD absolutely betters the HIFI SQ via Tidal. So, basically, how does MQA via the Bridge II compare?

I stream via an Aurender, but don’t have a bridge and I’m trying to figure out if the view justifies the climb.

Thanks, all!

1144
badbeef said

If the best thing out there now is some squashed version of the Mix Master made for CD in the 80’s, …


The squashing is done during mastering.

1145
Elk said ...
The MQA ideal is to share the sound of the "master" in the purest, most traditional sense--the output of the recording studio, that final mix that the artists listened to and gave signed off on. . . . This pre-mastered version--aka "the master", the "core master", the "flat master"--is typically quieter than the CD release, and it hasn't been "improved" by a mastering engineer.

This is the target version for MQA encoding/release.

This very much surprises me and seems unlikely. And I, for one, would not want the final mix as the end result other than as a curiosity to compare with the mastered version.

Here's a preview of a more recent interview with Stuart that I expect to publish soon. Remember that Ludwig is an MQA supporter:
The flat master is often the best-sounding thing you can get your hands on because it has high dynamic range, tends to not be quite loud enough, and yes, that would then go to a mastering step, and when that was released, it went to Ludwig when he was at Sterling Sound, and it was mastered for CD. That made it louder and compressed it. Other releases are done like that. ...
1146
I suppose a caveat is in order. There's no shortage of MQA critics around who, in order to win arguments, have suggested all sorts of malfeasance and dishonesty on the part of the MQA folks. Maybe what I'm telling you is all nonsense. Maybe it's all a scam. Maybe we'll drive the MQA used car home only to find it rusted out and heavily Bondo'd. Maybe they lie whenever they open their mouths. I find this implausible--but there you go. If one chooses to think they're lying, I can't convince them otherwise, and I have no interest in trying. But I think it's much more likely that they're honorable people with an idea they believe in who are trying to turn it into a viable business. I could be wrong.

Jim

Jim, I don't think it was developed as a scam, but in their zeal to sell it, they have made claims that have changed, many times. And that they are trying to be the format, needing manufacturer's DACs to implement it, is an odd way of doing things.

At first they had to implement it all in the DAC, but then the first unfold was able to be done with software. If they had not got Tidal onboard, I don’t think it would have generated anywhere near the discussion it has. I still think they should sell a piece of hardware that could be inserted between the source and any USB DAC. If they can sell the Explorer II for $199, then their little box could be priced at a similar price, with a pass through for non MQA files. That would make sense. Many of us have DACs we won’t replace just to get MQA. If I could afford a new DAC it would be a Directstream, so I could add the Bridge II, but I have no desire to buy any other DAC. So I will utilize the first unfold through Tidal’s buggy player. The amount of available music makes the small aggravations of their player worth living with. I think it is a shame that Tidal turned down JRiver, as I would prefer to have one player, to mix my music with theirs.

I am a few issues behind, but I started your review of the BHK preamp. I could not believe you were bothered by one button for the inputs, when you have direct access with the remote. My CJ CT5 has the same configuration. I believe you redeemed yourself further in the review. Please take it as just a little honest criticism. I do appreciate you participating in this and the Asylum’s forums. I’m sure you have knowledge, dealing directly with the manufacturers, that we will find of great value, if you choose to share it.

1147
JimAustin said Here's a preview of a more recent interview with Stuart that I expect to publish soon. Remember that Ludwig is an MQA supporter:
Thanks, Jim. i really appreciate the background information you are bringing to this thread.

In an unfortunate way this makes sense. If the final master is for certain going to be over compressed and otherwise compromised, the final mix/master is going to sound the best. A sad reality.

What bothers/concerns me is that there are many final masters which have been improved by having been given to a mastering engineer, both older recordings as well as current. A great modern example is Daft Punk’s Random Access Memories, not coincidentally beautifully mastered by Bob Ludwig. This is a spectacularly good sounding album, even with the amount of compression used - warm, deep, gorgeous bass, an enveloping sweet high-end. The EQ is awe-inspiring. (I have great respect for Bob’s work and his efforts to avoid over-compression and the loudness wars whenever he is given the freedom to do so.)

It appears that essentially everyone understands that excessive compression is a bad thing. In my perfect world, the final mix would go to a good mastering engineer and this final, gorgeous, polished version would then be processed by MQA.

But if MQA succeeds in getting final mixes to process, some recordings will sound better than the subsequently squashed final master. Others will sound inferior to the well-executed final master.

Most of this is theoretical for me as I listen primarily to classical and compression is used sparingly in this genre, but I would like all recordings to sound their best.

1148
jeffstarr said I am a few issues behind, but I started your review of the BHK preamp. I could not believe you were bothered by one button for the inputs, when you have direct access with the remote. My CJ CT5 has the same configuration.
I'm with Jim on this. I like clear, dedicated hardware buttons. I have nothing against remotes, but prefer direct non-remote access for all critical functions. I find being forced to pick up a remote annoying. Ideally the remote simply duplicates the existing hardware.
1149

Jim said: “The squashing is done during mastering.”

Apparently I didn’t state that clearly. I was saying that if the only version that exists is a squashed version (that is, a compressed final mastering) of the Mix Master, then an MQA remastering from that same (uncompressed) mix master has a better than even chance of being better, even if done by an algorithm.

Independent of all of this SQ stuff is the issue of what shakes out going forward. Let’s say your favorite label is a signatory to MQA, and they encode all their NEW material with it. Then, if you want to hear the full unfold, you MUST have an MQA device capable of doing it. That has been one of the issues all along. If everyone buys into it, we may be stuck with it.

Keep your CDs, folks, they will be worth hundreds on the black market in the future ; )

1150
badbeef said I was saying that if the only version that exists is a squashed version (that is, a compressed final mastering) of the Mix Master, then an MQA remastering from that same (uncompressed) mix master has a better than even chance of being better, even if done by an algorithm.
50/50 may be about right, assuming that the final mastered version is extremely dynamic limited. But, as you of course know, mastering involves so much more than dynamic compression that often the less compressed final mix will not sound as good as the compressed mastered version.

We also need to remember that mixing a recording involves a lot of track by track compression, especially on bass, electric guitar, vocals, and drums. This is the sound of a pop/rock/popular jazz recording. It is not as if the final mix is uncompressed before it goes to the mastering engineer.

I also have trouble accepting that the artist/producer is gong to sign off on the final mix, not care what happens in final mastering, and will be perfectly OK with a less compressed, lower volume MQA version. The artists want overall compression. They are driving it.

Mastering engineers would love not to heavily compress a track/album. I have yet to read an interview or see a post by a mastering engineer declaring squashed sounds better or stating “Even if the artist does not want it, I squash everything I can get away with and convince them this is what they want.”

The remarkable uncertainty of what you my or may not be getting with an MQA file serves only to make me even less comfortable with the concept. And frankly I do not want to hunt through and compare each MQA file to other recordings of the same album. If MQA is valid and truly makes a recording sound better we should be able to rely on the MQA file to possess better sound.

1151
jeffstarr said

I am a few issues behind, but I started your review of the BHK preamp. I could not believe you were bothered by one button for the inputs, when you have direct access with the remote. My CJ CT5 has the same configuration. I believe you redeemed yourself further in the review. Please take it as just a little honest criticism. I do appreciate you participating in this and the Asylum’s forums. I’m sure you have knowledge, dealing directly with the manufacturers, that we will find of great value, if you choose to share it.


I was not bothered by the little plastic button for the remote, or not exactly. However, I feel I have a responsibility as a reviewer to point to issues that might matter to some readers. I own a couple of old-school components, bought new within the last few years, that don’t have remote control at all. If you’re used to that–if that’s your aesthetic–then it might matter to you that apart from a volume control and a standby button (and the on/off switch on the back), the only control on the BHK pre is a little plastic button. The Stereophile readership spans a wide range of tastes. I try to alert them to issues that might affect them. (The intermittent noise on changing volume is a similar issue.)

Thanks for the kind words and the feedback.

Jim

1152
badbeef said Let's say your favorite label is a signatory to MQA, and they encode all their NEW material with it. Then, if you want to hear the full unfold, you MUST have an MQA device capable of doing it. That has been one of the issues all along. If everyone buys into it, we may be stuck with it.
There are a few small labels whose whole workflow is MQA aware, but even they don't use MQA as their archival format. In the 2014 conference paper that appears to outline the philosophy of MQA, Stuart and Craven made a clear distinction between archival formats and distribution formats, putting MQA in the latter category.

Distribution formats is a different question. I’m frankly happy to have access to these masters, whatever the format. I suspect most download companies (if they continue to exist) will continue to offer choice; I’m not that familiar with (eg) Unamas and Eudora, but the MQA-centric label 2L offers a wider range of download options than any other company I’m aware of, from mp3 to DXD. Streaming is another matter: For now, in the U.S. MQA seems to be the only option for high-res streaming. I suspect they’ll have competition soon (with Quobuz in Europe streaming high-res directly) but that’s mere speculation, and anyway, will they have access to the same master files? There’s reason to doubt it, since that’s part of MQA’s value proposition, part of what won the enthusiasm of record companies: It allows the companies to guard their “crown jewels”, the masters of those great albums.

And we should all probably keep in mind that streaming is a bit like online publishing: It wipes out traditional approaches (record stores, physical media) without necessarily providing a viable alternative. Not sure how Spotify is doing financially, but I do know they’re impoverishing their musicians. Tidal is apparently losing money; here’s hoping they hang on, expand, and soon turn a profit.

In such a climate, I’m happy for what I can get.

I’m gonna add one more point here. I was chatting a few days ago with an excellent (and proximate :wink: digital engineer, who was telling me about how reducing ultrasonic noise can improve sound quality. (Think Huron.) When you consider that music almost never extends above about 70kHz (and in the most extreme cases levels at those frequencies are tiny), only a handful of microphones can go higher (few of which–I’m thinking Neumann, are often used in recording music), and production tweeters never go above about 50kHz, there’s a compelling argument to be made that any sampling frequency above about 100kHz (or in extreme cases maybe a little higher) just RECORDS MORE ULTRASONIC NOISE, the kind of thing Ted has been working to get rid of. So explain to me again why it matters that above the first fold, MQA is slightly lossy.

Jim

1153
JimAustin said So explain to me again why it matters that above the first fold, MQA is slightly lossy.
Why, and if, it matters if we record and playback over the threshold of hearing remains a a fascinating question. By the time we reach adulthood and can start to afford expensive music reproducing toys, many of us cannot hear above 15kHz. One would this conclude 44.1/16 would be more than sufficient fidelity to capture anything we can hear. Yet, many (most?) of us claim we can hear the difference of higher sampling rates and better than 16-bit resolution.

I suspect it is not so much the frequencies themselves, but how we can push noise, dither, filtering anomalies, etc. into high enough frequencies they do not impact the audio band.

I predict at what point extended accurate frequency response and bit-depth becomes academic will long be debated.

1154
JimAustin said For now, in the U.S. MQA seems to be the only option for high-res streaming. I suspect they’ll have competition soon (with Quobuz in Europe streaming high-res directly) but that’s mere speculation, and anyway, will they have access to the same master files? There’s reason to doubt it, since that’s part of MQA’s value proposition, part of what won the enthusiasm of record companies: It allows the companies to guard their “crown jewels”, the masters of those great albums.
Jim,

Could you expound on this last point? What are the “masters” in this context and how does MQA help guard them? An how does this win over record companies?

I suspect I am missing something very basic.

1155
Frode said
Elk said
Yes - but to Jim's point - if it sounds better, it is better.
To the individual - yes, but not necessarily from a higher fidelity point of view.
'High fidelity' contains the seed of our doubt. Faithful… to what?

If the MQA process(es) are doing something worthwhile then the owners of MQA should be up front and supply demonstration files, clearly stating that the source of (say) a 16/44 version, a 24/192 version and an MQA version is one and the same, and of course giving the original bit depth and sample rate. Why don’t they? What are they afraid of?

Peter

1156

A great source of demonstration files is the 2L site, MQA, different formats, different resolutions. Click

1157
Elk said
JimAustin said So explain to me again why it matters that above the first fold, MQA is slightly lossy.

Why, and if, it matters if we record and playback over the threshold of hearing remains a a fascinating question.


I think you’ve missed my point. I’m arguing for preserving all the music–not all that we can hear. I said nothing about cutting off after 20kHz. I said, anything about 60-60kHz is demonstrably NOT music, as you can tell by watching the (rough) 1/f response descend into the noise floor. Above that frequency, musical information, if there is any, is far below the noise floor. So why keep it?

1158
Elk said

Jim,

Could you expound on this last point? What are the “masters” in this context and how does MQA help guard them? And how does this win over record companies?

I suspect I am missing something very basic.


I’ve never fully understood this “part of MQA’s value proposition”, as I put it. As to what the “masters” are, I’m using it the same way now as before: these are the best version of a piece of music that a record company has on hand. Often, Stuart told me, these are pre-mastered versions: “root” masters or “flat” masters.

Here I’m just repeating what others have said and assuming they know better than I, since I’m not an expert on the economics of the music business. In practical terms the point is simple: If the absolute best file you have is, say, 192/24, it’s probably best not to put that out in the wild where people can share it and adulterate it and do what they want with it. You can guard your masters (by not distributing them), just as in the analog era you guarded your master tape, and yet have the benefit of distributing very good-sounding tunes (equivalent to a high-quality analog-era LP). But then, MQA is supposed to sound better than the original master because the timing errors are fixed … so then I get confused. But if you ignore that claim and say they only sound as good, or just a hair away, then it kind of makes sense.

Assuming this is true–assuming record companies see holding back the master files as a good thing, and so see value in distributing a proxy for those files instead of the files themselves–well, MQA provides that proxy, so they might not want to allow another company to come in and distribute the master files themselves, which they’ve already decided they shouldn’t do. Understand?

1159
JimAustin said
Elk said Why, and if, it matters if we record and playback over the threshold of hearing remains a a fascinating question.
I think you've missed my point.
Not at all. Your point is straightforward. I am making a broader observation, not merely responding to you. :)
. . . assuming record companies see holding back the master files as a good thing, and so see value in distributing a proxy for those files instead of the files themselves--well, MQA provides that proxy . . .
Hmm . . . This is an interesting concept, particularly in light of MQA's claim their processing makes a file sound better. If true, who would want the company's original "best" file anyway? But you explained the idea well. Your reference to Quobuz potentially having access to "master" files is what confused me (final mix v. mastered). Thanks!
1160

In all the current discussion, about protecting the master, improving on the master, etc. I think there is one think being ignored.

M[aster] Q[uality] A[uthenticated] Part of the selling point, was that album, the MQA version was going to show you a little colored light verifying that it was the quality of the master. The master for us layman has always been considered the finished product that was being sent out for release. We kind of understand that the pro cutting the lacquer for LP release has to make adjustments so the vinyl is playable, and that if the CD plant gets a 24/96 master they have to convert it to 16/44, but they both are working off “the” master. Like I said a page or two back, that could mean a copy, or second generation tape when sent to other countries. Remember I used the Beatles as an example, the Parlophones being superior to the Capitol release. With straight DDD releases the quality of the master sent out world wide should be identical. Remember when CDs had, still might that three letter code, with all the jazz albums I so love, it would be either ADD, or AAD. My understanding has always been the first letter represented the recording medium, the second, the mixing medium, and the third was that it ended up as a digital product. The third kind of being obvious, unless it was really representing the mastering.

The other colored light on MQA releases was supposed to tell us whether the band/artist/ producer had approved the MQA release as the master,as they had created it. I don’t see that being mentioned anymore, or people with MQA DACs claiming which releases get that signed off on light to come on. Mostly I read that a light has come on.

So, my take has been that with the MQA, is that it is supposed to sound the same as the master. Any corrections for timing and deblurring having to do with the digital conversion made after an analog master has been made. Over at the Asylum there was a lot of discussion that HDCD discs had better sound, not so much from the HDCD processing, but that the superior sounding Pacific Microsonics AD had been used, rather than the inferior Sony AD converters. That the PM is still used but without any of the HDCD processing. The light or letters HDCD still show up in players that have the ability to decode HDCD as that marker goes into the code automatically, when PM AD is used. And that after, I think it was around 2009 that the HDCD processing was not turned on as very few players could process it. I know it has something to do with Microsoft buying the HDCD process, and basically putting it on the shelf.

The thing is that until recently only analog masters were archived in DSD or at higher bit rates like 24/192. Digital albums were mostly recorded at 16/44 or 16/48. Only audiophile labels were recording at 24/96 or higher.

And finally I thought it has become pretty much accepted that the real benefit of higher resolutions is that it pushes the brickwall filter up from 22k where it had a negative effect on the sound of digital for playback. The other benefits have more to do with mixing. One of the reasons DACs convert to DSD or offer upsampling which does not magically get back those ultrasonic data, is that allows the DAC to use different filters at a higher resolution, once again ridding us of the nasty 22k brickwall filter. That is one of the reasons the Directstreams sound so good, along with many other benefits that Ted has developed.

1161
jeffstarr said M[aster] Q[uality] A[uthenticated] Part of the selling point, was that album, the MQA version was going to show you a little colored light verifying that it was the quality of the master. The master for us layman has always been considered the finished product that was being sent out for release.
This had been my understanding as well. it never occurred to me that MQA would skip the final step of mastering and, instead, would be of the mix. I find this to be bizarre as the mix sounds very different after it has been mastered, even without any compression. While many would enjoy getting to hear the final mix of their favorite album - without mastering - often listeners are going to be disappointed. A mix is a lot more raw, less open, less polished, etc. than they expect. There is a reason good mastering engineers are highly respected and exceedingly busy.
. . . it would be either ADD, or AAD. My understanding has always been the first letter represented the recording medium, the second, the mixing medium, and the third was that it ended up as a digital product. The third kind of being obvious, unless it was really representing the mastering.
The SPARS code represents analog or digital for the three steps of producing a CD: 1) recording, 2) mixing, and 3) mastering. ! and 2 can be either analog or digital, and 3 is always (at least as i have seen) digital for CDs. (Although mastering could be accomplished in the analog domain with only an ADC converting it to digital - this strikes me as analog mastering even if it ends up on a CD. Anyway . . . )
The other colored light on MQA releases was supposed to tell us whether the band/artist/ producer had approved the MQA release as the master,as they had created it.
This is my understanding, but it was never clear to me whether this was a new approval of the MQA version, past approval of a previously released version, or something else. As a practical matter, it seems highly unlikely a bunch of artists are going to listen to sit and listen to the new MQA file for the purpose of providing approval. Thus, I assumed this "approval" was past approval of a released mastered version, not of the mix.

I am frustrated with MQA’s marketing and claims. There is no reason they cannot be transparent and directly tell us what they are up to. More and more MQA seems like yet another means to convince music lovers they need to buy yet another version of recordings they already have.

1162
jeffstarr said

In all the current discussion, about protecting the master, improving on the master, etc. I think there is one think being ignored.

M[aster] Q[uality] A[uthenticated] Part of the selling point, was that album, the MQA version was going to show you a little colored light verifying that it was the quality of the master. The master for us layman has always been considered the finished product that was being sent out for release. We kind of understand that the pro cutting the lacquer for LP release has to make adjustments so the vinyl is playable, and that if the CD plant gets a 24/96 master they have to convert it to 16/44, but they both are working off “the” master. Like I said a page or two back, that could mean a copy, or second generation tape when sent to other countries. Remember I used the Beatles as an example, the Parlophones being superior to the Capitol release. With straight DDD releases the quality of the master sent out world wide should be identical. Remember when CDs had, still might that three letter code, with all the jazz albums I so love, it would be either ADD, or AAD. My understanding has always been the first letter represented the recording medium, the second, the mixing medium, and the third was that it ended up as a digital product. The third kind of being obvious, unless it was really representing the mastering.

The other colored light on MQA releases was supposed to tell us whether the band/artist/ producer had approved the MQA release as the master,as they had created it. I don’t see that being mentioned anymore, or people with MQA DACs claiming which releases get that signed off on light to come on. Mostly I read that a light has come on.

So, my take has been that with the MQA, is that it is supposed to sound the same as the master. Any corrections for timing and deblurring having to do with the digital conversion made after an analog master has been made. Over at the Asylum there was a lot of discussion that HDCD discs had better sound, not so much from the HDCD processing, but that the superior sounding Pacific Microsonics AD had been used, rather than the inferior Sony AD converters. That the PM is still used but without any of the HDCD processing. The light or letters HDCD still show up in players that have the ability to decode HDCD as that marker goes into the code automatically, when PM AD is used. And that after, I think it was around 2009 that the HDCD processing was not turned on as very few players could process it. I know it has something to do with Microsoft buying the HDCD process, and basically putting it on the shelf.

The thing is that until recently only analog masters were archived in DSD or at higher bit rates like 24/192. Digital albums were mostly recorded at 16/44 or 16/48. Only audiophile labels were recording at 24/96 or higher.

And finally I thought it has become pretty much accepted that the real benefit of higher resolutions is that it pushes the brickwall filter up from 22k where it had a negative effect on the sound of digital for playback. The other benefits have more to do with mixing. One of the reasons DACs convert to DSD or offer upsampling which does not magically get back those ultrasonic data, is that allows the DAC to use different filters at a higher resolution, once again ridding us of the nasty 22k brickwall filter. That is one of the reasons the Directstreams sound so good, along with many other benefits that Ted has developed.


Rich post, lots to respond to. Not up to it now. Will try to remember to reply tomorrow. Best.