Just by upsampling to 192, it’s not the same as rendering it on MQA capable DAC. It might sound good to you, but it’s not because of MQA.
Once again You really need to read up on what MQA does. Outside of the MQA bubble.
MQA is really (regardless of what the marketing says) lossy 24 bit 48 kHz the rest is upsampling. Rendering is MQA’s name for upsampling.
For someone not wanting to defend MQA You do put a lot of effort into it, not just here
I would like to know Your thoughts on the video from RMAF and how You think a company with a viable product would act in that situation ?
Sure, which part of the video or which claim?
I don’t have a single MQA file, don’t subscribe to Tidal and my main rig doesn’t support it. I’ve heard few files on DAPs, but I still couldn’t form my opinion based on listening.
The concept of MQA is interesting to me. Fairly smart people worked on it, it’s obvious. It uses some very clever tricks.
You’re wrong about the 24/48 upsampling part. For the demo example they used upsampled sound to 24/192 just because this is the optimal resolution for MQA folding. That might have confused you.
The point of MQA is to start with 24/192 native file and “fold” it into little more than CD size.
I’m a DSD fanboy, but we’re not going to see DSD being streamed from Tidal and Qobuz any time soon. DSD is the closest representation of an analog signal in the digital domain.
Sure, which part of the video or which claim?
Would a CEO of a company with a product that is viable (i.e. it does what they claim) act like that banging tables, shouting and interrupt ? I for one would be embarrassed if it was someone representing my company
The concept of MQA is interesting to me. Fairly smart people worked on it, it’s obvious. It uses some very clever tricks.
If it could do what they claim, I would agree with You. However it’s been revealed as simply clever marketing.
You’re wrong about the 24/48 upsampling part.
Please do search for more info on the subject, outside of MQA’s marketing babble.
For the demo example they used upsampled sound to 24/192 just because this is the optimal resolution for MQA folding. That might have confused you.
Where did You get that demo ? Because the MQA format as used on Tidal is only capable of carrying a signal of 24 bit 48 kHz. MQA in their distortion filed calls upsampling, unfolding. However it is the exact same thing.
The point of MQA is to start with 24/192 native file and “fold” it into little more than CD size.
Well the part of little more should be taken with a grain of salt. For instance if You look at a newish song which is heavily dynamicly compressed (they become larger as FLAC files) I’ve chosen Pray You catch Me by Beyonce. As a 16/44 FLAC file it has a bit rate of 779 kbps. As a true lossless 24/96 file the bit rate is 2638 kbps. As a lossy 24/192 MQA file the bitrate is 1411 kbps. If You analyse the spectrum of the file You can see that there are no frequencies above 22 kHz. This means that they’ve taken a file sampled at 44.1 kHz and upsampled it and telling MQA users that it is a 192 kHz file. So nothing in this file couldn’t be contained in a standard 16/44 FLAC file (at 779 kbps) but as a lossy MQA file they have to use 1411 kbps.
Don’t read too much into it. They had previous exchanges.
Some of the claims the presenter makes are also pretty outrages, like when he claims that technical incompatibility is DRM.
Ultimately it comes down to does it sound better than redbook 16/44.1. Where is the line, is it better than 24/96 natively? Even that might be irrelevant, since Qobuz is already streaming in 24/192. Some even claim that MQA is better than DSD.
We’ll see, people are not rushing to get new DACs with MQA. Still many new DACs at all price levels are coming with MQA support as standard. It’s too early to tell but it might end up like with HDCD.
For a guy having no vested interest in MQA You sure do defend them.
Previous exchanges or not, that behavior isn’t suitable for any grown person. Especially a CEO of a company that wants peoples trust (and money)
Regarding DRM, You are aware that part of MQA has been developed with one of the worlds foremost DRM specialists ?
Ultimately it comes down to a company that overpromises and underdelivers. Going back to an earlier post:
They started out claiming it was lossless, it wasn’t
They claim that when not using either an app or any form of decoder it retains CD resolution, it doesn’t
They claim to deliver 24 bit and up to 384 kHz resolution, they don’t
They claim to be Highres, it is not
They claim to correct timing errors, they are not
It’s not whether it sounds better than CD (a FLAC file at 1411 kbps should, but is still only a 24 bit 48 kHz lossy file) it’s the fact that MQA is promoting a product that just as well could have been an 18 bit 96 kHz lossless FLAC file, with which You could deliver the same quality except that the 18 bit file would be lossless and the MQA would still be lossy.
However You twist and turned it MQA is and will always be a superMP3. Come on the company is so desperate that they even have issued MQA vinyl ??? Never heard of any MQA RIAA’s
A good example of Your promotion of MQA is Your “mention” of upcoming products coming out, how is this relevant ?
On the subject of DAC’s, initially they told everyone that all MQA DAC’s would have individual filters for each DAC. Another point where MQA has overpromised. It turnes out that all MQA DAC’s share the exact same filters.(Some have more, some less. But none of them individual to any DAC)
It always comes back to the fact that if MQA is all they claim, why not show all their critics ?
You seem pretty stuck on the lossless part. It’s maybe not lossless, but it’s also not correct to call it “lossy”. Some parts of the spectrum are compressed lossy, but the most important part is compressed using FLAC losslessly. Would you call FLAC compressed redbook 16/44.1 lossless or lossy?
Of course they’re embedding the folds in the part they losslessly compress. Whether that compromises the quality when being played without a MQA capable system, remains to be seen. They surely are convinced that the “folds” are inaudible, so if they’re correct the second claim is also true.
The other statement are just facts, it’s pointless to debate this. The last one about the timing, well they sure try. Whether you or others believe they succeeded is another question, but that’s what they’ve set themselves to do.
Also the renderer part is specifically tailored to the target DAC. This is also a fact. That’s why you have separation of the rendering the decoding part. This is also how Audirvana works, it’s a software decoder and when coupled with a MQA renderer like Audioquest DragonFly you get complete system. There are also DACs that do both MQA decoding and rendering in the DAC itself, like Meridian Explorer2.
MQA is indeed like “Super MP3”. What’s wrong with that? If it sounds better than redbook at roughly the same bandwidth, it’s fine with me.
I don’t understand why some people are fixated in trying to prove them wrong on a technical level. Some even without the necessary technical skills. Every piece of technology has it purpose. This is also true for MQA. It’s designed to make music sound better than CD at roughly the CD bandwidth. The only question that matters is did they succeed in doing this, and I don’t mean on the oscilloscope. Does it sound better to your ears, is it more engaging, enjoyable, pleasure to listen to? If the answer is yes, then it’s a success.
Again, MQA is irrelevant to me at the moment. I’m ripping my CD collection and still getting new CDs. I’m following closely the developments simply because I’m interested in technology in general, beyond just music reproduction.
You seem pretty stuck on the lossless part. It’s maybe not lossless, but it’s also not correct to call it “lossy”.
Alex, You can not have it both ways, either it’s a full reproduction of the original recording (within the format i.e.16/44 etc.) or You throw away unretrievable information. MQA is the later. So in any way shape or form MQA is lossy
Some parts of the spectrum are compressed lossy, but the most important part is compressed using FLAC losslessly.
You are aware that of the 24 bits in the original file (before conversion to FLAC) the eight least significant bits are not compatible with standard PCM. They are perceived as noise. Of these eight bits 1 is used for telling the MQA DAC how many times the remaining 16 bits (at 48kHz) should be upsampled. Another to tell the DAC to turn on the “authentication” light. The sad thing is that the eight least significant bits are part of the “lossless” part of the file. Meaning that when the file is played back on nonMQA products it has more noise in it than the original CD quality.
Nobody other than MQA and those who have paid to know (under NDA) what exactly happens, but when analyzing the upsampled file it looks like the part above 22kHz is random noise that does not look anything like what was in the original hires file.
Would you call FLAC compressed redbook 16/44.1 lossless or lossy?
That totally depends on what is compressed. FLAC is actually a container that can contain anything audio. The difference between a standard FLAC file and a FLAC file containing MQA is that MQA throws away information before packing the file into a FLAC file. If You take an AAC or MP3 file and convert it to FLAC it will still be a lossy file. Since the information was thrown away before converting to FLAC it can not be retrieved by converting to FLAC. The exact same is true of MQA files.
Of course they’re embedding the folds in the part they losslessly compress. Whether that compromises the quality when being played without a MQA capable system, remains to be seen.
No it has been measured, and MQA files have a certain sound print that shows that MQA is not really transparent. Whether this is due to the loss of information or the MQA process, that remains to be seen. Sadly this can only happen with MQA’s collaboration, which so far looks like a dream.
They surely are convinced that the “folds” are inaudible, so if they’re correct the second claim is also true.
Second claim ?
The other statement are just facts, it’s pointless to debate this.
Statements ?
The last one about the timing, well they sure try. Whether you or others believe they succeeded is another question, but that’s what they’ve set themselves to do.
Oh, You must mean the leaky filters that they pass on as a new invention. These filters have actually been around for years. Nobody have really been using them since they create a lot of problems.
Also the rendered part is specifically tailored to the target DAC.
I actually shook my head in disbelief when I read this, Then I thought: He must mean that Audirvana and other playback products set the max output rate to whatever the DAC is capable of.
The renderer does not change the output signal dependent on the individual DAC other than set the signal to for instance 16/44 for a TDA based DAC.
This is also a fact.
Please show me any independent source that conforms this, so far this is purely a statement of MQA
That’s why you have separation of the rendering the decoding part.
Actually, the reason is that MQA needed to get more people to see their DAC light up 24/96. To get people believe the hype. The album I mentioned earlier actually shows up as a 24/96 file when played through Audirvana. If You look at the frequency spectrum it is quite obvious that it is actually a 44 kHz recording, despite this if played back on an MQA DAC the "authentication light lights up and claims that the MQA file is a 24/96 recording. Which shows that “authentication” is worth nothing
This is also how Audirvana works, it’s a software decoder and when coupled with a MQA renderer like Audioquest DragonFly you get complete system.
Let me ask You, have ever gotten ANY information about MQA from independent sources ? Hans Beekhuyzen, Stereophile etc. are not independent. If so please tell me which.
There are also DACs that do both MQA decoding and rendering in the DAC itself, like Meridian Explorer2.
Which was the way MQA initially wanted it, that way they would get more revenue.
MQA is indeed like “Super MP3”. What’s wrong with that?
The fact that MQA initially tried to sell MQA as lossless should tell You the significance of the fact that MQA actually does throw away information. All communication from MQA, after they were outed regarding the “losslessness” of MQA. If MQA wanted to be a superMP3 do You think they would have started out, marketing themselves as lossless ?
If it sounds better than redbook at roughly the same bandwidth, it’s fine with me.
Roughly ? If that is how You percieve roughly, I will gladly supply You with anything You would want, at roughly the same price as Amazon. An average CD track converted to FLAC is somewhere between 700-800 kbps the same track streamed as MQA from tidal is 1411 kbps. In my book, that’s not roughly the same. The funny thing is that MQA has marketed themselves as a format that saves bandwith.
I don’t understand why some people are fixated in trying to prove them wrong on a technical level. Some even without the necessary technical skills. Every piece of technology has it purpose.
Your right every technology has it’s purpose, it just seems that the purpose of MQA is making money.
It looks like they’ve sat down around a table an made up a pretend product. And decided to market the sh_t out of it. The reason why people are showing why MQA can not and do not work like they are advertising it. Is simply because several digital engineers around the world have looked at MQA’s claim to discover that it’s simply not doable.
This is also true for MQA. It’s designed to make music sound better than CD at roughly the CD bandwidth. The only question that matters is did they succeed in doing this, and I don’t mean on the oscilloscope. Does it sound better to your ears, is it more engaging, enjoyable, pleasure to listen to? If the answer is yes, then it’s a success.
You do sound exactly like the mails I’ve received from Mike Jabara, MQA CEO
Again, MQA is irrelevant to me at the moment. I’m ripping my CD collection and still getting new CDs. I’m following closely the developments simply because I’m interested in technology in general, beyond just music reproduction.
Well, You keep saying so. Even though You’ve posted in this thread 10, all defending MQA repeating marketing babble.
Stating that Your marketing quotes are facts, even though these have been shown by engineers, programmers and DAC manufacturers to not hold up to scrutiny. There’s even DAC manufacturers that have come clean saying that the only reason they include MQA in their DAC is to add another bulletpoint to their marketing material.
Look at Paul McGovan, PS Audio who is even a friend of Bob Stuart. He’s publicly stated that he does like what MQA does to the music.
The lossless/lossy confusion is due to the fact that they wanted to simplify the marketing. This was very stupid and naive. They should have known how OCD and detail oriented audiophiles are. What they claim to have meant is that it’s lossless in the audible part of the spectrum. It’s already difficult to explain MQA as it is, so I find this explanation credible and reasonable (though stupid).
FLAC is not lossy, it’s never lossy regardless of the compression factor:
https://xiph.org/flac/features.html
The oversampling part you keep on insisting is just not true. MQA is not about upsampling. Unless you want to claim that what they purport to be their design is just marketing and what they implemented is just an “upsampling algorithm”. This is very naive. They’re betting their reputation on this, and some of this things are not that hard to prove.
If the file unfolds to 24/96 it means it was encoded from 24/96. The lossy compression of the upper part of the spectrum is not designed to extend the frequency response beyond the audible part of the frequency range. They claim that this is done to retrieve timing information based on some psychoacoustic research.
The renderer part is supposed to apply artefacts and timing correction, it doesn’t further increase resolution.
I’ve read quite a bit about MQA from multiple sources, some more some less reputable, including quite a bit of negative opinions. With some of them I agree, with some I disagree.
I’m not trying to defend MQA, I’m just correcting statements that are not true. You have to ask yourself why are you attacking it so strongly. What’s your problem with it?
I have great respect for Paul from PS Audio. It’s totally fine what he says, he doesn’t like it and I wouldn’t be shocked if I didn’t like it (after I get a chance to explore it). He never questions the technical aspects of it, that it really does what they claim it does. He just says that the result is not audibly pleasing for him (and that DSD does it better). I hope that DSD wins the high end formats war.
There’s no question that MQA is designed to make money. The end goal is to make MQA a standard and them as gatekeeper to the revenue stream. They have a history in this, remember the MLP?
The lossless/lossy confusion is due to the fact that they wanted to simplify the marketing. This was very stupid and naive. They should have known how OCD and detail oriented audiophiles are. What they claim to have meant is that it’s lossless in the audible part of the spectrum. It’s already difficult to explain MQA as it is, so I find this explanation credible and reasonable (though stupid).
You gotta get of the coolaid, MQA/Meridian wanted to ride the wave of hires popularity (relative) so they claimed that they had “invented” a new format that contained the full hires file in a smaller container. Nothing about “perceivably lossless” as the mantra is now after being found out. That’s what in the entire EU is called false marketing.
FLAC is not lossy, it’s never lossy regardless of the compression factor:
https://xiph.org/flac/features.html
MQA starts out as a MODIFIED Wav file that then gets compressed with the FLAC algorithm. The MODIFIED Wav file is a lossy file. Because MQA discards ((throws away) information that can not be retrieved. Hence the resulting MQA in a FLAC container is lossy. Consider this, if You took an MP3 file and converted it to a FLAC file would this become a lossless file ? Of course not, the information that You know the MP3 algorithm throws away can not be retrieved by converting it to FLAC. It’s just the same with MQA. Information/sounds that is discarded can not be retrieved.
The oversampling part you keep on insisting is just not true. MQA is not about upsampling. Unless you want to claim that what they purport to be their design is just marketing and what they implemented is just an “upsampling algorithm”. This is very naive. They’re betting their reputation on this, and some of this things are not that hard to prove.
Well You may think it’s naive, but that is in fact what is happening. (According to manufacturer that has implemented MQA in their DACs)
If the file unfolds to 24/96 it means it was encoded from 24/96. The lossy compression of the upper part of the spectrum is not designed to extend the frequency response beyond the audible part of the frequency range. They claim that this is done to retrieve timing information based on some psychoacoustic research.
Here’s the funny thing - They claim - You have all Your “knowledge” from the MQA website and the orchestrated “reviews”. You really should seek information outside of marketing. All true 24/96 files have better timing than 16/44 files (if the DAC can retrieve it) but it is not retained when You downsample it and throw away parts of the track.
The renderer part is supposed to apply artefacts and timing correction, it doesn’t further increase resolution.
Do You actually know what the renderer is ?
I’ve read quite a bit about MQA from multiple sources, some more some less reputable, including quite a bit of negative opinions. With some of them I agree, with some I disagree.
That’s quite unspecific, could You tell which ones You disagree with ?
I’m not trying to defend MQA,
Oh !
I’m just correcting statements that are not true.
All statements from me have never been refuted by MQA (trust me they are very aware of this info that is out there.) One should think that if this information was not correct they would show that the information was wrong. After all it is a hinderance to their business.
You have to ask yourself why are you attacking it so strongly. What’s your problem with it?
Where to start ?
They started out claiming it was lossless, it wasn’t
They claim that when not using either an app or any form of decoder it retains CD resolution, it doesn’t
They claim to deliver 24 bit and up to 384 kHz resolution, they don’t
They claim to be Highres, it is not
They claim to correct timing errors, they are not
Basically because they are selling a scam.
I have great respect for Paul from PS Audio. It’s totally fine what he says, he doesn’t like it
You did notice that he said they only implemented it because they had customers ask for it. When a company buys into MQA they sign a NON DISCLOSURE AGREEMENT. Which means that they can not legally tell anyone about what they know about it. So Paul saying that he does like the sound of it is really strong words.
and I wouldn’t be shocked if I didn’t like it (after I get a chance to explore it).
Go get that Tidal test week or two, use the Tidal app (it’s a renderer) see what You think. When You listen be aware that a lot of the MQA encoded albums are slightly louder than straight FLAC (louder is perceived as better sounding to most listeners)
He never questions the technical aspects of it,
Because he can’t he’s under contract
that it really does what they claim it does. He just says that the result is not audibly pleasing for him (and that DSD does it better). I hope that DSD wins the high end formats war.
I for one hope for the loudness war to end. As things stand there’s no need for hires. A standard 16/44 file is capable of 96 dB of dynamic. Do You have any albums that exceeds that, I don’t. Funny thing, one of the very first CD’s released was ABBA has an average DR of 15. Currently most albums are between 3 and 7
There’s no question that MQA is designed to make money. The end goal is to make MQA a standard and them as gatekeeper to the revenue stream. They have a history in this, remember the MLP?
I do and look at the success of that platform, several consumer electronics manufacturers have stopped producing Bluray players, the latest I know of is Samsung
Which is probably why they dreamed up this scam, or at least a hideous case of false marketing.
I have a feeling that you’re opposing the format more on philosophical than on technical grounds, and that’s fine. I also have problems with it.
PCM was also burdened with patents and licensing that have expired or expiring. Manufacturers had no problem paying the licensing to respective patent holders. In that respect MQA is nothing new.
What is new is that MQA imposes licensing cost on the whole music delivery chain. I’m not convinced that the customers are willing to pay additional money for MQA beyond the cost for plain 16/44.1 redbook content. I doubt that the music industry is willing to swallow the cost of MQA without the ability to pass on the cost to the end consumer.
In MQAs own words renderer handles compensation and filtering of the following D/A converter, which, depending on the model, may contain a cascade of signal processing, upsamplers or other forms of conversion and filtering
This is the only place the upsampling is mentioned. This is what my DacMagic does normally with all content. Cambridge Audio is very proud of it and it’s one of the selling points.
Also in their (MQA) own words This step in the decode process will be different for every product, in order to make the resulting analogue conform closely to the hierarchical target and to most accurately replicate what was approved in the studio.
Obviously we’re in the area of diminishing returns, higher the resolution the more lossy it gets. This is to be expected, especially looking at the file size which is closer to redbook.
This discussion got me interested, so I experimented a little bit. I downloaded some files for 2L.no:
All those files originate from the same DXD master recorded using Merging equipment. It’s interesting comparison.
With MQA file reproduced on non MQA system I can’t really tell the difference between the MQA file and redbook file (you can download CD flac). When playing the MQA file through Audirvana on non MQA DAC I also can’t tell the difference compared to redbook. Few times I thought I heard difference, but it was so minute I wasn’t sure my brain isn’t playing tricks on me. With my equipment I can’t definitively conclude that there is difference, let alone an improvement.
Haven’t tried on MQA capable DAC, because I don’t have one, but it would be an interesting test.
Paul from PS Audio is comparing MQA with DSD. I totally believe that, but there is a 5-6 fold difference in bandwidth compared to DSD64, and more than 10 fold when compared to DSD128. At DSD256 we’re talking about 1GB per 5min. playback time file.
On he 2L site you can also find files to compare DSD, DXD (PCM) and MQA directly:
In the age of AutoTune, the question is does HiRes even matter. Does it really matter to pursue higher quality just to better hear performers computer generated voices? Sure, maybe for live performances, classical and jazz. For the rest, the redbook is probably plenty enough.
Anyway, I like a good discussion, so keep on going…
The Tidal app is not a renderer, it’s a decoder. Just like Audirvana. As far as I’m aware there are no pure software renderers.
Not really, I oppose due to the fact that almost each and every usp they presented originally has been debunked.
Well each and every manufacturer from recording to playback had to pay fees. However PCM was actually a solution to a problem, MQA is a solution looking for a problem.
Well that’s marketing for You, it has been shown by others that what is actually happening is simple upsampling with a fancy name.
If this was anything more than marketing fluff, MQA DAC’s would sound more alike. Why would anyone want to pay for expensive DAC’s
Well if You by this, mean the WAV file. I’m not impressed that they are able to “squeeze” a 15.something/48k file into a FLAC container “only” the size of a uncompressed WAV file.
If MQA have got such a fantastic product, they should play with open cards. Show the world how it’s done. Release files with/without the various stages they claim to have. Let us hear the differences.
Instead of going to audio shows banging on tables
That’s very naive to believe that. If it was just upsampling, the file would be the size of the FLAC plus few bits of instructions.
Pretty much every DAC on the market upsamples on it’s own, but even those that don’t do upsampling (NOS) support MQA.
Here is the example:
https://metrumacoustics.com/mqa
Sorry, first they upsample (or should that be unfold) then they add random noise and run it through shallow digital filters.
It seems that Metrum doesn’t even understand what MQA does.
That’s funny.
Without all the marketing repetition, then explain what is happening
There’s not much to explain beyond the fact that the first unfold is definitely not a simple upsample.
The upsampling comes into play in the last phase of the unfolding in the renderer where together with filtering can be used to achieve the desired result (which is correction for DAC artefacts).
This is the theory at least. Whether this produces better sonic results, you can be skeptical of that.
If you read carefully what Metrum writes, they don’t even do that. Here is what Metrum (not MQA) writes:
Blockquote
The second unfold corrects the impulse of the digital to analog converter as well as extracting the fine temporal detail of the original recording. This process is tailored because every single DAC has its own behavior. This is most beneficial for oversampling DACs that use steep filtering. The DAC’s oversampling filters are always applied just “above” the music and have a negative impact on the impulse behavior. To solve this, one can use their own up-sampling technology to push the DAC’s steep filtering up to higher frequencies. As a result the DAC’s filter is being detached from the music and causes less disturbance to the music. Gentle up-sampling can also create artefacts, but these are generally considered less annoying than steep filtering.Now let’s talk Non Oversampling. Per definition a Non Oversampling digital to analog converter does NOT use any filtering and is entirely focused at the perception of the human ear. As NON Oversampling DAC do not use any filters, in combination with MQA Core Decoder (first unfold), a NON Oversampling output will only benefit, and this is the case with every single DAC in the Metrum Acoustics range.
That doesn’t tell us what it does then? I’m curious, if it’s not upsampling, what is it ?
It is burring information below the noise floor in the audible part of the spectrum. If they simply upsampled they wouldn’t have to do that, wouldn’t they?
To be able to know more beyond that, you’ll need to get a job at MQA. 