|
|
Post by rhale64 on Sept 21, 2017 19:05:38 GMT -5
I think that MQA hides information about the probable dynamic range and perhaps even higher frequency structure of reality vs the base recording medium capabilities, say a certain type of tape for instance. To recover with this additional information added in, MQA must expand the original 44.1khz file to a 96khz file containing new information and then expand that to 192khz to satisfy Niquist rules for errorless playback for the corrected file. Another fold due to other real world vs recording limitation may be added on top of this expanding the 192khz fill to 768khz. The improving information is hidden in the lowest bit or 2 of each word that is never heard when the file is played on a non MQA device because it is below the latent noise level. I'm trying to guess in terms we can easier understand because the MQA explanations written by Stuart are extremely complex. I think this guy is a true genius! And sir what you said is the best guess of what is probably going on. That actually makes a lot of sense. |
|
hemster
Global Moderator 
Particle Manufacturer
...still listening... still watching
Posts: 51,948
|
Post by hemster on Sept 21, 2017 19:08:31 GMT -5
I think that MQA hides information about the probable dynamic range and perhaps even higher frequency structure of reality vs the base recording medium capabilities, say a certain type of tape for instance. To recover with this additional information added in, MQA must expand the original 44.1khz file to a 96khz file containing new information and then expand that to 192khz to satisfy Niquist rules for errorless playback for the corrected file. Another fold due to other real world vs recording limitation may be added on top of this expanding the 192khz fill to 768khz. The improving information is hidden in the lowest bit or 2 of each word that is never heard when the file is played on a non MQA device because it is below the latent noise level. I'm trying to guess in terms we can easier understand because the MQA explanations written by Stuart are extremely complex. I think this guy is a true genius! I'm not sure if your assertion about hiding information is substantiated with actual evidence. I will say that having met the man, Stuart is a genius. But unlike most geniuses, he is actually very easy to talk to. |
|
|
|
Post by mgbpuff on Sept 21, 2017 19:28:23 GMT -5
I have no idea how the improving information is encoded and decoded, that is the great secret of MQA. But knowing Stuart is a pure designer and engineer, I believe this is actually being done. Why not give me some of your own thought unadorned by your skepticism?
|
|
|
|
Post by monkumonku on Sept 21, 2017 19:52:34 GMT -5
I have no idea how the improving information is encoded and decoded, that is the great secret of MQA. But knowing Stuart is a pure designer and engineer, I believe this is actually being done. Why not give me some of your own thought unadorned by your skepticism? I can't give you my thoughts unadorned by skepticism because everything I have read so far is what makes me skeptical. You yourself can't explain what is really going on, and it seems like no one else can either, except that somehow bytes are put into a magic box and come out transformed by mumbo jumbo. Also, no one so far in the Lounge or the reviews I have read can verify they heard an improvement by a true A/B test, that is, by using the exact same source but the A subjected to MQA and the B not subjected. There is no confirmation that what you are listening to is not a result of a different mix. MQA also seems resistant to scrutiny by experts such as Dr. Aix. That's why to me it seems no different than the guy who is selling the next secret to making a killing in the stock market. Or someone trying to sell you stock in a business, the purpose of which they cannot clearly explain. As for genius, the Wizard of Oz was considered a genius. And indeed he was - he was a genius at marketing himself. |
|
hemster
Global Moderator
Particle Manufacturer
...still listening... still watching
Posts: 51,948
|
Post by hemster on Sept 21, 2017 20:00:09 GMT -5
I have no idea how the improving information is encoded and decoded, that is the great secret of MQA. But knowing Stuart is a pure designer and engineer, I believe this is actually being done. Why not give me some of your own thought unadorned by your skepticism? I have already weighed in with my experience thus far with MQA. I found certain albums and tracks to sound decidedly better yet with others I couldn't tell a difference. So for me the jury is out. I have not ruled out this as a gimmick, nor will I run out and buy it.. just yet. |
|
|
|
Post by novisnick on Sept 21, 2017 20:22:19 GMT -5
I have no idea how the improving information is encoded and decoded, that is the great secret of MQA. But knowing Stuart is a pure designer and engineer, I believe this is actually being done. Why not give me some of your own thought unadorned by your skepticism? I have already weighed in with my experience thus far with MQA. I found certain albums and tracks to sound decidedly better yet with others I couldn't tell a difference. So for me the jury is out. I have not ruled out this as a gimmick, nor will I run out and buy it.. just yet. Im with you on this if you hadn’t noticed my friend. |
|
|
|
Post by rhale64 on Sept 21, 2017 20:45:04 GMT -5
I have already weighed in with my experience thus far with MQA. I found certain albums and tracks to sound decidedly better yet with others I couldn't tell a difference. So for me the jury is out. I have not ruled out this as a gimmick, nor will I run out and buy it.. just yet. Im with you on this if you hadn’t noticed my friend. I am too but I tell you that that Beyonce album in MQA is one of the best sounding of any genre that I have ever heard. I don't even really care for most of her stuff. But that is absolutely incredible sounding. |
|
|
|
Post by jefft51 on Sept 21, 2017 20:53:13 GMT -5
So am I understanding that they may be remastering and then hiding (folding) information several times so when you play it back, each "unfold' sounds better than the previous? Instead of remastering into a high bit rate/depth in the first place? That seems sketchy.
|
|
|
|
Post by novisnick on Sept 21, 2017 21:07:02 GMT -5
So am I understanding that they may be remastering and then hiding (folding) information several times so when you play it back, each "unfold' sounds better than the previous? Instead of remastering into a high bit rate/depth in the first place? That seems sketchy. I think that perhaps its folded to save bandwidth and make streaming hirez more available to more people, especially those with limited access to bandwidth. Im guessing and im sure somebody will be along shortly to correct me if Im wrong. 😋 More information, less packets? |
|
|
|
Post by mgbpuff on Sept 21, 2017 21:07:57 GMT -5
People seem to think MQA is a lossy compression scheme, but from what I've read it seems to me to be an expansion scheme instead which maintains the original file, adds corrective new information to that original file in a noise hidden codec and then upon play the original file is expanded to a higher sample rate that now contains all the new information. It could have been decoded into a higher sample rate file before transmission, but why do that since it is easier to transmit the lower undecoded sample rate file and the final result using MQA decoding DACs is the same.
|
|
|
|
Post by jefft51 on Sept 22, 2017 0:11:27 GMT -5
I didn't think bandwidth for audio was a problem, especially when there are compression schemes (lossless) out there that don't need a special DAC and licensing. Most people can stream Video, even on their phones over a cell connection and that takes a lot more bandwidth.
What does "noise hidden codec" mean?
How this all works is speculation though afaik.
|
|
|
|
Post by mgbpuff on Sept 22, 2017 7:46:44 GMT -5
Well the driving force behind MQA is NOT bandwidth minimizing, it's corrections for the original master recording's known limitations, adding temporal and psycho acoustical information that is lost in normal digital recording practice, and then restoring that information along with the normal digital sampling method presently used. How is it done? Is it snake oil? I think the AES would not accept Stuart's papers if it were nonsensical and not scientifically valid. So I may have made some dumb assumptions in how it works and is actually doing something just as many of you are making assumptions that it doesn't add anything except perhaps a new type of distortion that makes it sound different but not necessarily more accurate. So, is MQA scientifically based or not; that is the question!
|
|
|
|
Post by Boomzilla on Sept 22, 2017 10:40:05 GMT -5
Hi garbulky - I read your PM very closely. it said: "...if you feed an MQA file to the Dragonfly it ain't going to play MQA. It will just play regular PCM like any other DAC. The only way is to use the Tidal software or the Audirvana Plus 3 software to specifically decode MQA before it gets to the DAC." This implies (clearly) that if I have Audirvana Plus 3 AND the Audioquest Dragonfly, that I will get MQA analog output. So I downloaded the "full" trial version of Audirvana Plus 3. I played my MQA-encoded disc through the Audirvana Plus 3 program (NO iTunes - NO jRiver - just straight from the CD to Audirvana Plus 3 and output to the Dragonfly Red WITH the MQA update). What did I get? 44.1 CD-quality PCM output (green indication on the Dragonfly). I've also checked and double checked the settings per your link. Still no MQA. ALSO - NEWS FLASH - APPLE USB PORTS - NEWS FLASH I attended the Baton Rouge Macintosh Users Group (BRMUG) meeting last night and learned the following: If you have a MacBook Pro and you exceed the amperage demand limit on the USB ports (by charging more than one device at a time, for example, via the laptop's USB ports), the hardware in the machine disables ALL USB ports on the machine. They will not reset until the machine has been rebooted. After I rebooted my machine, my AudioQuest Dragonfly was immediately recognized in every USB port that I tried it in. So, just a FYI - If your Apple USB ports stop working, consider rebooting. Now one other peculiarity - While my MacBook pro was NOT recognizing the Dragonfly, the Dragonfly itself still worked! Despite not being "recognized" by the laptop, it still output analog music (but without any volume control available) and had a green status light, showing that it WAS outputting 44.1 PCM audio. Further, while the Dragonfly was plugged into the laptop, all other outputs (including the optical TOSLINK) were automatically disabled. I don't think that this particular anomaly is of any serious significance, but if you don't know what's happening, it can be confusing. So just a FYI... Finally, and this is back to MQA - Apparently the TIDAL desktop client and the Audirvana Plus 3 software are able to DECODE ("first unfold") MQA in their SOFTWARE. Otherwise, TIDAL, for example, would have to stream already decoded MQA audio. This would defeat the "packing" CODEC in MQA and lead to larger (MUCH larger) files. Therefore, if these softwares (TIDAL desktop client and Audirvana Plus 3) are DOING the decoding, then all that remains to get MQA analog out would be an MQA renderer, such as my AudioQuest Dragonfly. Now as I understand it, having a "full-MQA" DAC, such as the Mytek Brooklyn or the Meridian Ultra, would eliminate the need for the TIDAL desktop client or the Audirvana Plus 3 software. Why? Because those DACs can do BOTH the decoding AND the rendering in their hardware. If that is the case, then why would ANY player software (including iTunes or jRiver) NOT work as an MQA source? After all, the software is NOT changing the MQA-encoded PCM file - only sending it on to the "full-MQA" device that would then do both unfolds. In other words, with a "full-MQA DAC," the software needs to do NOTHING. All decoding and rendering is done in hardware, yes? Yet I'm told over and over that iTunes and jRiver are NOT MQA compatible. Why not? Boom |
|
KeithL
Administrator
Posts: 10,224
|
Post by KeithL on Sept 22, 2017 10:49:40 GMT -5
OK...... the metaphors are beignning to mix more and more... and the smoke it swirls... MQA is already a combination of vague and confusing - with a few nuances that really make your head hurt - and Cary Audio's use of certain terms makes it worse. In this case, they seem to be conflating several claims: 1) Being able to do full MAQ decoding properly. 2) Specifically, being able to decode MQA files up to 768k, while most other DACs can't handle sample rates above 192k (seems like a good thing). 3) UPSAMPLING MQA files (and other files) to a very high sample rate (768k) regardless of what they are to begin with. It's that third (exceptional) claim that I have a sort of problem with. UPSAMPLING DOES *NOT* CREATE NEW INFORMATION.
You CANNOT magically transform a 44k file into a 768k file. Well, to be strictly true, you CAN do just that.....
HOWEVER: 1) If you upsample the file accurately, then it will contain no additional information. In other words, you'll end up with a much bigger file that is no more accurate and sounds exactly the same. 2) You can deliberately apply specific filtering or other processes to ALTER the file as you upsample it.
Your alteration will be created by adding new information.
You CANNOT "get back information that was lost"; the new information you add will simply be new information.
(It can be based on a guess about what should have been there, or simply an idea about something that will sound good.)
In other words, if I start with a 24/192k file and convert it to 16/44k, I'm going to have to throw away some information.
THAT INFORMATION IS GONE AND CANNOT BE RECOVERED.
If you accurately upsample that 16/44k file up to 24/192k, it will sound EXACTLY the same as the 16/44k version, but will take up a lot more space.
You can upsample it INACCURATELY to 24/192, and get it to sound different, by making up new information based on your best guess of what should have been there to begin with.
Or you can upsample it INACCURATELY to 24/192, forgetting all about being true to the original, by simply adding new stuff that you hope will make it sound good.
However, you CANNOT reliably and accurately get back the stuff that was thrown away.... it's gone....
Virtually all modern DACs upsample the audio they receive internally (the only ones that don't are specifically referred to as "non-oversampling DACs").
(Inside DACs, it's usually referred to as "oversampling", which is simply a specific term that means "upsampling by an even multiple".)
They do this because it simplifies circuit design and enables the DAC to do a better and more accurate job of delivering the exact analog equivalent of the input.
A DAC does NOT upsample a 44k iput to 768k to make it sound like a 768k file.
The goal is to enable it to do a better and more accurate job of delivering the analog equivalent of the 44k file you gave it.
This is normally considered to be a fine distinction.... but not when DAC vendors start claiming that they're "actually turning that 44k file into a 768k file which sounds better".Well I asked Cary Audio what level of MQA unfolding does the AIOS deliver. Wow, talk about high sampling rates! The answer they gave - "The AiOS provides full decoding of MQA to achieve the full benefit of MQA encoding and decoding. The AiOS has full MQA certified decoding. Unlike other MQA products to date, AiOS will playback any MQA sample rate up to 16x. This means despite the original sample rate of the MQA file, the MQA decoder will extrapolate the file to either a 705.6 or 768 kHz sampling rate for stunning realism and studio quality sound. To add, MQA’s current limit natively is up to 384kHz and can be streamed via TIDAL. AiOS can easily handle that whereas many other products limit is 192kHz. Some can handle 384kHz. Ultimately, MQA would like to offer material natively up to 768kHz which AiOS is already setup to handle. As Dan points out, all MQA content whether natively 44.1 or up to 384kHz will automatically be rendered (upsampled) to 768kHz with AiOS and the DMS-500. MQA designed our MQA decoder to do this using MQA’s custom algorithms for 768kHz rendering because our digital ecosystem is already capable of handling such high rates!" |
|
|
|
Post by monkumonku on Sept 22, 2017 10:53:38 GMT -5
I would say the most telling thing so far is that no one seems to give a clear, concise explanation of exactly what MQA does and how it works.
If you were going to invest in a business, and no one could give you a clear, concise explanation of what that business did, would you put your money into it?
And I fully believe Hemster, Nick, etc. when they say the MQA versions of songs they have heard have been better. But is that compared to the same source without MQA or is that compared to another source like a CD or streaming or whatever. In other words, is the MQA version better because of the processing or because it is a remix or modified version of the original? Keith has been asking the same thing and no one so far as said anything about that.
MQA is supposed to make things better because of the processing it does on an original source. Therefore one ought to be able to listen to the non-MQA processed version of that source and compare it to the MQA processed version and make a judgment. But if you are comparing an MQA version to a different source then that's not a valid comparison.
|
|
|
|
Post by garbulky on Sept 22, 2017 11:00:18 GMT -5
Hi garbulky - I read your PM very closely. it said: "...if you feed an MQA file to the Dragonfly it ain't going to play MQA. It will just play regular PCM like any other DAC. The only way is to use the Tidal software or the Audirvana Plus 3 software to specifically decode MQA before it gets to the DAC." This implies (clearly) that if I have Audirvana Plus 3 AND the Audioquest Dragonfly, that I will get MQA analog output. So I downloaded the "full" trial version of Audirvana Plus 3. I played my MQA-encoded disc through the Audirvana Plus 3 program (NO iTunes - NO jRiver - just straight from the CD to Audirvana Plus 3 and output to the Dragonfly Red WITH the MQA update). What did I get? 44.1 CD-quality PCM output (green indication on the Dragonfly). I've also checked and double checked the settings per your link. Still no MQA. ALSO - NEWS FLASH - APPLE USB PORTS - NEWS FLASH I attended the Baton Rouge Macintosh Users Group (BRMUG) meeting last night and learned the following: If you have a MacBook Pro and you exceed the amperage demand limit on the USB ports (by charging more than one device at a time, for example, via the laptop's USB ports), the hardware in the machine disables ALL USB ports on the machine. They will not reset until the machine has been rebooted. After I rebooted my machine, my AudioQuest Dragonfly was immediately recognized in every USB port that I tried it in. So, just a FYI - If your Apple USB ports stop working, consider rebooting. Now one other peculiarity - While my MacBook pro was NOT recognizing the Dragonfly, the Dragonfly itself still worked! Despite not being "recognized" by the laptop, it still output analog music (but without any volume control available) and had a green status light, showing that it WAS outputting 44.1 PCM audio. Further, while the Dragonfly was plugged into the laptop, all other outputs (including the optical TOSLINK) were automatically disabled. I don't think that this particular anomaly is of any serious significance, but if you don't know what's happening, it can be confusing. So just a FYI... Finally, and this is back to MQA - Apparently the TIDAL desktop client and the Audirvana Plus 3 software are able to DECODE ("first unfold") MQA in their SOFTWARE. Otherwise, TIDAL, for example, would have to stream already decoded MQA audio. This would defeat the "packing" CODEC in MQA and lead to larger (MUCH larger) files. Therefore, if these softwares (TIDAL desktop client and Audirvana Plus 3) are DOING the decoding, then all that remains to get MQA analog out would be an MQA renderer, such as my AudioQuest Dragonfly. Now as I understand it, having a "full-MQA" DAC, such as the Mytek Brooklyn or the Meridian Ultra, would eliminate the need for the TIDAL desktop client or the Audirvana Plus 3 software. Why? Because those DACs can do BOTH the decoding AND the rendering in their hardware. If that is the case, then why would ANY player software (including iTunes or jRiver) NOT work as an MQA source? After all, the software is NOT changing the MQA-encoded PCM file - only sending it on to the "full-MQA" device that would then do both unfolds. In other words, with a "full-MQA DAC," the software needs to do NOTHING. All decoding and rendering is done in hardware, yes? Yet I'm told over and over that iTunes and jRiver are NOT MQA compatible. Why not? Boom So that the dragonfly was not recognized was probably the problem for the MQA. If you have the plus 3 then it should work. What I would do is do a hard restart of the macbook by fully turning it off and on again. Then try again. It should wok. The idea is that audirvana recognizes the dragonfly as a renderer in the software itself and sends the appropriate output. If it doesn't, there's a bug somewhere. Also for full MQA decoders, iunes and jriver should work fine. I don't see why it wouldn't. (It won't work for the Dragonfly though). |
|
|
|
Post by novisnick on Sept 22, 2017 11:41:24 GMT -5
I would say the most telling thing so far is that no one seems to give a clear, concise explanation of exactly what MQA does and how it works. If you were going to invest in a business, and no one could give you a clear, concise explanation of what that business did, would you put your money into it? And I fully believe Hemster, Nick, etc. when they say the MQA versions of songs they have heard have been better. But is that compared to the same source without MQA or is that compared to another source like a CD or streaming or whatever. In other words, is the MQA version better because of the processing or because it is a remix or modified version of the original? Keith has been asking the same thing and no one so far as said anything about that. MQA is supposed to make things better because of the processing it does on an original source. Therefore one ought to be able to listen to the non-MQA processed version of that source and compare it to the MQA processed version and make a judgment. But if you are comparing an MQA version to a different source then that's not a valid comparison. I have dont this to the best of my ability. This is how i found out that you can’t stream Tidal on two devices at the same time. I had to stop streaming on my Mac Mini in order to stream from my Oppo 103. The rest of the chain was Macintosh C220 to a VTA ST-120 to my Paradigm Studio 100’s v5 NOTE; the DAC’s of the Meridian Explorer2 and the DC@1 are so similar i would wager that it would be near impossible to distinguish the difference between the two, So Ive done as good as a comparison as i could and would say that at minimum 80% of the music sounded better via Full Master Studio ( fully unfolded / full MQA ) Some of the files were about as good as my Rega RP 6 with an Exact II cartridge, that is very High Praise from me if that has any baring. Say what you like but please do so with reason. Give it a listen, For $200.00 you can test drive for 60 days. |
|
|
|
Post by monkumonku on Sept 22, 2017 11:47:54 GMT -5
I would say the most telling thing so far is that no one seems to give a clear, concise explanation of exactly what MQA does and how it works. If you were going to invest in a business, and no one could give you a clear, concise explanation of what that business did, would you put your money into it? And I fully believe Hemster, Nick, etc. when they say the MQA versions of songs they have heard have been better. But is that compared to the same source without MQA or is that compared to another source like a CD or streaming or whatever. In other words, is the MQA version better because of the processing or because it is a remix or modified version of the original? Keith has been asking the same thing and no one so far as said anything about that. MQA is supposed to make things better because of the processing it does on an original source. Therefore one ought to be able to listen to the non-MQA processed version of that source and compare it to the MQA processed version and make a judgment. But if you are comparing an MQA version to a different source then that's not a valid comparison. I have dont this to the best of my ability. This is how i found out that you can’t stream Tidal on two devices at the same time. I had to stop streaming on my Mac Mini in order to stream from my Oppo 103. The rest of the chain was Macintosh C220 to a VTA ST-120 to my Paradigm Studio 100’s v5 NOTE; the DAC’s of the Meridian Explorer2 and the DC@1 are so similar i would wager that it would be near impossible to distinguish the difference between the two, So Ive done as good as a comparison as i could and would say that at minimum 80% of the music sounded better via Full Master Studio ( fully unfolded / full MQA ) Some of the files were about as good as my Rega RP 6 with an Exact II cartridge, that is very High Praise from me if that has any baring. Say what you like but please do so with reason. Give it a listen, For $200.00 you can test drive for 60 days. I'm not doubting you in the least when you say it sounds better, but reading over this thread plus other stuff on the net still makes MQA mumbo jumbo as no one can clearly explain it. And still not a true apples to apples when doing the comparison. That's not to say MQA isn't worth having since a device or system that makes something sound "better" to the listener is beneficial. My quibble is with the marketing and the vagaries of the explanation. |
|
DYohn
Emo VIPs 
Posts: 18,475
|
Post by DYohn on Sept 22, 2017 12:01:10 GMT -5
People seem to think MQA is a lossy compression scheme, but from what I've read it seems to me to be an expansion scheme instead which maintains the original file, adds corrective new information to that original file in a noise hidden codec and then upon play the original file is expanded to a higher sample rate that now contains all the new information. It could have been decoded into a higher sample rate file before transmission, but why do that since it is easier to transmit the lower undecoded sample rate file and the final result using MQA decoding DACs is the same. This is incorrect. MQA is a lossy compression scheme. en.wikipedia.org/wiki/Master_Quality_Authenticated |
|
|
|
Post by novisnick on Sept 22, 2017 12:09:36 GMT -5
Ah Ha!
So you really think Im going to reveal the secrets to the Secret sauce?
Bahahahahaha!!! (Said with an evil snicker)
|
|
