MQA is Vaporware

[John Dyson]

16 minutes ago, sandyk said:

Chris

 You have had the opportunity twice now to try and verify my reports, which Martin Colloms has already verified using the so called " Gold Standard" DBTs. The first time was 10 years ago when you actually downloaded the files then refused to listen to them.

 In the same thread, the highly sceptical Peter St. DID report hearing the differences along with a friend.

 (" Looking for a job outside I.T. now ")

 Do you REALLY believe that John who is highly qualified in  this area made such a monumental stuff up with his indepth reports ?

 

 You are also aware that Barry Diament who is now a good online friend of mine,. also reported hearing differences, along with his wife, using his studio gear for seamless switching between tracks with my supplied CD-R that he had ripped to HDD again, although it was intended to be listened to directly.

 

Alex

 

hearing and the brain are so very unreliable -- people can even hear things differently every time that they listen -- that is what got me, eveb though  I know about that fact of variable hearing.

I did a complete, bit by bit comparison of the 05 and 05b files -- exactly the same, and will produce exactly the same results -- it is HEARING that is variable for almost EVERYONE -- the answer to variable hearing is to run multiple tests, and do statistical analysis and also check at different times during the day.

This fact about variable hearing is one reason why myths get created and passed around over and over again.

 

PS: even headphones sitting differently on the head can make a big difference...

 

John

[John Dyson]

1 minute ago, Ralf11 said:

MQA sux

Sorry for diverting the subject -- but the discussion is Bad Bad Bad Good Bad Bad Good, etc.  There really isn't an answer or resolution about MQA.  I know that MQA doesn't have a benefit to the customer nowadays, but those profiting see a great benefit...

MQA is pretty much resolved -- unless people just like to argue :-).

 

John

[John Dyson]

14 minutes ago, Rt66indierock said:

 

John it may be solved technically in your eyes. I'm still working on the encoder so it isn't done. And the marketing issue is still on going.

 

And as the original poster cool it on discussion of depression.  Way off topic.

 

Raw fact summary:

 

Okay -- MQA no technical advantage, just download unmodified 96k/24bit or 192k/24bit or other rates.

MQA -- financial cost for users, financial advantage for IP owners.

 

If you enjoy just writing software, that is cool...  However I try to write things that actually might  help people.   MQA is of little interest to those who want freedom in using the music (without stealing -- not interested in that.)

 

Also, MQA recordings will most likely be mis-mastered like everything else -- how can the customer POSSIBLY correct the errors?  Not at all.

 

The arguments above, with a modicum of logic -- No advantage to consumers, lots of advantage to IP owners.

Okay, there is an advantage to consumers -- shorter files that have been distorted in one way or another -- but we have mp3 that compresses much better, and bandwidth is cheap anyway -- why would the consumer want something that is actually inferior to the quality that they want?.  Ethiopia probably has enough bandwidth, let alone EU, UK, Cn or the US.  Okay, the big customer base in Venezuela might have bandwidth problems -- big market there, right?

 

John

[John Dyson]

Abt MQA -- if MQA was the 'standard', more than likely, I couldn't easily get enoughtest material needed for my project.  Sure, 'scientific' measurement material is critical, but also crazy test material from the artists is also very--super important.  Someone might say -- 'who cares about your project?'...   (see below)

 

Resulting from having GOOD test material from the normal, unencrypted recordings -- I just found another breakthrough.  I cannot explain it for proprietary reasons, but for all practical purposes EVERY LAST BIT OF MODULATION DISTORTION IS GONE!!!  I have NO 'scientific' test material that puts the software through its paces like pop recordings/classical/etc.  The dynamics are very difficult to create (so many combinations.)   ABBA is crazy difficult to decode (my decoder can now do an incredibly good job on 'Dreamworld', but still isn't perfect of course -- it is damaged), but Carpenters and Carly Simon are also 'problem children' for DolbyA (amongst others.)

 

So, to answer who cares?  anyone who listens to music that is still DolbyA encoded should care (a lot of the older recordings, but certainly not all.)  Also, material that would improve from being decoded again would DEFINITELY benefit  the listeners.  I have been very busy after some revelations (cleaning up the code, preparing for the C4 work.)  The new code might also be applicable to C4 (if it is as dirty as a DolbyA is.)   Frankly, re-decoding even if already DolbyA HW decoding would make an incredible improvement (almost beyond belief.)

 

Well, why can't I just get licenses for proper, undistorted, clean material?  Because any non-standard license cost money, and people are happily paying for substandard and/or undecoded material all too often.  This means that the 'market' for my product is NIL (it really *is* very small, but existant.)  The market will also not pay megabucks if they get by with paying almost miniscule amounts for reasonable DAW software.  We have interest from organizations & recording people that most people wouldn't really believe (I cannot divulge.)

 

Projects like mine would not exist if everything was encumbered.  Of course, there would be work-arounds to decode the distorted/over complicated/some-snake-oil  mess called MQA  (would probably be technically illegal to bypass), but there is no benefit to the community at large given that I try to be ethical/follow rules (which I really try to be -- sometimes unknowingly fail.)  My project DOES NOT financially justify purchasing any specialty software or specialty recordings.

 

This MQA problem would definitely a 2nd order quality disadvantage, but shows an example about people trying to get full control, really EVENTUALLY would disadvantage a lot of people (it doesn't really affect me -- I'd be working on another project, just not audio.)

 

Of course, someone else could write the program that II have written (most likely, they would not have found some of the super-duper advanced  techniques because of the limited time & resources), rather than the software likely costing under $500 like mine, it would have to cost $10k or more, & almost definitely be inferior-- anyway, it would be dead in the water right there because of the cost/price.

 

MQA doesn't have evil intent (it is okay to siphon money from people), it is just the wrong thing (excessive control) at the wrong time (who cares about big files, unless excessively big, anymore?)  Like I wrote before, maybe people from impoverished nations might care about bandwidth costs, but they certainly couldn't pay for a greater tax anyway (MQA's overhead.)

Timing is wrong for MQA, and it just dirtifies the signal with insufficient benefit to the consumer.  (Some benefit to the IP owners -- even negligent benefit for the artists.)

 

John

 

[John Dyson]

1 hour ago, Jud said:

 

There were MQA tracks I liked better as well, but they were obviously different masters.  This made me think of how the better masterings could sound if they hadn't gone through the process that made the MQA White Album sound worse to you.

 

In my opinion, even though MQA might not be a good thing -- the mastering problem is often worse when it comes to technical quality. (That is, ignoring the financal/political/licensing aspects.)   I sure hope that people don't really blame the mastering engineers either -- blame the beancounters that keep the engineers from doing their due diligence and spending the time needed to do a good job.

As far as I know, there is frustration about this matter on both sides of the mastering/consumer side of things.

 

John

[John Dyson]

1 hour ago, mav52 said:

 Ok, Paul I really don't know if you use 2L or not, but their collected artist sold on their site used to be in all formats, but now its MQA.   ITS MQA on Tidal and its MQA on this site. https://www.highresaudio.com/en/label/view/e9303123-e18a-4694-bda5-2edd10004141/2l

 

You can still get their artist on HDTRACKS and Native DSD but some of that music was old.  

For me -- MQA highresaudio is zero res audio.  Might as well not be offered as a product as far as I am concerned.  They should just sell the undamaged product on appropriate medium -- much better alll around. 

 

John

[John Dyson]

I haven't read the technical details about MQA -- but just thought of something 'interesting' about their alledged usage of the term 'blur'?  I haven't been thinking about video -- but there is a kind of processing done when video has been rolled off (blurred) and some sharpness can be hopefully recovered (sharpness, NOT detail) but  a nonlinear processing technique.   'Sharpness in video is NOT detail, likewise a sharper or nonlinear expanded audio signal is also NOT detail.)  Video 'deblurring' isn't all that complex or complicated (either one), but must be done correctly.  The kind of technique used to 'deblur' video wouldn't be exactly applicable to audio (and probably not desirable except in very controlled cases for either forensic or recovery purposes.)

 

Think of it like this -- video is the world of edges and nonlinear detail.   Audio is the world of 'tones' (sine waves), even our ears detect tones and the various durations/combinations, not really edges and nonlinear details (maybe certain percussive transients -- but that is not the general case.)  Applying transforms for compression is also different between audio and video.  Being very very general: ypically for music material, the transforms/compression are done to the spectral type information vs time, while for video transforms are essentially done to the video itself vs. time.  (The speech type compression is sometimes different tha nwhat is used for music.)

 

Point being -- it worriies me that 'deblurring' means applying some kind of gratuitious nonlinear processing (or some analog of that) that happens to sound good to some people.  Deblurring sounds like 'adding artifacts' to me.

 

I could be very wrong, but other than peaking (ugly), some kind of nonlinear scheme (however fancy it might be -- e.g. AGC compressors have nonlinear characteritics), it might be some kind of nonlinear signal modification.  There is already too much signal modification used to hide defects, now there is signal modification to siphon more money from the customer base -- without giving anything useful in return?

 

The deblurring thing (or whatever it is called) really bothers me.  I hope that any follow up gobblty goop doesn't include discussions about Gibbs 'ringing' :-).

 

John

[John Dyson]

4 minutes ago, Ralf11 said:

This thread enforces my decision to stay away from streaming...

I was thinking EXACTLY the same thing -- your position on streaming is the same as mine.

 

John

[John Dyson]

8 minutes ago, Kal Rubinson said:

 

Why?  Just ignore the MQA stuff.  There is little of it, still, on Qobuz and there's lots of great music in lossless formats.

 

I isn't just MQA -- it is also the transient nature -- the lack of posession.  Probably a generational culture difference.  Too much control, alteration and dependency for my own emotional comfort.   (As my grandma used to say -- a little off color -- 'whatever blows your skirt up' :-)).

 

 

[John Dyson]

3 hours ago, Paul R said:

 

Unless you are always "on" - that is, always engaged in critical listening, then streaming has it's place. Critical listening is work, even if it is work one loves, it is still work. Casual listening on the other hand, is all about enjoyment. 

 

Streaming is for casual listening. Trying out stuff you want listen to, but don't know if you want to buy. After you listen to it, you may decide to buy it, or you decide not to buy it. You can even choose to just listen to it again. The choice is yours. And stuff that makes you run away with your ears starting to bleed?  That you know to stay away from pretty quickly, and at minimal cost.  

 

 

If listening to streaming material is like listening to the old time radio (moderate quality, casual), then why pay for it?  I'd just listen to free stuff.  Why pay for butchered material (like MQA?)  Why pay for something that just might go away?  For casual,  I'd listen to something free.  For something that I invest ANYTHING in, I want a copy that no-one can easily take from me.

 

[John Dyson]

8 hours ago, Ralf11 said:

 I could discover music and performances via streaming, but only a reduced subset of what I want to try out - mostly recent items.

 

Instead, I use the xlnt collection of CDs at our city library, with a university library for backup (which has a very decent music school - tho not competitive with the music school/library at Indiana univ. like this place thinks they are...).

 

But that may not apply to others.

 

There is also bandcamp.

For stuff that I really like (or use for reference), I tend to need specific release versions.   Not every version of Carpenters is mastered appropriately, not every version of ABBA has been processed/mastered correctly (for a really sickening time -- listen to the premium  'The Complete Studio Recordings -- Crest of 4.6, peak-rms of 13-14dB -- BAD!!!)  There is a lot of substandard material out there, and for stuff that I love (or use as a reference), I want the proper version.  On the other hand, if I don't care, then listen to free radio.   The in-between version of paying for online-listening (streaming) seems to be a lot like a jukebox, and I never liked jukeboxes.  The jukebox might not have my chosen selection forever, so why keep on supporting it?)

[John Dyson]

1 hour ago, bluesman said:

A lot of what's substandard out there is the music, not the recording and its presentation.  "You can't shine [expletive deleted by author]."

 

I don't think I'm narcoleptic.  I can listen through almost any music to hear the musicians' parts and appreciate the technical quality of the charts, the playing, and the recording.  I can stay awake listening to Kenny G, even if I do have to keep reminding myself that he's a skilled musician and an educated man who simply chooses to play sonic Sominex.  But I have a bit of trouble with the Carpenters and ABBA as reference source material

Every night I say a little prayer: "May ABBA never emerge from my speakers!"

 

PS:  Karen Carpenter was actually a fair drummer, but she couldn't play well enough to do studio work and apparently had some difficulty singing her best while also playing.

 

My primary interest in ABBA is not the music, it is the really difficult mix of mixed female vocals, wall of sound, and hard compression.  it is a total b*tch to decode.  Think about this -- lots of potentials for IMD because of multiple vocals near same frequency, then the wall of sound adding in a randomizing factor,, and then hard limiting (or intense compression) giving a nice relatively square  shape to the envelope.

 

(I am in the Indy area -- I know ALL ABOUT Wes M....)

 

I am actually quite proud that my decoder can produce ABBA results that sound surprisingly good -- definitely better than you have ever heard...  Would it improve the lyrics?  That would be a silly  question; Here is a silly answer:  Yes :-).

 

Karen's material also has problems -- they used DolbyA to enhance her voice -- bring up the highs from her relatively mellow vocals.  DolbyA sounds BAD if not decoded, then only sounds kind-of bad if it IS decoded (most of the pop material between the late 1960's through early 1990s')   Using my decoder actually cleans up the sound VERY VERY significantly -- got the idea?  The purpose is NOT ABBA, even though it helps their recordings also.

 

On a scale of 1 to 10 in my enjoyment of music -- I'd give ABBA a 3 (because it is either happy or bright, not because of trivial lyrics.)  I like Karen carpenters stuff better, and material like Nat King Cole even better.  (I did like Simon & Garfunkel until my deocder showed the recordings not as good as they initially sounded -- kind of ruined the experience.)  Linda Ronstadt vocals are often distorted by the Aphex distorter (I think that is the name, or should be.)

Almost every pop group has a problem (early Beatles -- often bad stereo or mono, pick one.)

There is nothing that is perfect -- I can criticize almost everything, sometimes it is best just to enjoy.

 

John

[John Dyson]

46 minutes ago, garrardguy60 said:

 

I am with you on ABBA and the Carpenters [though the latter isn't at the same level of abject suckitude, since Karen Carpenter had talent], but I have to step in and defend Wes Montgomery's California Dreamin'  [which was on Verve, not A&M]. I'll agree with you that, as easy listening/smooth jazz, we're not talking about breaking any ground here. But I'd argue that the [unique, wonderful] tone he got out of his guitar makes anything by Wes Montgomery worthwhile. Doubly so since he died so young. Interestingly, that album, which charted in 1966, has just been reissued on vinyl by Impulse Records, though it seems to be available only in Europe.  OK, sorry for off-topic.

The problem for my project about material with natural instruments and unprocessed vocals is that they make very poor, non rigorous test material.  Sounding 'nice by default' doesn't help my project.  (BTW, my best friend just recently had been working at the Walker Theater on Indiana Avenue (short street where Wes M played).)  Nice sounding music to begin with is SO VERY EASY to deal with -- way back when R Dolby created his DolbyA, I am sure that he didn't forsee the stuff that was going to be fed to his NR system.

 

This kind of unforseen limitation is the reason why schemes like MQA, mp3, opus and other lossy systems will tend to limit the choices for the future.  I am working hard to help mitigate the previous damage by DolbyA (as much as possible.)

 

Using nice, mellow music just does NOT test the dynamics processing, doesn't usually stress the attack/decay enough to intermodulate sufficiently to even produce much audible damage.  (The worst kind of bad thing that DolbyA did against typical, natural music was to fuzz up complex combinations of sources and also blunted the transients.)  DolbyA  can lose intensity in cymbals, and vocal choruses lose clarity/individual voices.  Typically, vocals turn into a 'blob'.

The way that my processing works is to do some very complex mathmatical tricks to keep the 'blob' from being so very bad.  There are the obvious 'tricks' like limiting the gain control slew rate -- that is trivial.  There are other techniques that treat the audio signal as being modulated -- and does some processing that removes undesired modulation sidebands (often called IMD, but it is a specific kind of IMD associated with gain control devices.)

 

I didin't intend to bring up the 'clarity and cleanness' of ABBA recordings (or their sophisticated lyrics), or the perfectly clean recordings made of Karen Carpenter (very sad, actually) -- but that my decoder can actually deal clealy with the damage -- and actually sound relatively clean, being able to more clearly hear individual voices, etc. (big smiley for the humor impaired.)

 

The processing DOES work on more normal chorus and instrumentals also -- some of my professionally sourced test recordings are complete, professional DolbyA  in a full 192k/24bit unmastered form.  (I do have some relatively well known professional recordings by moderately well known arties, some soon to be released -- I use them for testing, but it does NO-ONE any good to talk about them in detail.)  The clean, jazz type material is damaged by normal DolbyA decoding, but not enough to be used for anything except final quality control and estimate quality improvement.

 

Being able to cleanly decode DolbyA encoded ABBA is an achievement, because a DolbyA does it so very poorly.

Who really cares if ABBA might suck in some people's opinion -- I know people who think the opposite (E. european females of the 40-60yr age group) - music taste is based upon individual perceptions & experience.

Frankly, there is material that I know a lot of 'elite' listeners might like, but I totally dislike -- however, you'll NEVER hear me condemn or criticize their taste.   BTW -- ABBA for music quality doesn't rate super high on my scale either -- but that isn't the main reason for using it for test material.  I only have about 70-80 96k/24bit DolbyA ABBA recordings -- imagine needing to listen for IMD for every internal release/test pass that I do?  I am very tired of ABBA, Carpenters and almost every other copy of DolbyA material that I do have.

One thing that I can say for sure -- when the decoder actually makes something from ABBA sound clean/clear, it makes me feel really good, like I have achieved something -- will help other people also, who listen to much more popular groups than even ABBA was at one time.

 

Processing like SR is much worse than A -- and Dolby B/C being relatively unsophisticated, also has lots of quality issues.  Lots of damage went on back in the past.

 

John

[John Dyson]

I have put up a demo of the DA decoder to show the difference -- if you notice that the Polar version has swishy percussion and generally softer sound.  I don't know if vocals are cleaner on this example (sometimes there is already distortion in the recordings.)  There is a bit of compression in the polar version, so it isn't 100% fair.  But notice the positive, clear sound of the dhnrds, but the polar version is softer than even the compression should cause.

BTW -- I found that my copy of the Waterloo album has lots of noise glitches in  it (haven't heard them before, but the DHNRDS cleans up the material so much that they become obvious.)  Then, I listened to the Polar digital copy, and could JUST BARELY hear the glitches.  A normal DolbyA device fuzzes over the detail.

 

The dhnrds HAS been compared against other original master tapes also -- but not in this specific case.  All I have for ABBA are several properly DolbyA decoded examples of each song.  The DHNRDS USUALLY compares fairly well with an original -- and USUALLY blows away DolbyA HW.  DHNRDS isn't always perfect -- FOR SURE!!!

 

Sorry for using mp3, but it does show most of the differences.  The DHNRDS really does need .flac or lossless to show the quality -- mp3 does miss a LOT of detail.  The player on dropbox isn't very good -- so I don't know if will gloss over the differences.  I normally use sox or convert it with one of the mp3 tools.

 

The quality IS NOT GREAT -- but I think that you can see how difficult that ABBA is to process.  I know that the DHNRDS IS pretty accurate -- the quality of pop recordings is generally pretty low.

 

https://www.dropbox.com/s/prkz8ghg4jwqw2w/nameofthegame-dhnrds.mp3?dl=0

https://www.dropbox.com/s/ey502pue7kr1nk4/nameofthegame-polar.mp3?dl=0

 

[John Dyson]

Wont post any more links -- just wanted to show that the DHNRDS can sound pretty good also (th ABBA stuff is really a mess, REALLY.)

 

Here is a short Linda R snippet (I haven't massaged it -- this is raw, and not tweaked):

 

https://www.dropbox.com/s/v34yepx63x1n7zy/Pitiful-dhnrds.mp3?dl=0

 

You'' maybe notice that there is a spot where the vocals are a little odd sounding -- that is because I didn't set the calibration quite right.  Also, the DHNRDS has lower quality modes that sound VERY SIMILAR to a real DolbyA.

 

No more space wasting and off topic.

 

Bottom line -- it is SOOOO important to keep from doing things that will be regretted in the future (e.g. DolbyA, MQA, etc.)

 

John

[John Dyson]

5 hours ago, Don Hills said:

 

The State Of The YouTube Music Economy

 

<off sides for a minute -- and back on subject after this paragraph>

It is a moderately special case, but whole thing about licensed music has been tricky for me when trying to demo the results of my project.  On a micro (once in a long while -- here is hout it sounds, one-on-one) scale, I tend to be a little lax, but cannot broadcast entire songs without either feeling guilty or possibly even getting in trouble.  The results of my work WILL benefit the industry, but in the entire scheme of things, 99% of the time the material would have just been copied (I mean peer to peer -- maybe through a You Tube, then people not paying for the benefit of the music) and not reviewed.  Most of the real-world copying is NOT for review, but instead for simply adding to a collection (and not paying.)

 

Any deep resentments against the license holders do NOT count -- they own the stuff.  There are limitations as to how they can control it, but there are also limitations as to how the material should be distributed.

Excessive cheating WILL encourage the creating of Frankenstein monsters like MQA.  Perfectly clean material will cause restrictiveness like the control of HDTV signals (used to have unencoded HDTV around until stuff like HDMI.)  NTSC/PAL equivalents (the junk coming out of a home VHS//SVHS/Beta/EDBeta VCR is a very poor excuse for NTSC/PAL) have been available, but as the quality increases -- the restrictivness has also increased.  The risk of duplication is always in the back of the owners minds.  With VHS/Beta, there are only so many generations that one can go before the video is pretty much one big blob on the screen.  With properly done digital video, you can copy infinitely.  (Even my D9 decks where  I resolved  down to YUV analog and back, would do 10generations and barely decrease to Betacam SP level quality.)  That was just about the time that NTSC/PAL really started dying off anyway -- now who really cares much about NTSC/PAL master quality?

HDTV, on the other hand, starts coming close to the family jewels (Original Tron was done something like 600-700 lines AFAIR, many more recent movie masters with special effects were at 2k-4k -- now are probably much higher.  The stuff being displayed on consumer TVs has approaching master quality now.)

Right now, we often have access to pretty good audio -- in fact, by the stupidity and short sightedeness of some of the distributors, SOMETIMES they have let the family jewels out by distributing almost unmolested DolbyA copies of older material.  (Not really here nor there in this specific discussion.)

If there would be a culture where people would think twice about creating massive copies of their collections, material like MQA would have less strong justification vis-a-vis DRM.

DRM stuff more helps to minimize fear than anything else.  YouTube with their detection systems (IMO) seem to help more than hurt -- because people often do use YouTube to share.  The $$$ can still go to into the correct pockets.

(Guess just rambling -- and right now have troubles with demos BECAUSE of TRYING to follow the rules a little.)

 

John

 

 

[John Dyson]

17 minutes ago, bluesman said:

That thing was called the Aphex Aural Exciter - there was one in every studio rack in America back in the day. I’m lucky enough to live in Philly, the home of the original Sigma Sound Studios (where I got to record as a sideman on several instruments). I don’t remember if they had an AAE, but it wouldn’t surprise me to learn that even Sigma used it when producers asked for it. I suspect that Rudy VG never did, but I don’t know this for certain.

Yeah -- you are right about the so called 'Exciter' :-).  I wrote a program that can undo some of the phase distortion done to Linda Ronstadt and others, but was only partially successful (it takes multiple passes & teaking to properly undo each recording.  It makes it so she has less of an electronic lisp, but is too much work to use.  I still think that 'distorter' is a more accurate term 🙂 (really -- I mean it with humor, not arrogance :-)).

[John Dyson]

1 hour ago, Shadders said:

Hi,

I have asked this before - where exactly are the examples of the "blurring" or audio files in error, that have been corrected ?

 

Why doesn't the MQA website show a before and after example ?

 

Why hasn't MQA Ltd listed 10 albums from the past 30 years where there is the claimed "blurring" due to the filters so we can examine the claims ourselves ?

 

MQA is a fraud. There is NO evidence presented, and the AES MQA paper has significant flaws/errors, that MQA cannot therefore be a bona fide system. All we have are claims it is better using pseudo scientific statements, which have been misused.

 

If you examine the QSound website - they have demos, so why doesn't MQA, if it is meant to be the "New World Order" ???

 

All MQA is an effects processing algorithm - based on false claims that there is blurring.

 

Regards,

Shadders.

I know what MQA advocates will do to demonstrate fixing the 'blurring'.  1) simply fill in the high frequency loss caused by a linear phase low pass filter...  2) sharpen the edges in the audio by some kind of nonlinear process.  The only real 'blurring' caused by an LPF is when the LPF is not linear phase, then various frequencies have differing delays -- the phase shifts can then be partially fixed by an equalizer.  Each of these technologies is well known (nothing new here), and I sure hope that they don't try the nonlinear method.  (I am not even going to get started about the non-existent Gibbs 'ringing'.  To basically summarize -- Gibbs isn't ringing, but is a residual from removing the higher frequencies.  There is no 'regeneration' coming from sine exitation like true ringing has.)

 

Another thing that I have learned when playing with audio -- starting with DSP knowledge alone.  Doing certain things are more challenging than they seem to be.  Perhaps one thing that had confounded me at first -- the idea of splitting frequency bands -- processing each band differently, and then reconstituting the audio by gluing the bands back together.  That processing of mixing the bands back together is frought with land mines.  It isn't because the bands cannot be easily split and rejoined, but the different processing in each band can make the results more damaged than what would be initially guessed.

 

I'll give away one of my DolbyA processing tricks as  an example (I am diviulging only because it is so obvious as to be worthless from an IP standpoint) -- I sometimes split some of the 4 bands used in the DolbyA system.  So, there aren't just 4 bands being processed, but instead (on the current version) 6 bands.  Each band IS processed differently, but the results are superior than if the 4 bands are processed normally.  When doing the different processing -- there are lots of potential mistakes -- perhaps the most obvious might be relative phase shifts.  So, when joining the bands together (if the filter skirts are wide), there can be cancellation/re-enforcement that will create peaks/valleys in the freq response.  Time delays and even different levels than expected can also cause differing kinds of problems.

When someone willy-nilly talks about (the fixing the 'blur') either adding nonlinearity or processing different frequency bands differently (when expected to have the same processing), then I become worried about unforseen problems.   Think about the old NR systems that sounded okay when utilizing old technolgies, and now -- they sound terrible because our standards are higher.  If the ARE doing something obvious -- then their claims start looking more and more specious and silly (making a big deal out of nothing.)  They seem to be trying to create an advantage/improvement where there is none.

 

Chopping/splitting/processing/slicing/dicing audio is something that needs to be done carefully and/or use established techniques.  I am not including the 'stairstep' myth in this category, but there ARE distortions that are REAL -- and it is very easy to mistakenly cause problems.   A good, common example that slices/dices, saves some space destructively  and doesn't do the job perfectly might be mp3...    A good, common example that slices/dices, saves some space non-destructively and does do the job perfectly might be flac.  MQA is more like mp3 than it is like flac.

 

Anything that is destructive -- guess what?  It is destructive.  So, if it is claimed that something is non-destructive, and it isn't obvious that it is non-destructive -- significant proof is needed.

 

MQA is a gimmick whose time is long gone -- other than for benefitting those trying to control/limit the freedom of listening to high quality music or make money from intellectual property.  There just is no other reason for MQA to exist.  the existance of MQA CERTAINLY does not benefit the listener.

 

John

 

[John Dyson]

When I hear about 'audio guru this' or 'audio guru that', I don't quite understand.  All too often, I hear nonsense from people with that label.  I am not claiming that they are always wrong, but there is a lower signal to noise than should be.  Everytime I hear about 'audio guru said', I start thinking about 'Gibbs ringing' or just thought about 'pre-ringing' as another interesting misunderstood thing.  (Linear phase DOES sometimes sound different from minimium phase, but often the diagnosis is a little off.)  On the pre-ringing thing, it is more accurately that some frequencies are delayed differently in the minmum phase, and sometimes might sound better.  (When dealing with filters other than linear phase in complex signal processing -- it can be much more challenging...  When signals are split with linear phase,, and then re-assembled, there is no real difference other than delay -- so there is really NO ringing.  It is more of a difference sound and sometimes Gibbs effect -- NOT Gibbs ringing.  Most of the time, the 'ringing' is actually removal of expected signal components, thereby causing a residual near-sine-like wobble in the signal.)

 

Point being -- beware audio 'gurus', you really want to hear from EEs with long term expertise in DSP (actually writing code and/or designing -- NOT COPYING -- circuits.)  They might agree with an effect looking the same -- but the manifestation might be from a place different than common knowledge.

 

John

[John Dyson]

15 hours ago, crenca said:

 

 

I have to agree with Sonicularity here Rt66indierock, in that what "thoughtful" questions are left?  Certainly not technical ones - MQA is an empty proprietary rent scheme for the consumer, and a stab at DRM for the industry.  Perhaps you mean financial and such questions that might tell us a bit about how MQA plans to go forward?

Effectively it seems like a quality tradeoff (probably mitigated by some games) to get more bits for DRM type stuff.  There is NO benfit to the end consumer.  Anything talking about games that improve the quality to compensate are secondary to the actual DRM purpose.  There certainly is little/no benefit to a marginally effective lossy compression scheme.  (I mean margainally effective in the sense of saving space.)   Any loss of quality is not a good thing anymore -- when it can be allowed, then we have mp3/opus/aac/etc.

 

John

[John Dyson]

On 4/16/2019 at 3:16 PM, Ishmael Slapowitz said:

Wow.7digital is actually a really good FLAC download store. It would be a shame.

 

It also is further proof that even LOSSLESS Redbook music in a non physical format is dying a death. To the mass of consumers Mp3 IS lossless.

 

 

geesh, I have been putting together a beautiful demo for the ABBACHAT group, and tried to encode it into mp3 format (lame, in insane 320k mode, with all of the checks enabled, and carefully limited bandwidth to avoid any kind of possible artifacts.)  Mp3 sounds kind of bad (actually fairly bad, definitely not great.)  The material that I can produce is SIGNIFICANTLY better than mp3 (with 40yr old ABBA), let alone the more recent studios that could totally blow it away.  (I did careful processing & mastering -- much better than any demo that I might have done here.)

Mp3 might be okay for cars or listening while running, but it is really not all that great IMO.  And I do NOT have 'golden ears' at all.

 

John

[John Dyson]

5 hours ago, Paul R said:

 

To me, MP3 anf 320bps sounds very close to CD quality. Sometimes there is just something not quite right about it, but it is usually at least listenable.  What SRC did you use to convert it with? I have grown quite fond of iZotope RX7. (Which my spelling checker insists on changing to isotope.)

I used my own cobbled together research engineering type stuff.  I very very seldomly use stuff like Audacity, even though it is sometimes convienient for certain things.   My DolbyA decoder is as advanced as ever produced -- does do amazing things to correct the distortions, even some from encoding.  Software that was a 2wk project (for the audio processing portions -- usually it is GUI that takes the time) just cannot compare with carefully crafted material.  The DolbyA compatible decoder isn't just a single band fake (or a sloppy multiband) like some other decoders available for purchase.  This decoder is SUPER sophisticated and does mostly match as closely as various versions of DolbyA HW match (which is specifically not too difficult a goal.)  Except for improved distortion characteristics, the DHNRDS DA decoder mostly sounds like a DolbyA.

 

(My bragging results from a multi-year project, but my ego is just barely big enough to avoid Napolean complex type things.  After a fairly competent engineer sucessfully develops a difficult project like this, he/she deserves a bit of self congratulation. 🙂)  Even if the decoder is bad (which all feedback that I have received says that it is very good), I would self congratulate because I tried very hard.

 

The mp3 encoder that I compared the results with 'lame', which I have heard is one of the better encoders.  mp3 using that encoder left a lot to be desired even at 320k.  * I do have to admit that sometimes the DHNRDS decoder has produced especially difficult material to mp3 encode because of suboptimal quality -- and erroneously producing difficult-to-encode audio.  I don't think that this instance is such a case -- but at least I am trying to fully inform about possible mistakes.

 

I will try to post some comparisons today or tomorrow.  I found another quality optimization in the DolbyA decoder (also busy doing the security mechanisms -- my project partner tells me that the decoder (DHNRDS project) will be stolen otherwise.)  For only 10dB,10dB,10dB and 15dB processing for each band, there are LOTS of mistakes that the hardware, software and basic design can make...  This has been both a fun and frustrating project.

 

I will do EVERYTHING that I can to put together a clean demo today.  When I upload the results to Dropbox, remember that the default mp3 decoder used when on Dropbox is insanely bad.  It would be unfair against mp3 to depend on the Dropbox decoder.

 

John

[John Dyson]

On 4/19/2019 at 9:43 PM, Archimago said:

 

@John Dyson, could you perhaps post even just a few minutes of what you've been putting together in FLAC and the LAME encoded 320kbps MP3. Would be great to have a listen for comparison.

 

Thanks...

I am so sorry that this took so long to do.  I get distracted and forget things.  Lately, been writing some licensing software, and that is so far afield from what I know that I am becoming frustrated all of the time.

 

I put together some chopped (sorry, I might have forgotten to 'fade' them instead of chop them) demos of freshly and unEQed decoding results by the DolbyA compatible decoder.

 

First -- I did NO EQ or ANY KIND of compression of the decoding output.  This is RAW output, and I have just rebuilt the decoder correctly (I think...)  Been working on some very subtle performance improvements -- which often totally break the decoder for a while, but I think that it is put back together correctly.

 

There are also equivalent .mp3 results at 320k.  Listening carefully, you can tell that the output directly from a DolbyA encoded (then decoded) copy can be pretty darned high quality -- when compared to the original source material.

 

IF you do compare with equivalent commercial releases -- make sure to start with properly mastered releases -- a lot of releases are DolbyA encoded or badly EQed.  I usually try to use original vinyl for comparison.  The vinyl is not higher quality, but is a good basis of comparison for EQ.

 

ABBA is doubly problematical -- their music is consistently difficult to properly decode, but there are often flaws in the recordings also.  Another thing about ABBA -- I have 3 undecoded DolbyA copies of some ABBA songs, and each one sounds different.  I think that the SOS example that I uploaded is not mutiply encoded/decoded.  DolbyA starts getting ugly after multiple encode/decode cycles, so I have tried to find the best sources in my collection.

 

One other thing -- it is INCREDIBLY tricky to distinguish between poor mastering and leaked DolbyA material.  I can often be tricked into thinking that something is leaked DolbyA, when it really isn't.  I use multiple criteria to demonstrate the high liklihoood -- including looking at the noise spectrum -- because DolbyA creates certain patterns and shapes in the spectograms.  Also, I am pretty good at detecting the HF compression when using DolbyA, but pop material is already compressed -- so it can be confusing to make the correct judgement call.  So, when I don't actually KNOW that I am providing properly decoded material, I have super high confidence that this started with leaked DolbyA material. (Legally leaked/properly purchased is what I mean.)

 

All said, there is almost the best possible copies of some of the material, and the mp3 version -- encoded with lame in 'insane' mode.  I will also be trying to compare these today (again), and others might be interested in trying to detect differences.  THESE ARE SHORTENED to try to be a good citizen, but to avoid any kind of long-term distribution of even the snippets, I'll be deleting these examples within about a wk or so.  These are shortened enough that the material is a tease and might motivate purchase of the music :-).

 

I am notoriously bad at swapping the L+R channels also -- note that in any comparisons with normal releases -- sometimes the DolbyA encoded versions have swapped channels, and I don't know why that sometimes seems to be true.

 

Here are the demos (this is all 1960's/1970's stuff, I think pretty good for the timeframe):

 

https://www.dropbox.com/sh/ab9nhtqjforacd8/AABvt7IYgoob7VXxpN0ekK6ra?dl=0

 

 

John

[John Dyson]

1 hour ago, Rt66indierock said:

 

Actually not. Part of the purpose of this thread was to cover topics that impact recorded music more than the problems MQA claims to fix. I will give John a bit a latitude in this area.

Thanks -- there were probably three reasons for chatting about these decoding examples -- lossy compression  in ideal conditions, the mastering differences being MUCH more important than the last small amount of sample rate or bit depth, and trying to show the potential improvement available (using my decoder or perhaps some superior one that might yet be invented by someone who really IS an audio expert :-)).

After all of this -- still off topic, but an example of the h*ll that I am in -- the demos that I made were done by a version of the decoder with a bug remaining (just found the bug, and the decoder *IS* complicated - probably 10X more complicated than a typical compressor software package -- running in 20threads.)  I still think that the general quality is still well represented even with the bug.

 

We have a LONG WAY to go before some of the finer grained improvements are really needed -- this first and most important (and where we have least amount of control) is in the source material itself -- that is, the mastering part of the process.  The bit-rate (or with MQA, the bit-rot) is important, but not the biggest problem.

 

I'll probably end up updating the repository (it will be the same location) early tomorrow with the corrected material.  As it is, again -- the examples -- for looking at possible quality improvements, and comparing with what mp3 can do -- that is still valid.  (I could have probably kept the defect hidden -- but that is NOT in my nature -- I truly believe in transparency and attempting to be honest.)

 

I *think* that starting with the highest quality material possible -- mp3 (or opus) can handle any case where bit reduction is really needed nowadays.  Otherwise, we are VERY WELL COVERED with .flac and other lossless compression schemes.  The in-between data-reduction doesn't have a place anymore -- time has come and gone.

There is *ZERO* advantage of a lossy compression scheme that controls our freedom also.  All bases (for our needs) are already covered nowadays.

 

John

[John Dyson]

On 4/25/2019 at 9:49 AM, John Dyson said:

 

Here are the demos (this is all 1960's/1970's stuff, I think pretty good for the timeframe):

 

https://www.dropbox.com/sh/ab9nhtqjforacd8/AABvt7IYgoob7VXxpN0ekK6ra?dl=0

 

Demo snippets re-uploaded at around 5-6AM EST/USA time.

 

Just fixed a minor, but noticeable bug in the demos.  At this point (after my review this morning), I believe that IN GENERAL, the quality is as good as is available to the consumer -- except I believe that the Linda Ronstadt snippets might be mildly compressed.

The encode/decode distortion still persists in my copies of the Carpenter's albums -- not sure if the examples that I uploaded have an obvious case of the distortion.  The Carpenters did a bit too much processing, therefore obscuring Karen C's special voice.

 

Well, the specifics here are NOT on topic, but rather this ancient material would be good for encoding/decoding comparison.  I HAVE run into cases where mp3 is not transparent on the Brasil'66 material.  Multi-generation DolbyA encode/decode material like the Carpenters don't really tax mp3 all that much.

 

Mastering IS the big variable in consumer audio quality also -- we consumers need to somehow demand good quality remastering, and not just 'compress and compress more' for remastered releases.  The ABBA 'Complete Studio Recordings' show AM radio processor style compression (within a few minutes -- I'll upload the equivalent TCSR version of the ABBA snippets as an example.)  The Carpenters release available through HDtracks is also much further processed than the original albums.  I believe that the HDtracks copy of the Carpenters (not really HDtracks fault, but rather the distributor) is also leaked DolbyA, I believe that I did an analysis, but should probably recheck.

 

*added comment -- after listening again, I don't think that the TCSR recordings were DolbyA decoded before the extreme compression during mastering.  The TCSR is a real mess.

We gotta demand better mastering, then the 44.1k/16bits or 96k/24bits or whatever become a LITTLE more operative -- but still not really all that important.   14bits with stolen bits need not apply, however.  IMO -- my opinion only, quality issues are totally masked by mastering (or lack-off.)

 

John