Article: MQA: A Review of controversies, concerns, and cautions

[Doug Schneider]

30 minutes ago, adamdea said:

Interesting Doug 

The first 14 bit players were called 4 x oversampling I’m fairly sure. Does your first category mean changing the sample rate by an integer multiple?

The only really distinct meaning of oversampling  I am aware of is simply sampling by a higher rate than is needed to capture the information in the signal. 

 

I actually found two bits of information about this from Thierry Heeb (he was with Anagram): "Oversampling is an upsampling process where the ratio between output sampling frequency and input sampling frequency is an integer larger than 1. Upsampling is any kind of transformation providing an output sampling frequency that is higher than the input sampling frequency and not necessarily a ratio."

 

He also wrote an AES paper in 2006 on the subject, but I can't find it at the moment.

 

I also found this Spanish-language article that references Heeb and attempts to distinguish the two: http://www.revistacec.com/didactica/3014-upsampling-y-oversampling-3014.html

Doug
SoundStage!

 

[asdf1000]

5 hours ago, The Computer Audiophile said:

One other note about not engaging or not refuting data from an "anonymous" person, this is one of the first things Bob mentioned to me when he called.

 

Hi Chris, do you mean Bob didn't have an issue with this article from the "anonymous" Archi?

 

Or he did have an issue with it when he called you?

 

[The Computer Audiophile]

2 minutes ago, Em2016 said:

 

Hi Chris, do you mean Bob didn't have an issue with this article from the "anonymous" Archi?

 

Or he did have an issue with it when he called you?

 

Yes, he has issues with it. 

[asdf1000]

1 minute ago, The Computer Audiophile said:

Yes, he has issues with it. 

 

That's rather unfortunate.

[Doug Schneider]

14 minutes ago, John_Atkinson said:

I believe the correct pseudonym is Archimago.

 

And I still managed to find him easily enough with just a click or two of the mouse. Can't say we would've found information on Tellig that easily...

 

Doug 
SoundStage!

[Kal Rubinson]

1 hour ago, sullis02 said:

Ditto.  Hell, I didn't know until *this thread*.  

Tellig ~ Gillet(e). Seemed obvious to me from the get-go.

[mrvco]

1 hour ago, miguelito said:

Actually, I DID get "A/B" played to me at Meridian in NYC around March of 2014 or so (maybe 2015).

 

The MQA versions were substantially better. As I've mentioned elsewhere, the most amazing case was a 24/192 recording and it's MQA version. However, after listening to a lot of MQA in the last couple of months, I realize it MUST have been an MQA version from a different mastering altogether. 

 

On top of it all, I got such gibberish answers to my questions that I was left with the impression of a poorly set up magic trick. Sad.

 

I've had a Mytek Brooklyn DAC for nearly a year now and am still waiting for someone to identify that 'definitive' MQA album or track that I can compare to its equivalent FLAC counterpart and hear that (alleged) MQA magic   That being said, I much prefer the Fast Roll-off filter to the Minimum Phase Filter regardless and tend to leave the DAC set to FR.  Early on I A/B'd plenty of MQA and non-MQA tracks through Tidal with the mandatory MPH filter, but apparently I don't have the '4/7' golden ear .  Now when I do go through the hassle of switching the filter to MPH for an MQA track or album from Tidal (via Roon), it has never seemed like anything but time wasted.  I do wish there was a way to only see MQA or non-MQA versions of tracks and albums in Roon.

[Doug Schneider]

1 minute ago, Kal Rubinson said:

Tellig ~ Gillet(e). Seemed obvious to me from the get-go.

 

How could it be obvious from the get-go if someone never heard the name Tom Gillet?

Doug

[Kal Rubinson]

2 minutes ago, Doug Schneider said:

 

How could it be obvious from the get-go if someone never heard the name Tom Gillet?

Doug

It was always in the magazine.

[crenca]

30 minutes ago, John_Atkinson said:

Thank you for your comment, Doug.

I believe the correct pseudonym is Archimago.

 

John Atkinson

Editor, Stereophile

 

 

I had to give you a thumbs up with so succinctly sticking with misstatement of facts and non-engagement.  However, in this kind of medium the occasional misspelling or grammatical error is to be overlooked as it is part of the efficiency of communication....

[beetlemania]

7 minutes ago, Kal Rubinson said:

It was always in the magazine.

 

Waat?

 

I just pulled out the November 2008 issue. I see "Sam Tellig" on the masthead, not so much "Tom Gillette".

 

But this is way off topic. Mr. Atkinson has the chops to confirm or rebut the technical material in this article. Not a good look that he instead chooses to write about the author's choice of a pseudonym.

 

It's increasingly obvious that Mr. Atkinson and Stereophile made an embarrassing mistake by climbing on the MQA train. But instead of explaining how you were led astray you just keep digging. Oy.

[kumakuma]

10 minutes ago, Kal Rubinson said:

It was always in the magazine.

 

Where?

 

I checked the masthead of a number of back issues and only saw the pseudonym listed.

 

I'm guessing that his real identify was known by those in the industry but not common knowledge to regular folks like myself.

[mrvco]

4 hours ago, james45974 said:

Yes, but a no upside calculation for MQA or Stereophile is not necessarily neutral, there could be a downside, like to your reputation!

 

Clearly Bob and whoever invested in MQA has a bunch of money and pride invested in MQA, so I don't see them throwing in the towel easily, I can't imagine that there is much of a pivot that they can employ at this point either.  Maybe they can sell MQA to T-Mobile 

 

The pro-MQA 'journalists' have a bunch of pride (hopefully no financial interest) invested in MQA and they need a way to save face.  Maybe one of them could break the ice and declare that MQA is a technology designed for a 3G / 10M residential broadband world and that MQA is no longer worth the hassle in a 4G / 100M+ residential broadband world and hi-res FLAC is viable and the way forward.

[mansr]

46 minutes ago, Doug Schneider said:

I actually found two bits of information about this from Thierry Heeb (he was with Anagram): "Oversampling is an upsampling process where the ratio between output sampling frequency and input sampling frequency is an integer larger than 1. Upsampling is any kind of transformation providing an output sampling frequency that is higher than the input sampling frequency and not necessarily a ratio."

I've never seen that definition before. The fact is, there is no single, universally agreed upon definition of these terms.

 

In some contexts, oversampling means sampling an analogue signal at a rate higher than Nyquist, as opposed to undersampling, which means sampling at a lower rate. When talking about all-digital processes, the definitions I've mostly encountered, for instance in MATLAB, are:

• Resampling: band-limited interpolation to a new sample rate, either lower or higher than the original.
• Oversampling: resampling to a higher rate.
• Upsampling: increasing the sample rate by an integer factor by repeating samples or inserting zeros.
• Downsampling: decreasing the sample rate by discarding samples.

However, oversampling and upsampling are often, especially in audio, used interchangeably, just as downsampling is frequently used to mean band-limited resampling to a lower rate. It is generally obvious from context which meaning the author intends.

[kumakuma]

29 minutes ago, crenca said:

 

I had to give you a thumbs up with so succinctly sticking with misstatement of facts and non-engagement.  However, in this kind of medium the occasional misspelling or grammatical error is to be overlooked as it is part of the efficiency of communication....

 

Not to mention mixed metaphors...

 

19 minutes ago, beetlemania said:

It's increasingly obvious that Mr. Atkinson and Stereophile made an embarrassing mistake by climbing on the MQA train. But instead of explaining how you were led astray you just keep digging. Oy.

 

 

[Kal Rubinson]

19 minutes ago, kumakuma said:

Where?

I checked the masthead of a number of back issues and only saw the pseudonym listed.

 

I'm guessing that his real identify was known by those in the industry but not common knowledge to regular folks like myself.

I have not checked the old issues but I recall knowing this for many years prior to any association with the magazine.  So, perhaps, it was just due to a life-long fascination with word-play.  

 

Anyway, this has nothing to do with the thread.

[Doug Schneider]

6 minutes ago, mansr said:

I've never seen that definition before. The fact is, there is no single, universally agreed upon definition of these terms.

 

100% agree! When I spoke with Thierry about that, he represented it as his way to distinguish. I appreciated it, because it's the only time I've heard someone in the know take a stab at it.

Doug

[Doug Schneider]

5 minutes ago, Kal Rubinson said:

I have not checked the old issues but I recall knowing this for many years prior to any association with the magazine.  So, perhaps, it was just due to a life-long fascination with word-play.  

 

Anyway, this has nothing to do with the thread.

 

Actually, it has a lot to do with the thread. John brought up that he won't engage in discussions about Archimago's article because of the pseudonym. Seems like a double standard to me -- and rather odd given that he simply jumped into this thread, but didn't really want to partake. Nevertheless, to me:

 

1) You'd have to be high on something to believe Archimago is ???? ????'s real name.

 

2) You'd certain be given a lot of slack thinking that Sam Tellig was his real name since it actually sounds like it would be someone's name andhe wrote monthly for three decades and no one ever did what Connaker did with Archimago's article -- admit up front that it's not his real name, but that he vetted him first.

 

Doug

[Doug Schneider]

12 minutes ago, mansr said:

• Oversampling: resampling to a higher rate.
• Upsampling: increasing the sample rate by an integer factor by repeating samples or inserting zeros.

This is exactly opposite to what Thierry said. Also, 44.1 to 192 isn't integer. Are you sure that's not reversed?

 

Doug

[Doug Schneider]

50 minutes ago, Kal Rubinson said:

It was always in the magazine.

 

Where? I'd learned it on a forum somewhere.

 

Mind you, a Google search of "Sam Tellig Tom Gillett Stereophile.com" brought up this: "Sam begins his Stereophile career with a survey of moving-magnet phono cartridges, quickly becomes our most popular writer, and remains so to this day. But the magazine also benefits over the next 15 years from the circulation-promotion expertise of Sam's real-life alter ego, Tom Gillett."

So Sam was the face to the public. Tom increased circulation. Yet Achimago has a career outside of audio.

 

Doug

[mansr]

3 minutes ago, Doug Schneider said:

This is exactly opposite to what Thierry said. Also, 44.1 to 192 isn't integer. Are you sure that's not reversed?

I'm sure. See MATLAB documentation:

https://www.mathworks.com/help/signal/ref/upsample.html

https://www.mathworks.com/help/signal/ref/downsample.html

https://www.mathworks.com/help/signal/ref/resample.html

These definitions are probably the most widely used ones.

[Doug Schneider]

23 minutes ago, mansr said:

I'm sure. See MATLAB documentation:

https://www.mathworks.com/help/signal/ref/upsample.html

https://www.mathworks.com/help/signal/ref/downsample.html

https://www.mathworks.com/help/signal/ref/resample.html

These definitions are probably the most widely used ones.

 

Would you agree with this definition of interpolation?

 

https://www.mathworks.com/help/signal/ref/interp.html

 

The "integer factor" keeps coming up.

 

Doug

[Kal Rubinson]

21 minutes ago, Doug Schneider said:

Mind you, a Google search of "Sam Tellig Tom Gillett Stereophile.com" brought up this: "Sam begins his Stereophile career with a survey of moving-magnet phono cartridges, quickly becomes our most popular writer, and remains so to this day. But the magazine also benefits over the next 15 years from the circulation-promotion expertise of Sam's real-life alter ego, Tom Gillett."

I saw that and thought about posting it with an emphasis on the word "also" in the second sentence.  The only reason I posted about this at all was that I had always thought his identity was obvious to me.

[wdw]

1 hour ago, Kal Rubinson said:

Tellig ~ Gillet(e). Seemed obvious to me from the get-go.

 

but hey, really, who gives a squeak about this issue...Tellig was entertaining!  I am serious.  His Russian sweetie, his love of all those funny and noisy bits of electrical devices.  I alway went to his page right after opening the cover.  Was never looking to be persuaded on any issue but liked his prose and attitudes.  To my mind Dudley is in the same realm.  Great evocation of a musical ideal although I personally suspect his system has a fair bit of low level electrical noise that would be remarkable to most of us. 

Comforting, tubey, woody, the older noisy needle in the groove, etc...totally loved reading about it.  Then there is his tractor!

 

.....totally OT but just bought (3 hours ago)a  PONO Player at a serious discount and listening through Etymotic in ear monitor and the music (Takac  Quartet of Mr. Bs Late Quartets) is truly 3D...so very surprisingly so.  I had to look up a couple of time around the room.  

 

So very sad about Charley Hansen.

 

Run to get any that are now wholesaled out of the store.  

 

My young and dear hearted sales assistant was ernestly asking me why I was so interested in the product so I began by saying, well you know, Neil Young, but that wasn't working at all...so I regrouped and started in about levels of resolutions...but no further success.  She is taking music something at a local college but they haven't gifted her with any understanding of these very modest technical issues.

[mansr]

46 minutes ago, Doug Schneider said:

Would you agree with this definition of interpolation?

 

https://www.mathworks.com/help/signal/ref/interp.html

 

The "integer factor" keeps coming up.

Interpolation in general means calculating additional values between two given values. Applied to a sequence of samples, this naturally results in an integer multiple.