MQA creator Bob Stuart answers questions.

[Blumlein 88]

The Unpredictable Joys of Analog recording.
http://www.endino.com/graphs/

 

[March Audio]

The Unpredictable Joys of Analog recording.
http://www.endino.com/graphs/

I am quite shocked by that, didnt expect perfection but.....

 

[solderdude]

As I mentioned, you may want to study contemporary mobile bluetooth speakers, especially those made by JBL and Bose. Some of them seemingly violate laws of acoustics, by producing what is perceived as substantial bass, from 2" speakers, and playing that way for hours, using a tiny battery. The first distortion (dropped bone-fide bass frequencies) is perceptually cancelled by another (added higher frequencies not present in the original signal).

What's it have to do with MQA, and high resolution where you want to sample 0.5uS in several small steps in order to reproduce it more accurately ?
What's it have to do with hi-fi reproducton ?

 

[solderdude]

So, you disagree with https://en.wikipedia.org/wiki/Distortion?

"Distortion is the alteration of the original shape (or other characteristic) of something. In communications and electronics it means the alteration of the waveform of an information-bearing signal, such as an audio signal representing sound or a video signal representing images, in an electronic device or communication channel."

One can differentiate between linear and non linear distortion. These don't 'cancel out' despite you thinking they can seemingly do... they don't.
It can 'suggest' something. My son has a little toy car that makes sounds (you know, small ceramic buzzer in it) that can seemingly make the sound as if a subwoofer is playing music loudly but consists of pulse trains of a high frequency that appears to make LF sound.
It is fun but don't see this used in any music recording nor any serious equipment intended for serious audio reproduction where higher res audio is used.
This is used for toys (small boomboxes are toys) because the kids like it or it is used for background music.
Serious audio is big boy toys ... there is a difference.

 

[Wombat]

Perhaps as long as there are believers, customers and the money keeps rolling in it won't be dying until the next best format comes along.

MQA is all Rongorongo to me: https://www.historicmysteries.com/rongorongo-the-easter-island-script/

It could be Wrongowrongo, too.

 

[Sergei]

What's it has to do with MQA, and high resolution where you want to sample 0.5uS in several small steps in order to reproduce it more accurately ?
What's it has to do with hi-fi reproducton ?

My conjecture was that the high-end analog audio delivery chain, in the state it evolved by the mid-1980s, is capable of reproducing music at a higher subjective quality than the commonly used objective measurements, such as THD and SNR, suggest.

I hypothesized that the neurophysiological mechanisms partially explaining such gap between objective measurements and subjective perception could be same or similar to those deliberately targeted by some of the manufacturers of the contemporary bluetooth speakers.

Additional reason for the subjectively perceived high accuracy of the high-end analog audio delivery chain can be explained by its high inherent temporal informational density, expressed in bits per second.

The "bit", in the case of analog recording media, roughly corresponds to its "grain" - the smallest distinguishable physical unit affecting the level of recorded signal.

That same Information Theory approach explains why 192/24 PCM is preferable over 44/16 PCM for delivery of complex music: its temporal informational density is 6.5 times higher.

However, providing these temporally dense bit streams is pointless if the delivery chain can't ultimately transform them back into an accurate analog waveform of air pressure. The accuracy involves the amplitude and timing aspects.

As one of the criteria of accuracy, I suggested that a tweeter shall exhibit a meaningful response when the amplitude value changes from one 192 KHz sample to the next one - that is, its characteristic reaction time shall be 5.2 microseconds or less.

 

[solderdude]

The 'trickery' used in studios to make instruments and music productions sound 'better' has nothing to do with the format being used nor the sample rate/bit depth. Nor with audibility tests using artificial signals intended to isolate a single aspect of hearing that hasn't been proven to work in a similar fashion when listening to recorded music. Perception is an interesting topic though.
It seems to me that a lot of the issues you find you really want to blame on the 'digital' medium rather than the brain.
Both digital and analog recordings can make use of the above mentioned 'trickery'. This is not limited to specific formats but can be more accurately recorded in a digital manner because of technical limits of analog being 'worse' than those of digital (when of good quality)

Grain structures aren't in anyway similar to samples nor comparable in any way. In mechanical constructions the actual limitations of the medium in general isn't caused by the structures of the info bearing materials but rather the 'read/write' mechanisms/interfaces and technical hurdles/limitations to make use of those mediums which all have technical shortcomings and issues that do not exist or in other forms in digital formats.

You insist on the timing importance but have not offered convincing proof nor succesfully countered nor been able to persuede folks to see it your way. Maybe, just maybe you are barking up the wrong tree based on incorrect? assumptions.

I understand the wish to 'combine' all kinds of research (relevant to music or not) and invent a theory that eases your mind and like to have it peer reviewed here.

 

[Sergei]

One can differentiate between linear and non linear distortion. These don't 'cancel out' despite you thinking they can seemingly do... they don't.

Let me give you an example. Let's say you are recording a music instrument with a fundamental frequency of 100 Hz. There are also natural harmonics at 200, 300, 400, 500 etc. Hz, which give the instrument its characteristic timbre.

Let's assume that you are listening to the recording of that music instrument through a speaker that distorts in such a way that it mostly introduces even-order harmonics. So the harmonics at 200, 400 etc. Hz will be reinforced, changing the perceived timbre of the instrument.

Now let's assume that the recording media, such as a magnetic tape, has a property of introducing mostly odd-order harmonics. So, it will emphasize the harmonics at 300, 500 etc. Hz.

The combined effect of the distortions introduced by the speaker and the recording media could be such that all the relevant harmonics would be reinforced in approximately equal measure, thus perceptually keeping the instrument's characteristic timbre close to what it would have been if none of the distortions were present.

Interestingly enough, the adjustments on the high-end studio tape machines of the 1970s and 1980s allow to vary the level of odd-order harmonics for up to 3%, which corresponds well to the typical THD of 2% at normal listening levels, characteristic for studio speakers of that era.

Such combined effect is fragile: for instance, it will break down if the actual speaker the end-user possesses is less distorting or more distorting. But such mismatch can be cured by an additional distortion introduced, let's say, by a vacuum-tube-based preamp, which will add "missing" even harmonics in case the speaker is "too accurate".

Another consideration is that a speaker, or a distorting pre-amp, will introduce harmonics based on harmonics already reinforced by the recording media. In some cases this will be benign, in others it may change the perceived characteristic timbre, or add more noise at higher octaves.

Tinkering with distortions in the delivery chain really can nearly mutually cancel them in certain cases, especially if the music is of the simpler variety, which quite a few audiophiles seem to prefer, judged by the audiophile recommended playlists, yet this is a rather fragile correction.

However, we shall not doubt that a certain combination of highly distorting media, player, preamp, amp, and speakers will just happen to cancel out the distortions in a specific music material in such a way that it will make the material sound "heavenly". Then we can argue with the audiophile, who built and tweaked this system, until the cows come home, yet this won't change a bit his or her subjective reality.

It can 'suggest' something. My son has a little toy car that makes sounds (you know, small ceramic buzzer in it) that can seemingly make the sound as if a subwoofer is playing music loudly but consists of pulse trains of a high frequency that appears to make LF sound.
It is fun but don't see this used in any music recording nor any serious equipment intended for serious audio reproduction where higher res audio is used.
This is used for toys (small boomboxes are toys) because the kids like it or it is used for background music.
Serious audio is big boy toys ... there is a difference.

I like this example. Illustrates yet another perceptual illusion. Pulse trains of high frequency are also used by class D amps, right? There must be a biological low-pass filter somewhere, for the illusion you demonstrated to work. Perhaps middle ear? Hard to tell without knowing the details of the pulse modulation method used.

 

[solderdude]

You are combining info found here and there that you feel have a relation but differ so much that they aren't comparable at all.

Class-D amps for instance have NOTHING to do with tricking the mind. Yes, both are digital. That's the only thing they have in common.

Distortion levels in speakers exist and have no relation to recordings. An analog recording with certain levels of distortion (linear and non linear) differ from digital recordings with very low linear and non linear distortion.
Some folks may well find the analog recording more pleasant. That does not mean it is 'better'. It just means some folks prefer it.
In a technical sense the digital recording is MUCH closer to the original recorded event.

It appears that your gripe is 'resolution' and claim analog is 'better' in capturing 'impulse events'.
A digital recording of an analog recording can fully capture that exact same recording easily.
Using 192 or higher is more than sufficient.
Fine ... you don't feel 44/16 is enough that's fine. You are not the only one thinking this.
For MOST folks it is more than enough. Why would you mind ?

 

[RayDunzl]

distortions

Unless I am mistaken, for the levels at which I critically listen, measurable distortions on my main system neatly fall into the range considered inaudible.

Example:

Mains
1kHz



31.5Hz



JBL LSR 308

 

[Blumlein 88]

Unless I am mistaken, for the levels at which I critically listen, measurable distortions on my main system neatly fall into the range considered inaudible.

Example:

1kHz

View attachment 27848 View attachment 27849

31.5Hz

View attachment 27850 View attachment 27851

That's why your speakers do so poorly on the Harman tests. Not enough balancing distortion. The missing ingredient to go with the spin o Rama data.

 

[RayDunzl]

That's why your speakers do so poorly on the Harman tests.

I included the Harman speakers now.

Obviously better, harmonically richer. And they get even way more better when pushed another 10dB.

 

[March Audio]

I like this example. Illustrates yet another perceptual illusion. Pulse trains of high frequency are also used by class D amps, right? There must be a biological low-pass filter somewhere, for the illusion you demonstrated to work. Perhaps middle ear? Hard to tell without knowing the details of the pulse modulation method used.

Nope.

There is a low pass electrical filter in the amp and of course the speaker is an additional low pass electrical and mechanical filter. What reaches your ear from a class D amp is entirely low frequency analogue.

 

[March Audio]

Class-D amps for instance have NOTHING to do with tricking the mind. Yes, both are digital. That's the only thing they have in common.

Correct, nothing to do with "tricking" the mind but Class D isnt digital, very analogue

 

[restorer-john]

but Class D isnt digital, very analogue

Tell that to the flash-based, 8 bit CMOS microcontroller (PIC16F690) sitting in every NC-400 module...

 

[March Audio]

Tell that to the flash-based, 8 bit CMOS microcontroller (PIC16F690) sitting in every NC-400 module...

...

 

[Sergei]

Nope.

There is a low pass electrical filter in the amp and of course the speaker is an additional low pass electrical and mechanical filter. What reaches your ear from a class D amp is entirely low frequency analogue.

Repeating: There must be a biological low-pass filter somewhere, for the illusion you demonstrated to work.

The illusion is created by a toy. The toy itself doesn't appear to have a dedicated low pass filter like a D amp has. Or a speaker capable of reproducing the lower octaves. Per description, it has a smallish piezo transducer. Don't know what signal the transducer is getting, yet it ought to be a high-frequency one, because of such transducer's typical frequency response.

Low-cost piezo transducers are often driven by PWM signals. That was the core of my analogy with D class amp. Somewhere between such piezo transducer and inner hair cells sits a counterpart of the class D amp low pass filter.

The filter can be mechanical: piezo transducer can be glued to a larger vibrating plate, which could be a toy part, either flexible or attached to flexible supports.

Mechanical low pass filter would add cost though. More likely, the toy uses the missing fundamental effect, thus employing the "biological low pass filter", creating an illusion of a subwoofer.

 

[solderdude]

It is just a normal piezo speaker. The effect you hear creates an illusion. because the high frequency it emits short bursts in a low frequency.
If I can find the car maybe I can record it. Quite certain even a phone mic can pick it up and record it faithfully.

Class D and class T amps do not create an illusion they just deliver a clean AC voltage between 0Hz and their max freq. (between 30kHz and 100kHz depending on design). The 'digital' part inside is just and intermediate 'step' a bit like an ADC-DAC chain in DSD format.

 

[nscrivener]

Mechanical low pass filter would add cost though. More likely, the toy uses the missing fundamental effect, thus employing the "biological low pass filter", creating an illusion of a subwoofer.

What?

What has the missing fundamental effect got to do with this? A speaker that cannot reproduce the fundamental, but can produce the harmonics, can result in the brain interpreting the sound as including the fundamental. That is nothing to do with any so called "biological low pass filter".

 

[Wombat]

Subjectivists discount psychoacoustics. They relate it to being 'psycho'.