Can amplifier speed and resolution be measured?

[amirm]

What was the hi res equipment source / attenuator / power amp you were comparing with ? Did it have known spec above 20khz
to properly compare with CD ?

Testing was done with headphones. And yes, the content was mastered in high-resolution (AIX Records) with content above 20 kHz.

Original:

Converted version:

These were a public test to see if people could tell the difference between high-res and CD by the author of the content himself (Mark Waldrep).

 

[amirm]

Discovery is always the better approach... Amir's comment looked to me rushed and a generalisation, to not exactly say what equipment as hi-res was actually being used against CD to be able to say if extended bandwidth might cause bass to change.

You are doing that about me (rushing to judgment). Go a little slower, measure up who is in the discussion for a while and then draw conclusions.

 

[amirm]

I think his approach of discovery is the better one to take

Well, nothing came, did it? First step would have been controlled, double blind test. That is what is missing after nearly two decades.

 

[Headphonaholic]

Unless I am mistaken a simple null test would answer any question or doubt here yeah?

 

[stereo coffee]

The history of 20-20khz is tied to the restrictions of the Nyquist /Shannon sampling theorem.
David Blackmers article and video simply identified and looked beyond these restrictions which
I think is pretty clever to step beyond restrictions

My own experience is with with circuitry using cascode and common base or gate circuits.

 

[Blumlein 88]

The history of 20-20khz is tied to the restrictions of the Nyquist /Shannon sampling theorem.
David Blackmers article and video simply identified and looked beyond these restrictions which
I think is pretty clever to step beyond restrictions

My own experience is with with circuitry using cascode and common base or gate circuits.

What are those restrictions of the Shannon-Nyquist theorem which make David Blackmer's concern valid. I've seen those statements by him, but haven't read them in several years. They struck me then, as now, like he made some implied statements leaving out just enough you couldn't know what exactly he meant. Nor has he filled in the gaps since that time that I'm aware of.

 

[maxxevv]

The question from me would be what constitutes " speed" and " resolution" in those subjective comments?

By speed, in most engineering measures of equipment, we usually reference to the "impulse" response. Be it the inertia of a motor winding up to a specific revolution per second rating or for an electrical pulse, the time taken to reach a certain pulse plateau be it voltage or current.

Much like in cars, where the 0-60 mph timing is a far more useful reference of speed in a car than opposed to going 200mph top-out. Where the latter would be more akin to the maximum frequency range or power output of an audio device.

As for resolution, its so absolutely superfluous in terms of audio. Because at best, its a subset of impulse response, in that it being able to reach required peaks and troughs of frequency fast enough not to truncate any frequencies that are present in a signal.

Are such stuff even perceivable by ear ?

 

[Sir Sanders Zingmore]

Nor has he filled in the gaps since that time that I'm aware of.

It’s a bit like digital reconstruction except that rather than properly interpolating between samples, the reader can can fill the gaps with any damn nonsense they please.

 

[Blumlein 88]

The question from me would be what constitutes " speed" and " resolution" in those subjective comments?

By speed, in most engineering measures of equipment, we usually reference to the "impulse" response. Be it the inertia of a motor winding up to a specific revolution per second rating or for an electrical pulse, the time taken to reach a certain pulse plateau be it voltage or current.

Much like in cars, where the 0-60 mph timing is a far more useful reference of speed in a car than opposed to going 200mph top-out. Where the latter would be more akin to the maximum frequency range or power output of an audio device.

As for resolution, its so absolutely superfluous in terms of audio. Because at best, its a subset of impulse response, in that it being able to reach required peaks and troughs of frequency fast enough not to truncate any frequencies that are present in a signal.

Are such stuff even perceivable by ear ?

Yeah, in perceptible terms resolution as used in photography or video doesn't fit much with hearing.

We all know and can see what happens if you in step wise fashion reduced resolution of an image. What is the equivalent of step wise resolution reduction of a sound file? Would it be reducing frequency response gradually? Would it be reducing dynamic range gradually or similarly letting noise levels grow till it obscures sound? Would it be some combination of reducing dynamic range and frequency response?

Of course if any of that is true the real issue is we have all the dynamic range and frequency response already that any human can hear. So called hires might be higher resolution, but only in a measured sense. Not at all in a perceptible sense. Just like at some viewing distance photo resolution exceeds the eye's angular resolution, and increasing it more does nothing to the seen image.

 

[amirm]

I think resolution is easy to address: person focuses on the music and hears more individual notes and says that is higher resolution. In reality this is one's perception changing because they were comparing things and so started to focus on small details. A blind test make them appear in both pieces of gear but audiophiles don't go there.

On speed, I think this is one of those random terms people use because they can't explain what they like better and just use that term. Again, this is a perception thing that goes away in controlled testing.

 

[maxxevv]

Yeah, in perceptible terms resolution as used in photography or video doesn't fit much with hearing.

We all know and can see what happens if you in step wise fashion reduced resolution of an image. What is the equivalent of step wise resolution reduction of a sound file? Would it be reducing frequency response gradually? Would it be reducing dynamic range gradually or similarly letting noise levels grow till it obscures sound? Would it be some combination of reducing dynamic range and frequency response?

It would be somewhat like lossy audio codecs using lower and lower bitrates I guess ? 96kbps versus 128kbps versus 256kbps vesus 320kbps MP3 files being played back perhaps ?

Of course if any of that is true the real issue is we have all the dynamic range and frequency response already that any human can hear. So called hires might be higher resolution, but only in a measured sense. Not at all in a perceptible sense. Just like at some viewing distance photo resolution exceeds the eye's angular resolution, and increasing it more does nothing to the seen image.

The human eye has a resolution limit of approximately 330dpi at closest focusing distance unaided.

Is there such a thing as an equivalent for hearing ?

 

[stereo coffee]

What are those restrictions of the Shannon-Nyquist theorem which make David Blackmer's concern valid. I've seen those statements by him, but haven't read them in several years. They struck me then, as now, like he made some implied statements leaving out just enough you couldn't know what exactly he meant. Nor has he filled in the gaps since that time that I'm aware of.

Hi Blumein 88
The restrictions of the Shannon /Nyquist theorem were that the first digital sampling occur at twice the highest frequency determined. Hence sampling was sadly determined at 44.1 khz which dictated that the highest audio frequency could not exceed 22.05 khz, or equally it was thought that 22.05 khz was beyond our perception of hearing which determined 44.1khz as the original CD sampling rate.
As such a restriction was created - that equipment manufacturers then sadly feverishly followed.

If we view before this, Quad as one such manufacturer, the 303 power amplifier measured 30 hz to 35 khz
and the amplifier following it the 405, measured 20 -20khz

There are many viewpoints that the 303 was /is the better sounding amplifier - half a tick I think for bandwidth being actually why

David passed away in 2002, his main body of work was in companding seen in the DBX products,but in later years he founded Earthworks
https://earthworksaudio.com/about-earthworks/ His same philosophy can be seen to push the boundaries of what we can perceive in audio. THAT Corporation have continued with his licensed ideas http://www.thatcorp.com

 

[stereo coffee]

The question from me would be what constitutes " speed" and " resolution" in those subjective comments?

By speed, in most engineering measures of equipment, we usually reference to the "impulse" response. Be it the inertia of a motor winding up to a specific revolution per second rating or for an electrical pulse, the time taken to reach a certain pulse plateau be it voltage or current.

Much like in cars, where the 0-60 mph timing is a far more useful reference of speed in a car than opposed to going 200mph top-out. Where the latter would be more akin to the maximum frequency range or power output of an audio device.

As for resolution, its so absolutely superfluous in terms of audio. Because at best, its a subset of impulse response, in that it being able to reach required peaks and troughs of frequency fast enough not to truncate any frequencies that are present in a signal.

Are such stuff even perceivable by ear ?

I like the analogy to cars, as it reveals Why certain circuitry works and other circuitry does not. If we have a car travelling without a driver ( ignoring now a today's reality ) it is a travelling object with much potential for damage and dreadful consequences. However if we placed a skilled driver into the equation of its motion we have comparison to how a good circuit can use speed to its advantage rather than watching it just fly by.

As example we could build a otherwise fast and powerful with current power supply ( just like the car without a driver ) and not harness the supply for when and where it is needed in a circuit, to the casual observer we have done a great job, but to the experienced observer it is just like potential for damage and consequences.

 

[Blumlein 88]

Hi Blumein 88
The restrictions of the Shannon /Nyquist theorem were that the first digital sampling occur at twice the highest frequency determined. Hence sampling was sadly determined at 44.1 khz which dictated that the highest audio frequency could not exceed 22.05 khz, or equally it was thought that 22.05 khz was beyond our perception of hearing which determined 44.1khz as the original CD sampling rate.
As such a restriction was created - that equipment manufacturers then sadly feverishly followed.

If we view before this, Quad as one such manufacturer, the 303 power amplifier measured 30 hz to 35 khz
and the amplifier following it the 405, measured 20 -20khz

There are many viewpoints that the 303 was /is the better sounding amplifier - half a tick I think for bandwidth being actually why

David passed away in 2002, his main body of work was in companding seen in the DBX products,but in later years he founded Earthworks
https://earthworksaudio.com/about-earthworks/ His same philosophy can be seen to push the boundaries of what we can perceive in audio. THAT Corporation have continued with his licensed ideas http://www.thatcorp.com

Feverishly follow is another description for following standards.

BTW, the 405 was a bit wider band than the 303. The specs if you include db relative to 1khz show this. They were nearly the same on the upper end and the 405 extended further on the low end. .5 db vs 1db vs 3db and all that.

I record with some Earthworks mics. I know who is he is and his accomplishments. His statement's still lacking in regards to problems with Shannon Nyquist. Sounds like someone applying analog thinking to digital.

The reason for extended bandwidth in amps is to achieve truly flat response 20-20,000 hz. In conventional amps you need a -3 db down spec at 200 khz to keep 20 khz to - .1 db. In digital you can manage that with less excess bandwidth. A sampling rate of 50 to 65 khz might have been nice.

 

[solderdude]

@stereo coffee Could you please explain why cascode and common base or gate circuits would be superior over many other conventional circuits and how that would translate to a better sound ?

Do you have any proof in the form of nulls or simulations that shows the distortion is audibly lower (note the word audibly) with real world speaker loads ?

Can you post the circuit you claim sounds audibly better ?

 

[SIY]

Or, as I asked before, actual evidence of audible differences between engineered amplifiers?

Cascodes are a particular topology that is sometimes useful, sometimes not. It's a design tool, not an indicator of some magic amplifier quality. A standard Lin topology can also give vanishingly low distortion and flat frequency response (see Douglas Self's "Audio Power Amplifier Design" and "Baxandall and Self on Audio Power"). Ditto about a zillion other topologies (see Bob Cordell's excellent book, "Designing Audio Power Amplifiers" and Kolinumi Arto's "Audio Power Amplifiers").

 

[stereo coffee]

@stereo coffee Could you please explain why cascode and common base or gate circuits would be superior over many other conventional circuits and how that would translate to a better sound ?

Do you have any proof in the form of nulls or simulations that shows the distortion is audibly lower (note the word audibly) with real world speaker loads ?

Can you post the circuit you claim sounds audibly better ?

Your asking for quite a bit there, your first question is partially answered by referring to page 121 of Designing with field Effect Transistors 2nd edition Siliconix Inc where it is explained "common mode voltage , common- mode input resistance, and common mode rejection ratio can be improved by the addition of another pair of fets connected Cascode "

If we view any of the listed features of a cascode as possibly improving audio then Yes, its the real deal -better audio will be the result.

The circuit I use these techniques in is not an amplifier, so no speaker loads- rather a circuit between a 12v AC power supply, the anodes of a quad of NSL32SR3 LDR's , the cathodes of the same set of NSL32SR3 , a 50k potentiometer placed between the cathodes , a wiper at a lower potential difference, additional sensing circuitry and finally DC ground.

I have asked Armin to measure - so in the fullness of time your question concerning distortion may indeed be answered.

Here is some insight into the rectifier section of my circuit. It uses 2 x slave bridge rectifiers part MS06 , at the negative of each AC secondary phase a mosfet is placed in common source form, with attendant resistance to each gate to ground. The gate is connected via 220 ohms of resistance to each negative of the slave rectifiers. AC is between the drain of the mosfets and the anode of a thyristor pair for each phase, each gate of that thyristor pair is connected to the eventuated DC supply - as such the circuit requires additional small push to turn on - but is controlled by the voltage / current it then produces ( like stepping on board a train seemingly already at your destination ) A- P channel mosfet is connected as a common gate by an additional pair of thyristors and 10 ohms of resistance at each thyristor cathode. The gate of the P channel receives a small negative voltage derived from 2x TL431 connected with ref to the gates of the mosfet pair anodes to DC ground and cathodes to the P channel mosfet. Interestingly a PNP transistor passively at extremely high impedance has this capability, a TL431 passively slightly better
and a active TL431 pair produces over half a volt DC negative

The advantage even of this small section of the circuit is obvious to hear ( of course subjective ) and shows as enormous depth, and clarity with a pair of Quad 306 as amplifiers and speakers as Jr149 , sources being Marantz CD players CD7300, CD50 and Yamaha CDRHD1500
A earlier bunch of reviews and being discussed independently but of course subjectively is in many other forums.

 

[solderdude]

Thanks for the description of the circuit.

About LDR volume control: https://www.neurochrome.com/tortuga-audio-ldr3/

The question is what are the specifications and measurements. I am quite sure any decent sound card can be used to show superior performance.
Where is it shown that it is audibly 'better' than other designs and have you made any nulls showing the output is unchanged or has changed in a euphonic way.
Subjective impressions are just that. They can't be submitted as evidence of superiority. Hearing is flawed as an analyzer, that's why real analyzers exist and null tests can be used as 'proof' with real loads and real music.

 

[SIY]

The advantage even of this small section of the circuit is obvious to hear ( of course subjective )

By "subjective," you are meaning "without even basic controls"?

 

[stereo coffee]

Thanks for the description of the circuit.

About LDR volume control: https://www.neurochrome.com/tortuga-audio-ldr3/

The question is what are the specifications and measurements. I am quite sure any decent sound card can be used to show superior performance.
Where is it shown that it is audibly 'better' than other designs and have you made any nulls showing the output is unchanged or has changed in a euphonic way.
Subjective impressions are just that. They can't be submitted as evidence of superiority. Hearing is flawed as an analyzer, that's why real analyzers exist and null tests can be used as 'proof' with real loads and real music.

My product is not the Tortuga LDR3 ( I appreciate Morten's enormous work in LDR's and he and I on different sides of the globe -get on very well ) As I say I have asked Armin to measure my product by today, leaving a message at the designated spot. I appreciate your emphasis with this forum is with principally visual measurements, followed by audio impressions so we won't drift with subjective/ measured debates.