Stereophile | January 2017 Issue

[LL21]

Hi Gents, still enjoying the tennis...let me ask a few questions about SETs where I admit i am VERY curious.

1. I [believe] I read/recall that well-designed SETs in the 1 watt range can have extremely linear response...full spectrum, not flagging or weakening below 30hz or something.

Can someone (i am pretty Atmasphere has done this) actually give some scientific info around this?

2. If so, then with a 105db sensitive speaker, we are able to drive that horn to 117db at 1m with a 20 watt SET...add the 2nd speaker, reduce for the listening distance,etc, etc...but certainly loud. Then back it up to a base listening level of 85db, we remain well below the 1-watt range (where distortion is supposedly super low, and FQ super-linear) and with music that has a positive dynamic peak of 23db above our base listening level, we are still at only 2 watts of that 20-watter.

IFF (as mathematicians say, If any Only If) SETs at 1-2 watts are considered supremely linear (in all manner of various ways of measuring distortion, IMD, 2nd, 3rd, etc, etc)...isn't this an extremely interesting (or surely super-legit) place to start building a system (on the basis your speaker list is in the 100-105db sensitive range)?

3. And to Robh3606's point in his chart made about the potential for very attractively low distortion levels of a 96db speaker when driven by a capable amp (at certain listening levels)...does this not further support/make SETs an EXTREMELY interesting place to be?

Trust me, I am a SS Class A guy who loves what he has and considers himself blessed to have it...but i am trying to get my head around SETs properly. So thanks for a bit of math around exactly what the distortion levels are of super-well designed SETs AT THE 1-3 WATT LEVEL.

 

[morricab]

Hi Gents, still enjoying the tennis...let me ask a few questions about SETs where I admit i am VERY curious.

1. I [believe] I read/recall that well-designed SETs in the 1 watt range can have extremely linear response...full spectrum, not flagging or weakening below 30hz or something.

Can someone (i am pretty Atmasphere has done this) actually give some scientific info around this?

2. If so, then with a 105db sensitive speaker, we are able to drive that horn to 117db at 1m with a 20 watt SET...add the 2nd speaker, reduce for the listening distance,etc, etc...but certainly loud. Then back it up to a base listening level of 85db, we remain well below the 1-watt range (where distortion is supposedly super low, and FQ super-linear) and with music that has a positive dynamic peak of 23db above our base listening level, we are still at only 2 watts of that 20-watter.

IFF (as mathematicians say, If any Only If) SETs at 1-2 watts are considered supremely linear (in all manner of various ways of measuring distortion, IMD, 2nd, 3rd, etc, etc)...isn't this an extremely interesting (or surely super-legit) place to start building a system (on the basis your speaker list is in the 100-105db sensitive range)?

3. And to Robh3606's point in his chart made about the potential for very attractively low distortion levels of a 96db speaker when driven by a capable amp (at certain listening levels)...does this not further support/make SETs an EXTREMELY interesting place to be?

Trust me, I am a SS Class A guy who loves what he has and considers himself blessed to have it...but i am trying to get my head around SETs properly. So thanks for a bit of math around exactly what the distortion levels are of super-well designed SETs AT THE 1-3 WATT LEVEL.

An Australian mag measured the KR SXi at 0.49% at 1 watt and it was nearly all 2nd harmonic with 0.01% 3rd and anything else was buried in the baseline. Even lower powers will have lower distortion.

There are other SETs that will have lower THD at 1 watt but not a better spectrum.

 

[LL21]

An Australian mag measured the KR SXi at 0.49% at 1 watt and it was nearly all 2nd harmonic with 0.01% 3rd and anything else was buried in the baseline. Even lower powers will have lower distortion.

There are other SETs that will have lower THD at 1 watt but not a better spectrum.

http://www.stereomojo.com/KR Audio ...ew/KRAudioKronzillaSX1MkIIAmplifierReview.htm

Thanks! Sounds like a great piece...is it 50 watts Class A? Still probably want to keep it to those first few watts and leave the rest for dynamic peaks i suppose.

 

[LL21]

An Australian mag measured the KR SXi at 0.49% at 1 watt and it was nearly all 2nd harmonic with 0.01% 3rd and anything else was buried in the baseline. Even lower powers will have lower distortion.

There are other SETs that will have lower THD at 1 watt but not a better spectrum.

Is the KR DM http://www.kraudio.com/index.php?op...e&id=23:kronzilla-dm&catid=3&Itemid=6&lang=en their flagship SET Class A amp?

 

[opus112]

Both papers can be found on Geddes website

I'd appreciate a more specific heads up on the claim that 'IMD isn't correlated with SQ'. I've read various papers on his website and never come across this very general claim. So which detail am I missing?

Catching up on posts I see you've cited this from Geddes :

"Furthermore, 1% THD is not at all the same as 1% IM, but we will show that neither correlates with subjective perception. While some of the signal-based metrics may be “better” than others, it is our opinion they all fall short of what we are seeking."

However that's not the same as your claim that Geddes said IMD isn't correlated with SQ. He's just saying one particular measure of IMD isn't, which I agree with. So if this Geddes statement is what you've based your claim on, then I see you've misunderstood either him or me (or perhaps both).

 

[opus112]

IFF (as mathematicians say, If any Only If) SETs at 1-2 watts are considered supremely linear (in all manner of various ways of measuring distortion, IMD, 2nd, 3rd, etc, etc)...isn't this an extremely interesting (or surely super-legit) place to start building a system (on the basis your speaker list is in the 100-105db sensitive range)?

If Lloyd you're relying on the conventional measurements to gauge linearity then there are much better performers out there in terms of the numbers than SETs. (But I reckon they won't sound as good.) One example which has exemplary measurements is a chipamp called 'Modulus-86' which has been designed to get those numbers as low as humanly possible.

 

[f1eng]

From Geddes (2003):
"This is precisely where the signal-based distortion metrics fail. In our next paper we will show that .01% THD of one type of nonlinear system can be perceived as unacceptable while 10% THD in another example is perceived as inaudible. Even one of these simple examples is sufficient to invalidate THD as a viable metric for discussion of the perception of distortion. Furthermore, 1% THD is not at all the same as 1% IM, but we will show that neither correlates with subjective perception. While some of the signal-based metrics may be “better” than others, it is our opinion they all fall short of what we are seeking."

"One reason that the perception of nonlinear distortion is so complex is that the hearing mechanism itself is not linear and taken as a “system” it is also quite complex. It should thus be expected that it will be a difficult task to ascertain what levels and types of nonlinearity the ear can perceive and even more difficult will be the scaling of the subjective impression of these nonlinear functions."

I suggest you read the rest of the paper: http://www.gedlee.com/downloads/Distortion_AES_I.pdf

and it's followup:

http://www.gedlee.com/downloads/Distortion_AES_II.pdf

0.5% of pure 2nd harmonic added is almost certainly inaudible...

Basically, Geddes findings mirror CHeever's master thesis pretty closely. Both support what I have been telling you and not do support how firmly you feel about the subject.

I actually don't feel that strongly about it, just playing devil's advocate really.

It is quite easy to find a piece of work supporting almost any theory on the internet, we live in a "post-truth" information environment after all One researcher concludes one thing, a different researcher comes to a different conclusion.

In the case of lots of things, not just SET amps (which I have enjoyed when I listened to one, but not enough to buy despite using 109dB/watt horns) people seek confirmation of their point of view rather than a challenge.

I am an engineer. It seems to me likely that if a piece of kit has known inherent weaknesses in performance but is preferred it is much more likely that it is the manifestation of these shortcomings which is being preferred rather than some yet-to-be-properly-explained mysterious superiority of some sort.

That is all, just the opinion of an engineer who has investigated several things up to the (at the time) state of the art (I am now retired) who never found a single thing which was not in the end explained by the basic science.

In the normal commercial world if you can convince customers of the superiority of your product sufficiently well you will have success whether the product actually is superior or not.

In Formula 1 all the convincing, marketing and press interviews come to nothing if your car isn't fast or relable enough. That is a world I know and have had the elation and pain of that reality. My other world was noise and vibration research which, in the early days, meant often designing and making one's own custom transducers and recorders. It was fun and educational. Nowadays both transducers and recorders are incredibly cheap and freely available in a multitude of types. This is great but it does mean most of the people using them do not deeply understand how they work and their limitations.

Anyway I have been devil's advocate long enough. IME HiFi is in many ways like fashion and politics rather than engineering

 

[morricab]

I'd appreciate a more specific heads up on the claim that 'IMD isn't correlated with SQ'. I've read various papers on his website and never come across this very general claim. So which detail am I missing?

Catching up on posts I see you've cited this from Geddes :

"Furthermore, 1% THD is not at all the same as 1% IM, but we will show that neither correlates with subjective perception. While some of the signal-based metrics may be “better” than others, it is our opinion they all fall short of what we are seeking."

However that's not the same as your claim that Geddes said IMD isn't correlated with SQ. He's just saying one particular measure of IMD isn't, which I agree with. So if this Geddes statement is what you've based your claim on, then I see you've misunderstood either him or me (or perhaps both).

Read the papers first then shoot from the hip. Seek and ye shall find.

 

[Folsom]

Distortion... the most boring conversation that was dead in the 70's... lives on like some galaxy consuming dead horse...

But I'll note it's only voltage distortion being measured, ah-thank you.

 

[LL21]

If Lloyd you're relying on the conventional measurements to gauge linearity then there are much better performers out there in terms of the numbers than SETs. (But I reckon they won't sound as good.) One example which has exemplary measurements is a chipamp called 'Modulus-86' which has been designed to get those numbers as low as humanly possible.

Hi Opus112, great to see you here again and remember well all of your many helpful and specific technical posts. In the case of my questions, I am not looking for conventional measurements...in fact, the reason for my posts is trying to get to some 'basic understanding' of the correct technical measurements (which i am now also suggesting from the various posts here MUST include both the equipment, the interface with other equipment, AND the physiology of the human ear).

And on that basis, if we at least seek to produce the correct technical measurements for the AMP to start with (whatever those technical measurements are for accurate reproduction...in layman's terms, distorting the signal as little as possible), the following question is if it is more straightforward to build a low-watt, supremely accurate/honest/truthful, etc amp, then does that take us back to the overall system design of a super-efficient speaker with a low-watt, super-low distortion or 'super high fidelity' amp?

 

[opus112]

Hi Opus112, great to see you here again and remember well all of your many helpful and specific technical posts.

Thanks Lloyd, I also remember well your interesting and thoughtful questions which made the other 'side' of the 'tango'.

In the case of my questions, I am not looking for conventional measurements...in fact, the reason for my posts is trying to get to some 'basic understanding' of the correct technical measurements (which i am now also suggesting from the various posts here MUST include both the equipment, the interface with other equipment, AND the physiology of the human ear).

I don't believe there are such things as 'correct' measurements. There are helpful ones, and less helpful and lastly, largely irrelevant ones. There's a spectrum of usefulness.

And on that basis, if we at least seek to produce the correct technical measurements for the AMP to start with (whatever those technical measurements are for accurate reproduction...in layman's terms, distorting the signal as little as possible), the following question is if it is more straightforward to build a low-watt, supremely accurate/honest/truthful, etc amp, then does that take us back to the overall system design of a super-efficient speaker with a low-watt, super-low distortion or 'super high fidelity' amp?

In my experience yes, its more straightforward to build a low wattage transparent amp than a high wattage one. I've been working on such a thing for the past few weeks, turns out to be about 15W per channel. Getting more power without losing quality is the next design challenge. Its a widely appreciated observation that high efficiency speakers sound more dynamic than low efficiency ones, I put this down to the difficulty of building high powered amps with the highest level of dynamics. That in turn stems from the physical/electrical limitations of capacitors which are the primary power source in practically all amps.

 

[bonzo75]

Its a widely appreciated observation that high efficiency speakers sound more dynamic than low efficiency ones, I put this down to the difficulty of building high powered amps with the highest level of dynamics. That in turn stems from the physical/electrical limitations of capacitors which are the primary power source in practically all amps.

It is always widely appreciated that for many high efficiency speakers it is tough to integrate the bass with the mids and tops (hence the various hybrids), affecting overall dynamics. When the tweeter moves fast and the bass does not - there are extremely few high efficiency speakers out there that do dynamics across the spectrum.

 

[microstrip]

(...) I am an engineer. It seems to me likely that if a piece of kit has known inherent weaknesses in performance but is preferred it is much more likely that it is the manifestation of these shortcomings which is being preferred rather than some yet-to-be-properly-explained mysterious superiority of some sort.

That is all, just the opinion of an engineer who has investigated several things up to the (at the time) state of the art (I am now retired) who never found a single thing which was not in the end explained by the basic science.(...)

Did you you ever investigate any thing that involved psychoacoustics or any kind of perceptual science?

 

[opus112]

It is always widely appreciated that for many high efficiency speakers it is tough to integrate the bass with the mids and tops (hence the various hybrids), affecting overall dynamics. When the tweeter moves fast and the bass does not - there are extremely few high efficiency speakers out there that do dynamics across the spectrum.

Makes perfect sense yes, even though it was something I was previously unaware of. Either you have to slow the HF down by deliberately selecting a less dynamic amp or you go in search of the more dynamic bass amp....

 

[microstrip]

(...) I don't believe there are such things as 'correct' measurements. There are helpful ones, and less helpful and lastly, largely irrelevant ones. There's a spectrum of usefulness. (...)

Good point!

 

[Believe High Fidelity]

Hi Gents, still enjoying the tennis...let me ask a few questions about SETs where I admit i am VERY curious.

1. I [believe] I read/recall that well-designed SETs in the 1 watt range can have extremely linear response...full spectrum, not flagging or weakening below 30hz or something.

Can someone (i am pretty Atmasphere has done this) actually give some scientific info around this?

2. If so, then with a 105db sensitive speaker, we are able to drive that horn to 117db at 1m with a 20 watt SET...add the 2nd speaker, reduce for the listening distance,etc, etc...but certainly loud. Then back it up to a base listening level of 85db, we remain well below the 1-watt range (where distortion is supposedly super low, and FQ super-linear) and with music that has a positive dynamic peak of 23db above our base listening level, we are still at only 2 watts of that 20-watter.

IFF (as mathematicians say, If any Only If) SETs at 1-2 watts are considered supremely linear (in all manner of various ways of measuring distortion, IMD, 2nd, 3rd, etc, etc)...isn't this an extremely interesting (or surely super-legit) place to start building a system (on the basis your speaker list is in the 100-105db sensitive range)?

3. And to Robh3606's point in his chart made about the potential for very attractively low distortion levels of a 96db speaker when driven by a capable amp (at certain listening levels)...does this not further support/make SETs an EXTREMELY interesting place to be?

Trust me, I am a SS Class A guy who loves what he has and considers himself blessed to have it...but i am trying to get my head around SETs properly. So thanks for a bit of math around exactly what the distortion levels are of super-well designed SETs AT THE 1-3 WATT LEVEL.

If you are trying to get your head around SET's I am surprised that distortion is the main topic of interest in motivating you to understand it better.

 

[morricab]

It is always widely appreciated that for many high efficiency speakers it is tough to integrate the bass with the mids and tops (hence the various hybrids), affecting overall dynamics. When the tweeter moves fast and the bass does not - there are extremely few high efficiency speakers out there that do dynamics across the spectrum.

You are not talking about actual time delay I hope. The delay from the crossoer is for sure greater than the "acceleration" differences between a bass driver in a vented box and a horn loaded mid or tweeter. That "delay" will not be what is causing a perception of slowness. Overhang and Qts of the box system though is another story. If someone is trying to get a more "powerful" sounding bass they might opt for a loading scheme that gives a high Q, which will sound slower because of the stored energy.

A crtically damped box will have the "speed" one needs but probably not the sensitivity. Or if it gets the sensitivity it probably won't have as much deep bass due to a higher Fs.

The matching, which is hard is either the Q of the design to get max slam and sensitivity or good enough quality but either too low sensitivity or not deep enough bass.

A horn solves this if it is long enough...if not then it won't go deep enough.

 

[morricab]

I actually don't feel that strongly about it, just playing devil's advocate really.

It is quite easy to find a piece of work supporting almost any theory on the internet, we live in a "post-truth" information environment after all One researcher concludes one thing, a different researcher comes to a different conclusion.

In the case of lots of things, not just SET amps (which I have enjoyed when I listened to one, but not enough to buy despite using 109dB/watt horns) people seek confirmation of their point of view rather than a challenge.

I am an engineer. It seems to me likely that if a piece of kit has known inherent weaknesses in performance but is preferred it is much more likely that it is the manifestation of these shortcomings which is being preferred rather than some yet-to-be-properly-explained mysterious superiority of some sort.

That is all, just the opinion of an engineer who has investigated several things up to the (at the time) state of the art (I am now retired) who never found a single thing which was not in the end explained by the basic science.

In the normal commercial world if you can convince customers of the superiority of your product sufficiently well you will have success whether the product actually is superior or not.

In Formula 1 all the convincing, marketing and press interviews come to nothing if your car isn't fast or relable enough. That is a world I know and have had the elation and pain of that reality. My other world was noise and vibration research which, in the early days, meant often designing and making one's own custom transducers and recorders. It was fun and educational. Nowadays both transducers and recorders are incredibly cheap and freely available in a multitude of types. This is great but it does mean most of the people using them do not deeply understand how they work and their limitations.

Anyway I have been devil's advocate long enough. IME HiFi is in many ways like fashion and politics rather than engineering

Its interesting you run your Anima's (we can agree on that speaker at least) with SS. The only time I didn't like their sound was with SS. They showed SS harshness and grain all too clearly IMO.

No disrespect, but your F1 engineering experience was focused on achieving one goal: Getting around a race track as fast as possible. It does not necessarily translate well to questions about human perception, although there are of course tangential points of interest, like vibrations.

I am of the thought that nearly all things in audio can be described accurately except one thing: Human perception, which is in the end the "detector" for the whole system. That reliance on a superbly sensitive but utterly unreliable detector makes the whole thing much more diffcult than any other objective type of engineering.

Because audio is ultimately only about human perception, of course salesmenship, marketing and visual "bling" matter a lot because people can be fooled, at least in the short run, by such things. That doesn't mean there is any less engineering in getting electronics that model human perception accurately...its just that most engineers do not even bother trying.

 

[andromedaaudio]

Its a bit off topic F 1 eng , but can you explain with basic science how this is made , mysteries exist after all , imo .

https://www.youtube.com/watch?v=CxgHeh9Mlrg



QUOTE=f1eng;429956]I actually don't feel that strongly about it, just playing devil's advocate really.

It is quite easy to find a piece of work supporting almost any theory on the internet, we live in a "post-truth" information environment after all One researcher concludes one thing, a different researcher comes to a different conclusion.

In the case of lots of things, not just SET amps (which I have enjoyed when I listened to one, but not enough to buy despite using 109dB/watt horns) people seek confirmation of their point of view rather than a challenge.

I am an engineer. It seems to me likely that if a piece of kit has known inherent weaknesses in performance but is preferred it is much more likely that it is the manifestation of these shortcomings which is being preferred rather than some yet-to-be-properly-explained mysterious superiority of some sort.

That is all, just the opinion of an engineer who has investigated several things up to the (at the time) state of the art (I am now retired) who never found a single thing which was not in the end explained by the basic science.

In the normal commercial world if you can convince customers of the superiority of your product sufficiently well you will have success whether the product actually is superior or not.

In Formula 1 all the convincing, marketing and press interviews come to nothing if your car isn't fast or relable enough. That is a world I know and have had the elation and pain of that reality. My other world was noise and vibration research which, in the early days, meant often designing and making one's own custom transducers and recorders. It was fun and educational. Nowadays both transducers and recorders are incredibly cheap and freely available in a multitude of types. This is great but it does mean most of the people using them do not deeply understand how they work and their limitations.

Anyway I have been devil's advocate long enough. IME HiFi is in many ways like fashion and politics rather than engineering [/QUOTE]

 

[LL21]

If you are trying to get your head around SET's I am surprised that distortion is the main topic of interest in motivating you to understand it better.

Hi...i am very intrigued by Aries Cerat. That aside, the reason i am interested in SETs and distortion is because:

1. the cliche is that they have high distortion and those who love them dont care...but in fact, proponents argue that at low wattage, they are not only very low distortion, but can compete or be superior [AT 1-3 WATTS] to that of many SS designs.

2. the kinds of distortion SETs produce...plus the physiological way the human ear perceives/processes sound...combine to possibly make a different kind of scientific case for SETs