Amplifiers bench tests results-should we care or not, regardless of audibility, are those thresholds such as -100, -80, -40 dB THD+N, IMD etc.useful?

Are bench test results useful for decision making regardless?

  • Only if the results are all too close to or exceed thresholds of audibility

  • No, well designed amplifiers have flat response 20-20 kHz and THD+N below -60 dB (worst conditions)

  • Yes, because not all of the thresholds of audibility for various measurements are known


Results are only viewable after voting.
AcuDefTechGuy

AcuDefTechGuy

Audioholic Jedi
Sarcasm and jokes aside, the Blade One (the origin big one) was the best sounding speakers I have ever heard in a dealer's room. No subs needed imo but I think for you, at least one 18" sub will be needed to help those speakers.

I know THD>0.5% or even 1% would bother me, but I don't like any kind of noise. When I downgraded from my $3,000 Marantz AV8801 to the Denon AVR-X4400H, I was happy to "hear" the quietness of the Denon.

But no, I am not aware of speakers prone to noise. Also, speakers harmonic structures are quite different. They typically are mainly 2nd, 3rd harmonics, may be a little bit of 4th and 5th. So I understand speaker's harmonic distortions are far less offensive then amplifiers.

Since I know KEF speakers (the midrange and up models) in general have relatively low distortions so I will use the R3 as an example, the Blade will most likely be much better:

KEF R3 Speaker Review | Audio Science Review (ASR) Forum

See that the 4th and 5th harmonics were vanishingly low, well below 0.5%!

Keep in mind this is Amir's bench test using high quality instruments. If you use an amplifier that has 0.5% THD and if a significant % of that are the higher order 5th, 7th, 9th harmonics, the speaker will produce those and that's what you will hear. No one can hear the distortions of amplifiers, they hear them through their speakers.

Even if the amplifier's THD are only made up of the almost pleasing 2nd 3rd harmonics, since the speakers also produce 2nd and 3rd harmonics distortions, you will end up hearing the 4th, 8th, 9th, 15th harmonics etc., and it may not sound good to even you. (okay now I may be joking..).


View attachment 59410
Well, this is all just academic since in real life our AVR/AVP/Amp will likely have extremely low THD+N. Even the Denon x3800 has THD+N of 0.004%.

Some people probably think that I don’t mind having electronics with higher THD+N. But I do mind. Always have. My Yamaha CX-A5100 has THD+N of 0.003%, and I hope the next CX-A5300 has THD+N of 0.001% or better.

Also can’t wait to see SINAD on the Marantz AV-10.
 
D

dlaloum

Full Audioholic
That is absolutely not true. Toby and Dinsdale proceeded Peter on this by a wide margin. The problem is that those early designs blow up on a really regular basis.

Peter's achievement was to design and build the first really reliable solid state amp, the Quad 303 in 1969. These amps are incredibly reliable and seldom fail. The BBC replaced all their line amps with 303s. I bought my first 303 in 1970. I used it for years. I have not used it in a while, but as far as I know it still works. I have two of them.

Quad 303 domestic version.



Quad 303 pro version.



Those amps are so stable that you can parallel the inputs and outputs to bridge them. I know of no other power amp you can do that to. The rest blow up instantly.

They still fetch good money. If people want a really reliable amp, then they look for a Quad 303 still. They were class AB and I think the only class AB amp he produced. Then went on to produce the 405, 405-2, 606, and 909. He also made pro versions of all of them I think, for the BBC and studios. He also designed the amps for the legendary BBC transmission line monitors at Maida vale. They were an active design, in which Peter, Raymond Cooke of KEF, and Jim Rogers all cooperated. All had their own companies.

While we are mentioning these, four of the amp boards from those legendary speakers came up for sale, and I could not resist buying them. I have installed them in two 405-2 amp cases, as they are basically a special version of the Quad 405-2. I could not resist adding them to my museum. They are a special part of audio history.

In the 405 and 405-2 the output transistors are biased class C! There is feed forward error correction from a small class A amp. Performance is class A. The 606 and 909s output transistors are biased class B, again with feed forward error correction from a class A amp.

Nelson Pass did infringe on Peter's patents, but Peter never mentioned it himself or made a fuss. He was always conflicted about patents, and basically was of the view that good ideas should be freely distributed.

I have finally found a picture illustrative of the sort of cabinetry with high WAF seen in the 1960s

I believe the 303 is a very well done pure class B design - not AB....

(I too have one in storage...along with a pair of ESL's to match which need a service)
 
D

dlaloum

Full Audioholic
What’s the lowest SPEAKERS THD 20Hz-20kHz you’ve ever seen?
Quad ESL63

Quad used to advertise that THD was below 0.1% above 100Hz... below that, from memory, it peaked at something like 0.5% - an order of magnitude lower than almost all "dynamic" speakers. - I don't recall at what SPL level this was measured.

Later models(ESL2812) - were quoted at 100db SPL/wm - under 1% including the bass from 50 Hz up (higher frequencies at more moderate levels, are likely to be the same as for the ESL63...)
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
I believe the vast majority don't care about distortion levels that are not audible. Some will be willing to pay more for lower distortions regardless up to a point.

For me the only reason is I want the best specs I can afford regardless of audibility, and I did pay more for my Purifi amp vs the Hypex amp. The difference in cost was about $300. If it was more than that, I would have opted for the Hypex. Both have THD below the threshold, and they do sound the same.
THD has been one of the 'buzzword' specs since the late-1970s and consumers used to constantly come into the stereo store where I worked spouting specs they didn't understand. Questions like "How many Amps does this receiver put out?" were common, as was "What's the damping factor?" or "What's the Slew Rate?" with the latter two because of Sansui, who used to harp on them incessantly in their ads.
 
highfigh

highfigh

Seriously, I have no life.
Okay then, I guess based on published specs, a good example for the relatively more affordable speakers could be the US$35,000 KEF Blade:

Harmonic distortion 2nd and
3rd harmonics (90dB, 1m)
<0.5% 40Hz - 100kHz
<0.2% 200Hz - 2kHz
<0.1% 2kHz - 20kHz

The US$225,000 KEF Muon doesn't have published harmonic distortions specs so it may or may not be lower:

B&W's flagship, the $70,000 Nautilus does not have published specs either.

Expensive exotic speakers such as Wilson Audio's, Focal's don't published much specs neither.

My guess is, KEF Blade's 0.1 to 0.5% are probably among the best regardless of price.
I might be able to find out about the KEF speakers in greater detail soon- a friend recently had a great interview with them for the Midwest Regional sales manager role.
 
highfigh

highfigh

Seriously, I have no life.
Well, this is all just academic since in real life our AVR/AVP/Amp will likely have extremely low THD+N. Even the Denon x3800 has THD+N of 0.004%.

Some people probably think that I don’t mind having electronics with higher THD+N. But I do mind. Always have. My Yamaha CX-A5100 has THD+N of 0.003%, and I hope the next CX-A5300 has THD+N of 0.001% or better.

Also can’t wait to see SINAD on the Marantz AV-10.
You'll never hear the difference between .001 and 003%.
 
AcuDefTechGuy

AcuDefTechGuy

Audioholic Jedi
You'll never hear the difference between .001 and 003%.
Preaching to the choir. :D

Nor can we hear a difference between 0.0001% vs 0.007%. :D

But I think for some, this audibility isn’t the point of discussion. They want the best measurements. And so do I. But I am just not as hardcore as some guys. IOW, I’m not spending $3K on some 100W amp just for a few more SINAD numbers. I would rather spend it on more power, warranty, etc., and accept a few less SINAD numbers.

Again, aiming for SINAD of -100dB is their prerogative, just like accepting SINAD of -90dB is my prerogative (although I wouldn’t mind -200dB :D) or -80dB for others.

So maybe this thread is asking, what’s YOUR NUMBER? :D
 
highfigh

highfigh

Seriously, I have no life.
Preaching to the choir. :D

Nor can we hear a difference between 0.0001% vs 0.007%. :D

But I think for some, this audibility isn’t the point of discussion. They want the best measurements. And so do I. But I am just not as hardcore as some guys. IOW, I’m not spending $3K on some 100W amp just for a few more SINAD numbers. I would rather spend it on more power, warranty, etc., and accept a few less SINAD numbers.

Again, aiming for SINAD of -100dB is their prerogative, just like accepting SINAD of -90dB is my prerogative (although I wouldn’t mind -200dB :D) or -80dB for others.

So maybe this thread is asking, what’s YOUR NUMBER? :D
-200dB is impossible- I have heard that core of the Earth S/N is -164dB. Maybe, Gene or someone who has experience in electronic Engineering can confirm or deny this.
 
P

PENG

Audioholic Slumlord
Well, this is all just academic since in real life our AVR/AVP/Amp will likely have extremely low THD+N. Even the Denon x3800 has THD+N of 0.004%.

Some people probably think that I don’t mind having electronics with higher THD+N. But I do mind. Always have. My Yamaha CX-A5100 has THD+N of 0.003%, and I hope the next CX-A5300 has THD+N of 0.001% or better.

Also can’t wait to see SINAD on the Marantz AV-10.
I know you do mind, though not for audible concerns. You do tend to cite those single S&V style THD+N numbers that in general, don't tell much though they are still useful for quick apple to apple comparisons. I used to think that if S&V's, for example, the NAD T758V3's THD+N 0.4% at rated output, 1 kHz means there will probably be no audible concern. After seeing so many ASR's, Stereophile, and others more detailed measurements that show in some cases the distortions could be much higher when listening to real music, I no longer focus on those single THD numbers alone, unless there are other collaborating evidence provided in the reviews.

Instead of using the NAD that has unusually higher distortions even based on S&V's single number, take a look of the AVR-X3800H that, as you observed, has very low number, i.e. 0.004% measured at 1 kHz.

If you focus on the 1 kHz test, you can see that in the 10 to 135 W into 4 ohm output range, you do get about as you cited, 0.004%, +/- 0.001% or so.

Now look at the 10 kHz test you can see that the number would jump to 0.025.0%, still low enough for me to detect such distortions, but it is a whopping 6.25X jump! The story may not end there either, as such tests are typically done using resistor loads. Difficult speaker loads would most likely make things worse, not better.

Now that's starting from the 1 kHz test of the very low 0.004% level. So if you had the NAD T758V3 that a lot of their users like the sound, that has 0.4%, about 100X higher than the Denon's, what would it be if tested with a 10 kHz, or even 5 kHz signal? If it jumps 10X, that would be 4% and that would definitely be audible even to me and for sure to you since you are younger.:) Since even the 40+ population could hear high frequencies up to say 10 kHz or even 15 kHz, I would really like reviewers to use not only the 1 kHz but also the 5 kHz for THD measurements.


1673013086292.png


On the other hand, having audible harmonic distortions or even intermodulation distortions may not be always bad to all people.

As Nelson Pass mentioned on the Passlab website:
Pass LabsAudio distortion and feedback - Pass Labs

However when given a choice between the sound of an amplifier whose characteristic is dominantly 2nd harmonic versus 3rd harmonic, a good percentage of listeners choose the 3rd.
Personally, I don't want my amp's harmonic structure dictated by the manufacturers who don't aim for the lowest possible distortions. If I happen to enjoy some harmonic distortions (which I don't), I would much prefer to have the option to customize my own harmonics so that I can choose to add targeted harmonic distortions according to the structures I prefer, or choose "flat", that is, no intentional distortions. That's not unlike the now popular room correction software's feature that allows the user to customize the target curve.
 
P

PENG

Audioholic Slumlord
I might be able to find out about the KEF speakers in greater detail soon- a friend recently had a great interview with them for the Midwest Regional sales manager role.
I would also be interested to know why KEF seems to be the only popular speaker brand that published their distortions specs except for their top Muon models. Why aren't more manufacturers include distortions in their specs?
 
AcuDefTechGuy

AcuDefTechGuy

Audioholic Jedi
I know you do mind, though not for audible concerns. You do tend to cite those single S&V style THD+N numbers that in general, don't tell much though they are still useful for quick apple to apple comparisons. I used to think that if S&V's, for example, the NAD T758V3's THD+N 0.4% at rated output, 1 kHz means there will probably be no audible concern. After seeing so many ASR's, Stereophile, and others more detailed measurements that show in some cases the distortions could be much higher when listening to real music, I no longer focus on those single THD numbers alone, unless there are other collaborating evidence provided in the reviews.

Instead of using the NAD that has unusually higher distortions even based on S&V's single number, take a look of the AVR-X3800H that, as you observed, has very low number, i.e. 0.004% measured at 1 kHz.

If you focus on the 1 kHz test, you can see that in the 10 to 135 W into 4 ohm output range, you do get about as you cited, 0.004%, +/- 0.001% or so.

Now look at the 10 kHz test you can see that the number would jump to 0.025.0%, still low enough for me to detect such distortions, but it is a whopping 6.25X jump! The story may not end there either, as such tests are typically done using resistor loads. Difficult speaker loads would most likely make things worse, not better.

Now that's starting from the 1 kHz test of the very low 0.004% level. So if you had the NAD T758V3 that a lot of their users like the sound, that has 0.4%, about 100X higher than the Denon's, what would it be if tested with a 10 kHz, or even 5 kHz signal? If it jumps 10X, that would be 4% and that would definitely be audible even to me and for sure to you since you are younger.:) Since even the 40+ population could hear high frequencies up to say 10 kHz or even 15 kHz, I would really like reviewers to use not only the 1 kHz but also the 5 kHz for THD measurements.


View attachment 59420

On the other hand, having audible harmonic distortions or even intermodulation distortions may not be always bad to all people.

As Nelson Pass mentioned on the Passlab website:
Pass LabsAudio distortion and feedback - Pass Labs
Personally, I don't want my amp's harmonic structure dictated by the manufacturers who don't aim for the lowest possible distortions. If I happen to enjoy some harmonic distortions (which I don't), I would much prefer to have the option to customize my own harmonics so that I can choose to add targeted harmonic distortions according to the structures I prefer, or choose "flat", that is, no intentional distortions. That's not unlike the now popular room correction software's feature that allows the user to customize the target curve.
Yeah, no way I’m buying electronics with 0.4% THD+N. I will compromise a little (like 0.007% :D), but that’s asking for a lot more than I am willing. :D
 
M

Movie2099

Audioholic General
Yeah, no way I’m buying electronics with 0.4% THD+N. I will compromise a little (like 0.007% :D), but that’s asking for a lot more than I am willing. :D
The amps I want are only 0.02% THD......what should i do??? :p
 
highfigh

highfigh

Seriously, I have no life.
I know you do mind, though not for audible concerns. You do tend to cite those single S&V style THD+N numbers that in general, don't tell much though they are still useful for quick apple to apple comparisons. I used to think that if S&V's, for example, the NAD T758V3's THD+N 0.4% at rated output, 1 kHz means there will probably be no audible concern. After seeing so many ASR's, Stereophile, and others more detailed measurements that show in some cases the distortions could be much higher when listening to real music, I no longer focus on those single THD numbers alone, unless there are other collaborating evidence provided in the reviews.

Instead of using the NAD that has unusually higher distortions even based on S&V's single number, take a look of the AVR-X3800H that, as you observed, has very low number, i.e. 0.004% measured at 1 kHz.

If you focus on the 1 kHz test, you can see that in the 10 to 135 W into 4 ohm output range, you do get about as you cited, 0.004%, +/- 0.001% or so.

Now look at the 10 kHz test you can see that the number would jump to 0.025.0%, still low enough for me to detect such distortions, but it is a whopping 6.25X jump! The story may not end there either, as such tests are typically done using resistor loads. Difficult speaker loads would most likely make things worse, not better.

Now that's starting from the 1 kHz test of the very low 0.004% level. So if you had the NAD T758V3 that a lot of their users like the sound, that has 0.4%, about 100X higher than the Denon's, what would it be if tested with a 10 kHz, or even 5 kHz signal? If it jumps 10X, that would be 4% and that would definitely be audible even to me and for sure to you since you are younger.:) Since even the 40+ population could hear high frequencies up to say 10 kHz or even 15 kHz, I would really like reviewers to use not only the 1 kHz but also the 5 kHz for THD measurements.


View attachment 59420

On the other hand, having audible harmonic distortions or even intermodulation distortions may not be always bad to all people.

As Nelson Pass mentioned on the Passlab website:
Pass LabsAudio distortion and feedback - Pass Labs



Personally, I don't want my amp's harmonic structure dictated by the manufacturers who don't aim for the lowest possible distortions. If I happen to enjoy some harmonic distortions (which I don't), I would much prefer to have the option to customize my own harmonics so that I can choose to add targeted harmonic distortions according to the structures I prefer, or choose "flat", that is, no intentional distortions. That's not unlike the now popular room correction software's feature that allows the user to customize the target curve.
If the music has pure tones from specific instruments like pipe organ or synths, IM is a bad thing when two frequencies are played and the amp runs out of headroom- the sum & difference frequencies can be audible and those are far from musical, unless it's a song by Yoko Ono, Bjork, Erik Dolphy, etc.
 
P

PENG

Audioholic Slumlord
If the music has pure tones from specific instruments like pipe organ or synths, IM is a bad thing when two frequencies are played and the amp runs out of headroom- the sum & difference frequencies can be audible and those are far from musical, unless it's a song by Yoko Ono, Bjork, Erik Dolphy, etc.
I understand that but what is the threshold for IMD? By the way, if THD is low enough, IMD will be most likely be low too. Regardless, that's one metric reviewers often missed, one reason I prefer ASR's.:)
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
I understand that but why is the threshold for IMD? By the way, if THD is low enough, IMD will be most likely be low too. Regardless, that's one metric reviewers often missed, one reason I prefer ASR's.:)
IM causes distinct tones when pure tones are used for the test, as I mentioned- one is the sum of the frequencies and the other is the difference. If these are used without other signals to clutter the test, they're very audible, but when more complex signals are amplified, it's harder to hear, although it can still be noticeable. A lot of people want 20Hz from their subs, but what happens when IM causes frequencies that are lower than the speakers can handle? Well, if these infra-aural frequencies fall below the port tuning frequency, it can be very bad and this is a great argument for using a subsonic filter.

How low the IM is depends on the signal, level of amplification and ability of the amplifier to produce clean amplification of the original signal. Lower is better, but again, it's academic when the number is below some point.
 
AcuDefTechGuy

AcuDefTechGuy

Audioholic Jedi
The amps I want are only 0.02% THD......what should i do??? :p
You should get the ATI AT527NC since it has 0.001% THD+N on ASR. :D

 
Last edited:
M

Movie2099

Audioholic General
You should get the ATI AT527NC since it has 0.001% THD+N on ASR. :D
Hahaha, well you got the company right! I was looking on their website and all they say is "<0.02%"

I need some of those Class D amps!
 
Trell

Trell

Audioholic Spartan
-200dB is impossible- I have heard that core of the Earth S/N is -164dB. Maybe, Gene or someone who has experience in electronic Engineering can confirm or deny this.
Just randomly inserting some zeroes when when writing THD as % and it looks kinda dumb. :D But then as long as THD of solid state electronics is lower than the THD of speakers we should be happy! :D Must be hard to sell underperforming overpriced audio devices when today there are so many excellently performing ones at a reasonable prices. :D:D
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
Just randomly inserting some zeroes when when writing THD as % and it looks kinda dumb. :D But then as long as THD of solid state electronics is lower than the THD of speakers we should be happy! :D Must be hard to sell underperforming overpriced audio devices when today there are so many excellently performing ones at a reasonable prices. :D:D
I started telling customers "Listen to the music, not the specs" a very long time ago.

WRT overpriced- I avoid that stuff like the Plague.
 
highfigh

highfigh

Seriously, I have no life.
I would also be interested to know why KEF seems to be the only popular speaker brand that published their distortions specs except for their top Muon models. Why aren't more manufacturers include distortions in their specs?
If the distortion specs only seem to be OK or bad, who would want to make that well-known? After a third party reveals it and says "These speakers suck", about all the manufacturer can say is "You're right, they suck, somethin' fierce".

As I posted, if my friend gets the position, I'll ask about these.
 

Latest posts

newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top