The Sound of the Machine The Hidden Harmonics behind THD

[WBear2GCR]

Jim,

When an amplifier output stage has no feedback there will be great interaction between the output stage and the speakers.  That means that a given speaker cabinet/system will have a different frequency response and sound with a different feedback-less amplifier. I don't consider this a good situation at all.

You can't simply draw this conclusion.

It is not true in every instance.

In some cases, a variation in frequency response (usually in the LF only) is beneficial, since it results in an increase in low bass, and an improvement in the overall Q of the bass curve...

BUT, you're thinking only of tubes - in the case of a solid state amp, you can make an output stage with extremely low output Z compared to the load, no global feedback, and you won't get this same effect. Of course you can make an output stage with even lower Z using global feedback techniques.

If you happen to biamp or triamp then using a ZFB tube amp on say the tweeter is a very different thing than trying to run a multi-way speaker from a small ZFB tube amp.

Proper application is key to any situation, and improper application usually results in undesireable results. The idea behind the typical feedback amp is to make it universally applicable. Sometimes special purpose designed gear (amplifiers in this instance) are better or best.

                   _-_-bear

[WBear2GCR]

The speaker's distortion is so much greater than the amplfier's.  When you believe you hear IMD products (which is most probably the difference in the "good" and "bad" sounds) this is most likely from the speakers.

Bear, you agree that the critical points in the audio system are the recording session and the speakers.  Even with the most mediocre listening system you can  point out a poor recording session.  We can't do anything about the recording.  Speakers is what the listener need so concentrate on.  This is what Peter is saying.  Yet people are obsessed with the amps!  Go figure.

P.S.  I can't hear the 18th harmonic of 1 kHz, and neither can anyone else.  But all of us can hear the 2nd and 3rd harmonics of 1 kHz very well.

I agree that speakers can and do create IMD products.

It is insufficient to say that you have IMD products without defining their spectra.
This is important!
Non-trivial point.

I had a set of tweeters here that ran from 1500Hz up past 20kHz. They were printed ribbons ("leaf ribbons") that could do 128dB SPL/1M at full power, with LESS than 0.1%THD. Consider that I typically ran them at 28dB down from there... perhaps these tweeters had distortion levels on par with many amps??

Anyhow, it turns out that it's not so much which harmonics, but what the ratio of harmonics are. It turns out it's not so much HOW MUCH (amplitude) but what the ratio of harmonics are!! You can hear some combinations of harmonics (typically 9th and down, especially 5,6 &7th) in certain ratios very easily. Whereas the same harmonics in a different ratio but much higher in amplitude are almost indetectable by the ear.

This is why absolute levels of IM or THD are very poor indicators or predictors of audibility of artifacts.

As I said go search on the "Geddes Metric" - his work follows some British research from the 70s and arrives at the same conclusions.

It explains the whole issue very neatly, imho.

                     _-_-bear

[W1VD]

"The two terms, known as the sum and difference components,
cause the unpleasant sound from an overloaded amplifier. The process
by which they are introduced is known as intermodulation, the two input
signals acting as the carrier and modulating signal, to produce sum and
difference terms which are equivalent to upper and lower sidebands res-
pectively in amplitude modulation.
As an example, suppose two input signals are applied to an amplifier,
the frequencies being 256 cfs and 320 c/s. These frequencies are those of
middle C and E and blend harmoniously to give a chord known as a major
third. Due to non-linearity in the amplifier, the output contains components
with frequencies of 64 cfs (the difference term) and 576 c/s (the addition
term]. The difference frequency is that of C two octaves below the input
signal and harmonises with 256 c/s and 320 cfs, but the addition frequency
is that of D one octave above the input signal and produces a discord with
both 256 cfs and 320 cfs."

<2 cents>

That's just the sum and difference. How about the 'real' IMD

3rd order:  192 & 384 Hz

5th order:  128 & 448 Hz

7th order:   64 & 512 Hz

Unless you're running a serious class A amp these are going to be an issue.

Always been curious if anyone has run double blind tests with different levels of 3rd, 5th, 7th etc order IMD to see what level is noticeable. Bet 3rd order products at 1% (-40 dB) stand out like a sore thumb to someone with good ears...

</2 cents>

[Tom WA3KLR]

Bear I think you have settled everything.  No need for anyone to do any more evaluations.  Just go for the highest MSRP.

[Steve - WB3HUZ]

Anyone who runs their amp into clipping doesn't know what they are doing. No reading required to know that. Obfuscation and double talk doesn't get it done. Sorry.

Of course we're talking about unclipped amps. But the so called expert tried to bring an overloaded amp into the equation. This is typical snake oil salesmen talk. It's garbage and should be pointed out as such.

Steve,

What in the world are you talkin' about??

You need to read up some on the more recent thinking, engineering and testing on these things. Find and read what Dr. Earl Geddes says.

The way an amp clips or even a modulator is indeed important, but that's not the main discussion, since the main discussion revolves around unclipped amplifiers.

However, if you consider that many CDs are recorded with peaks that are 20dB up from the average level, and then you go to play back said CD on your home system at say 90dB SPL at the listening position how much headroom do you need?

Assume to simplify the matter that you manage to get 90dB SPL at 1 watt because you listen very close to ur speakers (yeah, right    ) then we need to reproduce unclipped 20dB up from there. That means we need 100 watts. But if it is a real world speaker and it takes say 4 watts to get to that level, then you need 400 watts, and if it is a typical high-end speaker with only mid 80s reference sensitivity, or you like to play at a slightly higher average level, like 93dB SPL average, you then need to double all those powers - you gotz a 200 watt amp in ur system? You gotz a 1000watt amp? Whatcha got?

I'll bet dollars to donuts that most amps in most systems are clipped far more often than people would like to believe - like all the time.

What's the spectra of distortion of ur amp, no matter what the power rating, as the power gets within 6dB of clipping, how about 3dB of clipping? Betcha that you start to see much larger differences between amps when you look at them this way, and you can hear the same differences just as well...

If one doesn't care, doesn't listen that critically, doesn't notice any differences, that's fine, but there are measureably, objectively, scientifically, and audibly present differences.

                           _-_-bear

                

[WBear2GCR]

Tom & Steve,

I really don't get it.

MSRP has nothing to do with anything I said.
Sorry you don't like to see what prices are for high-end gear, I don't set the prices. We can talk prices if you like, in terms of running a business. But that has nothing to do with sound. If you want lower cost high performance gear you can build ur own.

If neither of you two gentlemen have bothered to take a few minutes and search out what Dr. Geddes' research says, then there's not much more I can say.

Most amps that are in use in the home environment run into clipping all the time.
That's a simple fact. The way an amp sound when clipped plays a role in practical home use. EVEN IF you drop the issue of clipping, you still have to consider the distortion of an amp when driven to somewhere near clipping, like -3dB of clipping. Look at any amp's distortion curve (with a few minor exceptions) and the distortion goes up by maybe two orders of magnitude when dealing with HF?

So, let's not call any of this "garbage" or "snake oil salesman" talk.

Come back with facts please, not name calling.
It's innapropriate and uncalled for.

Thank you.

[ka3zlr]

All I ever wanted to know is how does Bose do it in such a Small Box...the Big Room zound...

[Tom WA3KLR]

My definition - The listening system ideal audio amplifiers – a perfectly linear amplifier with zero output impedance and flat response from 20 Hz to 20 kHz.  Who wants to make an argument against this?

Any and all audio amplifier designs that that come very close to this ideal will all sound the same.  Peter Aczel is saying that “today’s amps” do.  I’m not sure if he is excluding all tube amps in ”today’s amps” or not.  And just because someone puts out an amplifier doesn’t means it is not a poor implementation and can those units can be excluded also; all amplifiers existing today cannot be put into the “today’s amps” grouping.

Any amplifier that does not sound like the rest of the pack is deficient.  It’s that simple.

Jim said “He also makes the case where a solid state Crown DC-300 will look almost perfect on the scope, but will be almost un-listenable with music as the source when the amp is driving quality speakers. Those same speakers and music will sound much better with an old Dynakit ST-70 stereo tube amplifier. The article makes a good case for why this is. This is not a case of SS versus tubes, but more a case against high gain and huge amounts of global negative feedback.” 

From your BBC reference Jim, page 35 - “To obtain low distortion and a very low effective value of output impedance, both desirable for high quality reproduction, feedback must be applied over at least two stages of amplification.”

So this is exactly a case for global negative feedback.

Jim, I found nothing in the BBC pdf in support of deriding the use of negative feedback.
- - -  -
I said “When an amplifier output stage has no feedback there will be great interaction between the output stage and the speakers.  That means that a given speaker cabinet/system will have a different frequency response and sound with a different feedback-less amplifier. I don't consider this a good situation at all.”

Bear said  “You can't simply draw this conclusion.”
Yes I can.  It is quite simple and fundamental.  Passive circuits at work here Bear.  I’ve measured it too.  With no feedback at the output to lower the output impedance to around 100 milliOhms, it is probably in the neighborhood of 4 Ohms.  The speaker at the low frequency end of the spectrum will go though a series resonance.  Only with a very low output impedance will the frequency response be guaranteed to have proper flatness and roll-off.

”It is not true in every instance.”
There is a chance that one’s hi-Z output amp and speaker happen to work together o.k  Otherwise, it is possible to select a speaker and amplifier combination that does not exhibit this.  But you have to have a number of drivers on hand and have a noise generator and acoustic spectrum analyzer or sine generator and SPL meter on hand and find a “good” combination.  How many home owners can do this?  So they probably have a screwed up bass response otherwise, if they own a hi-Z output amp.  Wouldn’t it be better to be able to put an amp and speaker together and know that they work together and produce a flat response.  This is called good system engineering.  Most people own low-Z output solid-state amps today so they don’t have this problem that the no-FB output amp owners do.

”In some cases, a variation in frequency response (usually in the LF only) is beneficial, since it results in an increase in low bass, and an improvement in the overall Q of the bass curve...”
Why people would want to deliberately color the frequency response is beyond me.  Raising the Q of bass curve is bad to me, I can’t think of any good argument for doing this, Bear.

”BUT, you're thinking only of tubes - in the case of a solid state amp, you can make an output stage with extremely low output Z compared to the load, no global feedback, and you won't get this same effect. Of course you can make an output stage with even lower Z using global feedback techniques.”
No I’m not thinking only of tubes amps.  Negative feedback lowering output impedance applies to tube and transistor amps both, can be done to both.  You can make a transformer-coupled output amp. lower impedance by reducing the secondary turns; with no feedback.  This is more in the domain of tube amps today; but you will loose a lot of output power capability though.

Bear, do you have any of the following test equipment:
a.  THD meter
b. noise generator
c. acoustic measurement meter – audio spectrum analyzer or SPL meter
d. pc-based audio analyzer software for a. and c. above.

- - - - - -
Shane said, referring to Carver “Of course, he did a double blind comparison of the audio phools with 5 or 6 amplifiers.  After modifying the neg feedback and output Z, the audio phools couldn't figure out which was the t00bed amplifiers and which where the SS.

Most thought all where t00bed, when in fact, only one was.

He was poo pooed as an audio phool, as well, but who knows.”

Right on Shane.  Since people said they liked the sound of tube amps, Bob did the study to define the difference in tube amp transfer function to make a solid state amp sound like a tube amp (marketing and sales for profit).  I think it was a big hit at the time.  Of course if the tube amp sounds different than the solid state amp, that means there is a difference. And they both can’t be the ideal linear amplifier, or they would sound the same; they both can’t be right (a common HUZ-ism here)!

- - - - -
It’s interesting to see that the audio community does not resort to the 2-tone IMD measurements when analyzing audio amplifiers (which are or should be called because of the application demand – "ultra-linear” amplifiers).  THD measurements are easier to make, and useful, but 2-tone IMD measurements are what is needed also for meaningful and critical evaluation of amplifier non-linearity.  Why this technique isn’t more common in the audio world is baffling.  Audio isn’t magic, some metrics are missing.
- - - -
Jim,

 Your first post with the study on various amplifier configurations and the harmonics is a good and meticulous study.  The researcher was looking for an explanation for system listening differences based on different amplifiers that appear the same.  Interesting to note is that the variations at the 10th and 11th harmonics, is 10 to 15 db.  With this being 10 to 11 kHz and in the region of 100 dB to 120 dB down, I am skeptical that this is the explanation for differences heard.  This study probably was done with a resistive load and recent analyzers.  I doubt that home listening systems can realize more than 80 dB dynamic range, keeping peaks below clipping.  The high order harmonic products would therefore be below the noise floor, regardless of the amplifier.  This is why I still think that listening differences are based on speakers rather than the amplifiers.  There’s plenty of possibility there for the drivers and cross-over components to generate IMD and harmonics from what is being fed to them.

Now if the researcher can re-do the tests based on IMD testing and show significant amplifier IMD performance differences that correlate with the listening system results then he is on to something about specific amplifiers being the problem.  And the problem would be demonstrated amplifier nonlinearity.
- - - - - - -
Bear, I haven't read Dr. Geddes paper yet, but I will.  I don't think you posted the link, otherwise I would have read it by now. 

My home amp in the last 5 years doesn't get driven into clipping, historical fact.  I don't listen to hard rock either.  If I drove my amp into clipping in routine use Bear I would say that I was either deaf beforehand or would be deaf by now.  Perhaps you and your friends should get hearing tests.

[WD5JKO]

Tom,

   Thanks for putting all this together. I'm packing for a trip to Boston, so I'll review the details later. The first thing you say struck me as not universally true and I stopped reading there:

"My definition - The listening system ideal audio amplifiers – a perfectly linear amplifier with zero output impedance and flat response from 20 Hz to 20 kHz.  Who wants to make an argument against this?"

  In order to get 20-20K at low THD out of many "low end amps" you need lots of global NFB to cover up all the shortcuts taken. If it is SS, maybe as much as 40 db NFB. With this much NFB the bandwith (open loop) needs to be much wider to maintain stability. There are consequences of doing this that compromise other spec's not stated like IMD and TIMD. Also the damping factor has a range that it needs to stay above, but not necessarily too high either. A zero source impedance amp is not always what you want. Some speakers might perform better if Rs  is ahigher. You do want some (not a lot) interaction between the woofer /enclosure resonance and the amp. With your "perfect" amp there will not be any. The low end response will be sterile.

Look at sheet 10 downward and zero in on DF:
http://www.turneraudio.com.au/monobloc845se55.html

The amp described has a Rs of 0.5 ohms with only 8 db of Global NFB.

More later,
Jim
WD5JKO

 

[WBear2GCR]

My definition - The listening system ideal audio amplifiers – a perfectly linear amplifier with zero output impedance and flat response from 20 Hz to 20 kHz.  Who wants to make an argument against this?

Insufficient bandwidth.
You'll have problems if you don't have response above 20khz.

Any and all audio amplifier designs that that come very close to this ideal will all sound the same.  Peter Aczel is saying that “today’s amps” do.  I’m not sure if he is excluding all tube amps in ”today’s amps” or not.  And just because someone puts out an amplifier doesn’t means it is not a poor implementation and can those units can be excluded also; all amplifiers existing today cannot be put into the “today’s amps” grouping.

Peter Aczel is wrong in real practice.
Read what Dr. Earl Geddes says and get off this bandwagon.

Any amplifier that does not sound like the rest of the pack is deficient.  It’s that simple.

Incorrect again.
Many amps that measure technically very well have issues that make them sound like other amps with the same issues, per Geddes' research (he merely has reduced to scientific measurement what others have long been saying). Similarly many amps that measure technically very well sound different than the first group I mentioned.

What will confound you, not having read Geddes' JAES peer reviewed papers, or his books, is that some amps that do not measure particularly well also sound the same as those in one of the two above groups of technically good measuring amps!

You're working with old information that has been rendered obsolete and incomplete. The old idea that any "good" amp that measures less than some threshold of THD (like 0.001%?) is indistinguishable from another. This old idea is false.

Jim said “He also makes the case where a solid state Crown DC-300 will look almost perfect on the scope, but will be almost un-listenable with music as the source when the amp is driving quality speakers. Those same speakers and music will sound much better with an old Dynakit ST-70 stereo tube amplifier. The article makes a good case for why this is. This is not a case of SS versus tubes, but more a case against high gain and huge amounts of global negative feedback.” 

From your BBC reference Jim, page 35 - “To obtain low distortion and a very low effective value of output impedance, both desirable for high quality reproduction, feedback must be applied over at least two stages of amplification.”

So this is exactly a case for global negative feedback.

It is possible to get low output Z without global negative feedback.
It is also possible to get low distortion without global negative feedback.
Feedback has both benefits and deficits and its overall benefit depends on the specific implementation - it is neither good nor bad, it is a tool.

Jim, I found nothing in the BBC pdf in support of deriding the use of negative feedback.
- - -  -
I said “When an amplifier output stage has no feedback there will be great interaction between the output stage and the speakers.  That means that a given speaker cabinet/system will have a different frequency response and sound with a different feedback-less amplifier. I don't consider this a good situation at all.”

Bear said  “You can't simply draw this conclusion.”
Yes I can.  It is quite simple and fundamental.  Passive circuits at work here Bear.  I’ve measured it too.  With no feedback at the output to lower the output impedance to around 100 milliOhms, it is probably in the neighborhood of 4 Ohms.  The speaker at the low frequency end of the spectrum will go though a series resonance.  Only with a very low output impedance will the frequency response be guaranteed to have proper flatness and roll-off.

I have no idea what output stages you are familiar with.
The "average" SS amplifier that has say 2 or 4 output devices per channel will have a high output Z without feedback, agreed. But this is not the case where there are multiple output devices in parallel. Say something like 24 or 48? There are quite a few amps like this...

Otoh, the Crown MacroReference amp uses both multiple output pairs and a double loop of feedback which yields a crazy DF of something like >60,000!!

There are some speakers that happen to be "overdamped" - meaning their Qt = <0.5 for example. In this case a high DF (low output Z) amp will appear to have anemic bass, whereas an amp with a low  DF will have actually greater bass extension and perhaps even a "flatter" response.


Let's keep one thing in mind, very very very very few speakers have a "flat" response at all, and even fewer have a "flat response" in the room, and fewer still have a "flat" response in the room and off axis. So, what exactly do you mean by what you've said WRT "flat response" or "proper roll-off"? Please don't tell me that the speakers that are like this are not properly designed...

"It's not true in every instance"
There is a chance that one’s hi-Z output amp and speaker happen to work together o.k  Otherwise, it is possible to select a speaker and amplifier combination that does not exhibit this.  But you have to have a number of drivers on hand and have a noise generator and acoustic spectrum analyzer or sine generator and SPL meter on hand and find a “good” combination.  How many home owners can do this?  So they probably have a screwed up bass response otherwise, if they own a hi-Z output amp.  Wouldn’t it be better to be able to put an amp and speaker together and know that they work together and produce a flat response.  This is called good system engineering.  Most people own low-Z output solid-state amps today so they don’t have this problem that the no-FB output amp owners do.

Actually today anyone who wants to have excellent test gear and owns a computer can do this and more. But there are no combinations of amps and speakers that anyone can put together and know a prori that the results are going to be just fine. Sure, amps with identical or nearly identical DF will produce a frequency response curve that is nominally identical, but so will cars with similar power to weight ratios produce similar 0-60 times. It doesn't really tell us much.

Similarly, anyone who has a ZFB power amp and doesn't know how or when to apply it, is no different than someone who buys anything else that they don't know how to apply it. Audio is no different, you get the proper item for the application. The guy with the Bose Wave Radio is on the opposite side of the audio world from a person who builds his own or buys a ZFB tube amp... should ZFB amps be illegal??

....

  variation in frequency response (usually in the LF only) is beneficial, since it results in an increase in low bass, and an improvement in the overall Q of the bass curve...”
Why people would want to deliberately color the frequency response is beyond me.  Raising the Q of bass curve is bad to me, I can’t think of any good argument for doing this, Bear.

Quite the contray, you're making a false assumption.
The assumption is that the only way to make an accurate or even good sounding system is one way - with a high FB low output Z solid state amp. Again there are speakers that demand low output Z. Bass horns are one possible system that this works for. Another is the so-called "EBS" alignment. A fellow named Bob Cordell champions this idea. Do you recognize the name? Check his website. I'm not saying that it is good in every case. Raising the Qt of a speaker system depends entirely on what it IS to begin with!

”BUT, you're thinking only of tubes - in the case of a solid state amp, you can make an output stage with extremely low output Z compared to the load, no global feedback, and you won't get this same effect. Of course you can make an output stage with even lower Z using global feedback techniques.”
No I’m not thinking only of tubes amps.  Negative feedback lowering output impedance applies to tube and transistor amps both, can be done to both.  You can make a transformer-coupled output amp. lower impedance by reducing the secondary turns; with no feedback.  This is more in the domain of tube amps today; but you will loose a lot of output power capability though.

I'm saying the opposite. You can get a low output Z without adding feedback.
I'm not talking about tube amps here either, I'm talking about SS amps.

Then there is Hawksford's feedforward correction...

Bear, do you have any of the following test equipment:
a.  THD meter
b. noise generator
c. acoustic measurement meter – audio spectrum analyzer or SPL meter
d. pc-based audio analyzer software for a. and c. above.

Does it matter?
I don't have a particle accelerator, but I can read and know the physics?
But yes, I have measurement capability. And what I don't have I can get my hands on, need be.
What are you getting at?

- - - - - -
It’s interesting to see that the audio community does not resort to the 2-tone IMD measurements when analyzing audio amplifiers (which are or should be called because of the application demand – "ultra-linear” amplifiers).  THD measurements are easier to make, and useful, but 2-tone IMD measurements are what is needed also for meaningful and critical evaluation of amplifier non-linearity.  Why this technique isn’t more common in the audio world is baffling.  Audio isn’t magic, some metrics are missing.

Say what??
Huh??

Look up CCIR...
Report back.

- - - -
Jim,

 Your first post with the study on various amplifier configurations and the harmonics is a good and meticulous study.  The researcher was looking for an explanation for system listening differences based on different amplifiers that appear the same.  Interesting to note is that the variations at the 10th and 11th harmonics, is 10 to 15 db.  With this being 10 to 11 kHz and in the region of 100 dB to 120 dB down, I am skeptical that this is the explanation for differences heard.  This study probably was done with a resistive load and recent analyzers.  I doubt that home listening systems can realize more than 80 dB dynamic range, keeping peaks below clipping.  The high order harmonic products would therefore be below the noise floor, regardless of the amplifier.  This is why I still think that listening differences are based on speakers rather than the amplifiers.  There’s plenty of possibility there for the drivers and cross-over components to generate IMD and harmonics from what is being fed to them.

Now if the researcher can re-do the tests based on IMD testing and show significant amplifier IMD performance differences that correlate with the listening system results then he is on to something about specific amplifiers being the problem.  And the problem would be demonstrated amplifier nonlinearity.
- - - - - - -
Bear, I haven't read Dr. Geddes paper yet, but I will.  I don't think you posted the link, otherwise I would have read it by now. 

My home amp in the last 5 years doesn't get driven into clipping, historical fact.  I don't listen to hard rock either.  If I drove my amp into clipping in routine use Bear I would say that I was either deaf beforehand or would be deaf by now.  Perhaps you and your friends should get hearing tests.

How many watts is your amp?
How far from your speakers do you sit?
What is the typical listening level in SPL at ur listening position?
What are the legit specs for your speakers (sensitivity, impedance, etc.)
Let's figure out if you ever clip ur amp or not...

           _-_-bear

[Tom WA3KLR]

Bear,

Are these the Dr. Geddes papers you are referring to?  See the 2 pdf's below.

I will make amplitude measurements on my listening system today if I can and report back.

[Tom WA3KLR]

My definition - The listening system ideal audio amplifiers – a perfectly linear amplifier with zero output impedance and flat response from 20 Hz to 20 kHz.  Who wants to make an argument against this?

Insufficient bandwidth.
You'll have problems if you don't have response above 20khz.

I never made any limitations on bandwidth.  I stated a minimum (perhaps implied) frequency response only.

[Tom WA3KLR]

Bear said -  "You're working with old information that has been rendered obsolete and incomplete. The old idea that any "good" amp that measures less than some threshold of THD (like 0.001%?) is indistinguishable from another. This old idea is false."

I never made such a claim.  My stance is that the "low" THD numbers do not tell the whole story.  One needs to advance to IMD measurements.  The human hearing is much more sensitive to IMD than harmonic distortion and the amps produce noticeable IMD at "low" THD levels.  * But the situation is that most people's hearing is poor and all over the lot.  They can't tell the difference in the "good" amps.

* What I really mean to say here is that most people do not or cannot judge sound quality.

[Steve - WB3HUZ]

But you must first drive your amp into distortion, then tell everyone it sounds better than other amps. 

[WBear2GCR]

Bear said -  "You're working with old information that has been rendered obsolete and incomplete. The old idea that any "good" amp that measures less than some threshold of THD (like 0.001%?) is indistinguishable from another. This old idea is false."

I never made such a claim.  My stance is that the "low" THD numbers do not tell the whole story.  One needs to advance to IMD measurements.  The human hearing is much more sensitive to IMD than harmonic distortion and the amps produce noticeable IMD at "low" THD levels.  * But the situation is that most people's hearing is poor and all over the lot.  They can't tell the difference in the "good" amps.

* What I really mean to say here is that most people do not or cannot judge sound quality.

CCIR is just one standard used to do IM tests, iirc it is 19kHz & 20 kHz tones... fyi.
No one who designs a "serious" amplifier uses mere THD anymore - that's for the marketing people for mid-fi companies, imho.

I think those are probably the Geddes papers, if they are from his site, and they include the term "Geddes Metric"... I didn't look yet.

I agree that most people can not tell the difference.
That doesn't mean that the difference isn't there, or isn't real, or isn't audible.

Thus the Bose Wave Radio is said to be all that you need... ha ha.

However, I believe that if an average person, like you or me, got to hear two amps on a system that is sufficiently clean (whatever that threshold is...) you could hear such a difference quite clearly, without ambiguity.

Personally, I think that it is cool that Geddes has brought to light what many people have been saying for a long time. It reveals something very interesting about sound and the human ear. It also shows why some tube amps can sound much better than some solid state amps of seemingly lower distortion!!

         

                            _-_-bear

[WBear2GCR]

But you must first drive your amp into distortion, then tell everyone it sounds better than other amps. 

Ummm... no.

Obviously, once driven into clipping, the distortion levels increase momentarily. This is usually more audible than whatever spectra of distortion the amp has the rest of the time. Unless the spectra of distortion happens to conform to the Geddes Metric, and is essentially inaudible the rest of the time... in which case the amp only begins to sound noticeably distressed when clipped on peaks...

Then it depends how much other masking effects you are getting from your speakers (a big source of distortions) - but again much depends on the spectra of distortion not the absolute level of distortion!, and the rest of the signal chain.

For most average folks with mid level systems, especially those that are not set up optimally (like speakers kind of randomly placed for example) these things don't make a whole lot of difference.

It's once you begin on the path to optimization and getting better or higher performance gear that you suddenly are confronted by the stark realization that things are not as simple as they seem when you started.

It's akin to watching an old 1960s color tv program broadcast using 1960s technology and then moving up to (or beyond) today's HD, or maybe a high res computer monitor and suddenly you can see things that formerly were glossed over and were unimportant. You can see them and they're there and you notice them, even if you would prefer to not.

Good example, there is tape on the floor on The Price Is Right, you couldn't see it at all in the old days... 

Or maybe it is the difference between SSB and AM?
Do you notice the difference in sound quality?
Sure that's a gross difference, but in sound reproduction there are differences of several plateau levels, it's not just a binary thing...

                     _-_-bear

[Steve - WB3HUZ]

Umm, yes. Let me take you back in the thread.

From: http://www.pmillett.com/Books/bbc_feedback.pdf

"The two terms, known as the sum and difference components,
cause the unpleasant sound from an overloaded amplifier. The process
by which they are introduced is known as intermodulation, the two input
signals acting as the carrier and modulating signal, to produce sum and
difference terms which are equivalent to upper and lower sidebands res-
pectively in amplitude modulation.

I added the bold on overloaded. You defended the above statement with great vigor. Now you say no. Which is it?

[WBear2GCR]

Sorry, Steve, I'm totally confused...

The books on Pete Millet's site are useful, but generally old books that have no copyright. In some cases the information there has been eclipsed.

There are various issues at play when an amp is clipped (overloaded?), especially when a NFB loop is involved. I don't know what ur pointing at. Clipped amps produce distortion. Unclipped amps produce distortion.

The two questions are:
- how much
- what spectra
- oh and a third, is it audible?

This is for the unclipped and clipped states... is there something else I am missing about it?

                 _-_-bear

[Tom WA3KLR]

Bear,

I just looked at my stereo system which is a slighty modified Harman Kardon 630 receiver and modified Utah HS4-B 3-way bass-reflex speakers. 
As I recall, the amps are rated at 30 Watts per channel.
I sit 13 feet from the speakers.

The audio meter’s SPL function is set to:
SPL Weighting = FLAT
SPL Response = FAST

The amp, no load puts out 60 Volt peak-to-peak.

I ran the receiver in mono, listening to a local FM station, only one speaker channel on.  This eliminates any question of channel imbalance.
Adjust for 85 – 90 dB SPL reading.  I wouldn't care to listen any louder than this.  I believe you mentioned the same SPL.
I saw maximum of 10 Volts peak-to-peak.
This is 16 dB headroom.  I assume running stereo for the same SPL would produce a little more headroom.

Next, I ran the receiver from the CD player.
Adjust for 85 – 90 dB SPL reading (perhaps a couple dB louder than FM actually was).
I saw a maximum of 20 Volts peak-to-peak.
This is 10 dB headroom. 

- - - - - -

There's always been tape on the studio floors.  It was visible.  It may be more visible today due to the solid state technology.

[Tom WA3KLR]

Looking at the Audio Engineering Society website (aes.org) a bit I spotted these 2 items:

AES Journal Forum

1. A New Method for Measuring Distortion Using a Multitone Stimulus and Noncoherence

Read comments here:
http://www.aes.org/journal/online/comment/?ID=14381


2. A paper that must be purchased but just the description is interesting.  Another listening study by Geddes, this one 3 years more recent than the Geddes Metric paper.

AES E-Library:
Audibility of Linear Distortion with Variations in Sound Pressure Level and Group Delay by Geddes, Earl R.; Lee, Lidia W.

Recent psychoacoustic studies of nonlinear distortion have yielded some new insights into what audible problems in loudspeaker might be related to. This paper will show the results of recent subjective tests which extend the work of various previous works to show that sound level significantly affects the perception of linear distortion in audio systems. This means that the hearing system itself is nonlinear and what has been thought of as being nonlinear distortion in the audio system may actually be a nonlinear perception directly in the receiver itself.
Paper Number:   6888    AES Convention:   121 (October 2006)
Authors:   Geddes, Earl R.; Lee, Lidia W.
Affiliations:   Eastern Michigan University; GedLee LLC    (See document for exact affiliation information.)

AES Store: Purchase Full Paper
$5.00 AES Member      $20.00 Non-Member

The "receiver" being the human listener.

[Tom WA3KLR]

A paper on perceived distortion from Geddes' website by Keith Howard below:

[WBear2GCR]

Bear,

I just looked at my stereo system which is a slighty modified Harman Kardon 630 receiver and modified Utah HS4-B 3-way bass-reflex speakers. 
As I recall, the amps are rated at 30 Watts per channel.
I sit 13 feet from the speakers.

The audio meter’s SPL function is set to:
SPL Weighting = FLAT
SPL Response = FAST

The amp, no load puts out 60 Volt peak-to-peak.

I ran the receiver in mono, listening to a local FM station, only one speaker channel on.  This eliminates any question of channel imbalance.
Adjust for 85 – 90 dB SPL reading.  I wouldn't care to listen any louder than this.  I believe you mentioned the same SPL.
I saw maximum of 10 Volts peak-to-peak.
This is 16 dB headroom.  I assume running stereo for the same SPL would produce a little more headroom.

Next, I ran the receiver from the CD player.
Adjust for 85 – 90 dB SPL reading (perhaps a couple dB louder than FM actually was).
I saw a maximum of 20 Volts peak-to-peak.
This is 10 dB headroom. 

- - - - - -

There's always been tape on the studio floors.  It was visible.  It may be more visible today due to the solid state technology.

20 volts peak to peak? take 1/2 and then take .707 of that, square it and divide by eight. That gives us 6 watts into 8 ohms.  Take the 60v p-p take half and then .707, square and divide by 8 and you get 56 watts. So, must be a 60 watt per channel amp.

Yeah you have an apparent ~10db (not 16dB) of headroom, assuming there are no peaks over the average level, which is a poor assumption using a CD that is not compressed up the yinyang, which apparently is not the sort of music ur listening to?

You don't want the flat setting on the SPL meter, you want the one that rolls off the lows... but we can still work with these readings for now... also it is best to use the slow setting to average the level, we want the average level...

You might want to use pseudo random noise as the level setting source, FM interstation is almost right, but it suffers (as you probably know) from the de-emphasis boost built into the receiver... I think you can probably generate a .wav file using shareware that will do that trick, if you wanted to. Heck, if ur not in a hurry I can email you a wav file or mp3 that will work...

You can see the problem though, if you pushed that CD playback to just 3 dB more (two just noticeable increases) you'd have dropped the headroom down to only 6dB.
That's a guarantee that peaks will be into the amps clipping...

Those old Utahs are likely up in the high end of the 90-100dB sensitivity range, above 95 almost certainly, maybe as high as 100dB/1w/1m...btw.

Which, interestingly enough makes them a candidate for tube amplification - which if it a modern tube amp, using modern power supplies and modern resistors and caps might be very very surprising to you! I suggest that you try that, not some old tube amp, some time?

The tape on the floors is more visible due to higher resolution, regardless of the technology that produces higher resolution. It was always there, but it was not very noticeable at all - that is the point of my analogy.

                 _-_-bear

[WBear2GCR]

Thanks for the link(s) Tom!

I recall reading the last link when it came out in Multimedia Mfr mag - an Ed Dell publication...

I hope that what I've been saying, and really what Geddes and Shorter before him are/were makes sense now?

                   _-_-bear

[WD5JKO]

Hi All,

   I am up near Boston on business. Boy the WX sure stinks!

It seems that the division among us posting on this thread has narrowed a bit. I started this, and to be honest, I have learned quite a bit considering all the various point of views. Understanding good sound is like trying to understand what motivates my XYL. Science doesn't always get you there. ;-)

 One point worth making is that IM can be high WITHOUT OVERLOAD. All we need is a non linear transfer function on a gain stage. This is kind of represented with my old 8417 modulator when I had the idle current too low. I had plenty of power, and it looked OK on the scope. The audio reports were terrible and it sounded like a carbon Mic that needed to be whacked. After raising the idle current it sounded pretty decent. This was NOT OVERLOAD.

   On a slightly different twist, some years ago Joe Curico published a paper, "The Audibility of Power supplies". see link:


http://www.curcioaudio.com/Audibility_of_PwrSupply.htm

 
    I read this a few times over several years and said to myself, yea yea yea, BS, snake oil, since his business is selling kits.
Well I've had a Sherwood S8000 tube amp / FM tuner for 30 years (30w / channel amp). About 5 years ago it showed serious fatigue as some tubes were aging, caps were drying out, and some of the wiring had insulation falling off. The output tubes are 7868's (Novar version of the Octal 7591), and are the same Sylvania's since 1981 which are going strong to this day.

   This amp has been the center of my main sound system. For several months in 2004 I took the amp out of service and did some cleaning up and updating. I used some of Joe Curcio's ideas using my own designs. I beefed up the plate voltage (450V) with 200 mfd, and electronicially regulated the 7868 screen voltage, audio driver, FM tuner, preamp filaments, and 7868 grid 1 bias settings.

   The important point here is that the basic Sherwood tube amplifier circuitry remained stock except for the ability to set the bias on each 7868 output tube. My changes concentrated on power distribution. A summary of my changes are at this link:

http://pages.prodigy.net/jcandela/Sherwood_S8000/HV_REG/S8000_Mods2.htm

   So did any of those changes make a difference? Well at conservative volume levels, I'd say no. If I crank, especially with deep bass, I can say absolutely yes. This amp is now playing HD DVD's, TV, CD's, Play station 2, XBox 360, and Wi. When the kids play games the house shakes. The other night we played "King Kong" on HD DVD, and the noises from the big Ape (mostly under 50 hz) were resonating things around the house. That Sherwood low end was never like this when it was stock and much newer.

    Back to the point. A low end amp is likely to skimp on the power supply whereas a high end amp might have better magnetics (pwr transformer) and bigger electrolytics. Also the low end amp will have less heatsinking (SS amps) and hope for providing high peak power when the average is very low. I don't think the "Snake Oil" salesmen are as prevalent as some think selling $1000 amps that are no better than $100 amps. Instead I think they have moved to selling $100 HDMI cables, and other high end cables for HDTV installations.

Good info on this thread from all,
73
Jim
WD5JKO

[Tom WA3KLR]

Jim,

No doubt power supplies or lack of regulated supplies in the amplifier is a source of many problems and complaints, transient phenomenon.  For many years when I hear of the line cords, penny in fusebox etc. improvements I said to myself well maybe there is power supply problems.

My receiver/amp here is a Harmon Kardon 630 which is 38 years old now.  Many years ago after noticing the significant hum level (saw-toothed also) at the speaker terminals, I investigated and took out the simple series pass transistor/zener diode regulator for the +22 Volt supply and replaced it with a LM723 circuit I think it was.  Now pure dc.  Also the main speaker amp supplies had bad ground loops involving the filter capacitor currents.  I put all of the raw filter stuff on isolated buss wire which came to a single point ground on the chassis.  I moved the speaker ground return points.   Afterwards the hum was in the noise level and sinusoidal.

How many high end amps use a linear regulated supply ahead of the output amps?  Does anyone at all?