Tlf,
Masking. If you add some low order harmonic distortion, H2/H3, it will supplement the intrinsic distortion in the ear and dupe it into missing any low amplitude high order harmonics which are present.
Very low levels of THD, as Nelson alludes to, don't necessarily sound good, because they have been achieved by use of feedback, which creates a cascade of higher order distortions, skewing the spectrum towards the higher order, albeit at very low amplitudes. These are highly audible as long as the fundamental puts them within the audio range.
The H2 will make an amp sound full and rich, H3 will sharpen it up, giving the perception of more detail. Both will mask the presence of higher order distortions.
This is why reducing the feedback factor can actually sound better despite a higher measured THD.
Like Elso, I design my amps by ear, with scrupulous input from the Function Generator, CRO and DMM. I push for very low high order distortion, and try to induce a little low order for its masking and richening effects. The acid test is to listen, and listen hard.
This all relates to the analysis of how sound is perceived in the human animal. It's been done before, but this work is deliberately obfuscated by high end marketing hype. It would be interesting to measure the THD of some of the net codecs, such as MP3....... I'm confident it's high, yet it sounds, well, OK considering.
Perception is actually more important than reality in most things, and particularly in marketing.
Cheers,
Hugh
Masking. If you add some low order harmonic distortion, H2/H3, it will supplement the intrinsic distortion in the ear and dupe it into missing any low amplitude high order harmonics which are present.
Very low levels of THD, as Nelson alludes to, don't necessarily sound good, because they have been achieved by use of feedback, which creates a cascade of higher order distortions, skewing the spectrum towards the higher order, albeit at very low amplitudes. These are highly audible as long as the fundamental puts them within the audio range.
The H2 will make an amp sound full and rich, H3 will sharpen it up, giving the perception of more detail. Both will mask the presence of higher order distortions.
This is why reducing the feedback factor can actually sound better despite a higher measured THD.
Like Elso, I design my amps by ear, with scrupulous input from the Function Generator, CRO and DMM. I push for very low high order distortion, and try to induce a little low order for its masking and richening effects. The acid test is to listen, and listen hard.
This all relates to the analysis of how sound is perceived in the human animal. It's been done before, but this work is deliberately obfuscated by high end marketing hype. It would be interesting to measure the THD of some of the net codecs, such as MP3....... I'm confident it's high, yet it sounds, well, OK considering.
Perception is actually more important than reality in most things, and particularly in marketing.
Cheers,
Hugh
I always think the low-order/high-order distortion is coming in one package, I imagine it like a water pillow.
For the same amp, if you press the low-order ones, the high-order ones will rise. But if you don't press the low-order ones, the high-order ones will not show up.
It doesn't matter if the initial balance is at 0.1% or at 0.001%, I think the mechanism is still the same at any initial THD balance, which is determine by the linearity of the (openloop) design.
For the same amp, if you press the low-order ones, the high-order ones will rise. But if you don't press the low-order ones, the high-order ones will not show up.
It doesn't matter if the initial balance is at 0.1% or at 0.001%, I think the mechanism is still the same at any initial THD balance, which is determine by the linearity of the (openloop) design.
Feedback discussion is a "never ending story" here 😀
We have some very good experts here.
Maybe this can help understanding what happens, in "easy to understand" reading. Read patent #4,476,442 by Iwamatsu, seen in www.freepatentsonline.com.
In the text, 1st-2nd column, (especially section 50-55 left) it can helps to understand that while the feedback concept is a noble concept, but in reality it's performance will be limited by the nature of magnifying curve of a transistor (which is non-linear). Feedback will loose grip at higher frequencies (usually in higher order harmonics). Unfortunately, our ear is sensitive to these higher order harmonics/dissonant harmonics.
Maybe that's why now NP is forcing to use Lovoltech power Jfet, a "switch", not "for audio" product, even it is used to work only below 4V. 😀 This device has more linear magnifying curve.
We have some very good experts here.
Maybe this can help understanding what happens, in "easy to understand" reading. Read patent #4,476,442 by Iwamatsu, seen in www.freepatentsonline.com.
In the text, 1st-2nd column, (especially section 50-55 left) it can helps to understand that while the feedback concept is a noble concept, but in reality it's performance will be limited by the nature of magnifying curve of a transistor (which is non-linear). Feedback will loose grip at higher frequencies (usually in higher order harmonics). Unfortunately, our ear is sensitive to these higher order harmonics/dissonant harmonics.
Maybe that's why now NP is forcing to use Lovoltech power Jfet, a "switch", not "for audio" product, even it is used to work only below 4V. 😀 This device has more linear magnifying curve.
Obfuscation is part of the game - muddy the waters! Destroy scientific credibility.
All in the name of...
Cheerrs,
Greg
All in the name of...
Cheerrs,
Greg
forr said:
This is intersting. Did you have the opportunity to compare the microphone signal during the recording with the musical signal before A/D conversion?
Or do musical instruments not have high order harmonics? But then, why does a Steinway sound different from a Bösendorfer?
A little confused 😕
Tino
planet10 said:
It is typically 2nd & 3rd that make an amp sound richer. Higher harmonics are grating to the ear.
Here is an example of a spectrum that collapses to a low number, but has a scary amount of high order harmonics.
dave
Very sorry, this is nothing else than an example of a horrible amplifier. There is nothing like a spectrum that collapses to a low number.
AKSA said:
Hugh,
I have to disagree. H2 and H3 applied to more complex music will create disharmonic components, nothing like "full and rich" sound. The result is melting and confusing sound.
PMA said:
Ahh, MikeB's Symasym 5_1 amplifier is "horrible", ok, I learned something new today!
Cheers Michael 😉
AKSA said:
I think there are three separate issues here:
a) the threshold (or audiability) of THD: I suppose there is a level of thd below which no one, even those golden ear types, can reliably hear. I don't personally know what that number is but I suppose 0.1 (or 0.01)% is a pretty good guess?
b) the composition of THD: the higher order harmonics vs. low order harmonics, the odd vs. even order harmonics, how much we like some of them and how much we hate the others. I haven't seen much concrete evidence of it but I have heard a lot about people claiming to hate the odd order or high order harmonics. I will take that as a given for now.
c) creation of harmonics by feedback: we are in agreement here that the use of feedback will create more high order haromnics. Two things need to be examined here however:
c.1) what if the use of feedback reduce thd below the audioable levels we talked about earlier? In this case, wouldn't you agree that there is an improvement of sound quality (as the system goes from having audioable distortion to NOT haveing audioable distortion)?
c.2) if the use of feedback isn't sufficient to reduce distortion below the audioable levels, does there exist a trade off such that the sound is better due to a lower distortion but worse composition of the distortion?
There is also a third question: why different amps exhibit different thd profile when feedback is applied? that is, are there amps better suited for the application of feedback than others? if so, what are the characteristics of those amps?
AKSA said:
if so, you would agree that a trade off between lower levels of distortion and worse compsotion of distortion can be benefitial.
AKSA said:
That is a little simplistic. Music is nothing but a bunch of harmonics. That's why a G on a piano sounds different t han a G on a trumpet. So the signal an amp sees has in itself a lot of harmonics to begin with. When that composition of harmonics goes through an amp that has just H2 distortion, the output will have distortion in a twice the frequency spectrum than the original signal has -> higher order harmonics from an amp that generates seemingly just low order distortion.
AKSA said:
and as importantly to figure out if there is actually any preference for those high-end amps, or how important is distortion composition vs. levels.
I suggest forum guys, the ones that do not believe in nothing, that
will be more productive to start to construct things to have a real evaluation.
As those theories are never ending discussions, that really serves only to create less harmony between us.
Music was made for ears...so...construct and evaluate, and them, with more "real" informs, you will have better condition to tell some.
This way, the ones that enter only to produce problems, as do not want to do more than that....those will construct nothing, maybe cannot solder nothing because do not know how to do.
The others, will have strong argument associating theories to real audition.
Symassym is a very good amplifier, despite i believe that low distortion amplifiers are so bad as sinthetic Orange Juice..... reproduction remember a good sound...but do not reproduce a good sound..... Symassym, a exception, really very strange, distortions not so high, and producing good sound.
I am one those that believe that we can feel good with some distortions, so Symassym can be really strange, because scaped from my beliefs...so.... as i use to do...construct and do it by yourself.
Ultima...good.... very good...he deserve a longer one...a longer message...ahahahhaa.
regards,
Carlos
will be more productive to start to construct things to have a real evaluation.
As those theories are never ending discussions, that really serves only to create less harmony between us.
Music was made for ears...so...construct and evaluate, and them, with more "real" informs, you will have better condition to tell some.
This way, the ones that enter only to produce problems, as do not want to do more than that....those will construct nothing, maybe cannot solder nothing because do not know how to do.
The others, will have strong argument associating theories to real audition.
Symassym is a very good amplifier, despite i believe that low distortion amplifiers are so bad as sinthetic Orange Juice..... reproduction remember a good sound...but do not reproduce a good sound..... Symassym, a exception, really very strange, distortions not so high, and producing good sound.
I am one those that believe that we can feel good with some distortions, so Symassym can be really strange, because scaped from my beliefs...so.... as i use to do...construct and do it by yourself.
Ultima...good.... very good...he deserve a longer one...a longer message...ahahahhaa.
regards,
Carlos
There an interesting reading, coming from a person who has sufficient measurement equipment.
With Google, type "John Curl Audio", then in one of the result there is a .pdf of "John Curl Interview".
Hugh and Elso are always aksed to defend their products and explain everything so that it reflects measurments.
Many of the users of their equipment have had amps that measure very well from a scientific standpoint, but they have abandoned that in search of something a little more relaxed and emotionally fulfilling.
Some people prefer sound which is richer, others stop at nothing to get it clean and measure well - then spend more time trying to explain why, rather than listening to the amp.
Kinda reminds me of the scientist who is so caught up in his research , he forgets about his wife.
Is this not really down to human feeling, and some not having a desire to completley understand why something sounds so good?
There can never be a universally accepted conclusion to this topic.
Many of the users of their equipment have had amps that measure very well from a scientific standpoint, but they have abandoned that in search of something a little more relaxed and emotionally fulfilling.
Some people prefer sound which is richer, others stop at nothing to get it clean and measure well - then spend more time trying to explain why, rather than listening to the amp.
Kinda reminds me of the scientist who is so caught up in his research , he forgets about his wife.
Is this not really down to human feeling, and some not having a desire to completley understand why something sounds so good?
There can never be a universally accepted conclusion to this topic.
PMA said:
I should have known better and not post that chart... The peaks
above 10khz are also present without signal/amp, these are introduced
by the measuring, might be the soundcard, fft-window or something
else. The only valuable infos are h2-h7, where h7 is already at the
same level as measured without amp. (loopback-cable,-110db)
Also, i had written in my posting that the right channel has a defect...
The left channel (not well visible in this chart) has 15db less h2.
Okay, will continue building horrible amplifiers, these seem to sound best... 😀
Mike
MikeB said:
The sound card - this is what I would expect. The RightMark itself is not a problem. We are collecting results of different soundcards and I know that the result can be much better than the one shown here.
For the reason you stated it is difficult to evaluate amplifier itself, so there is no much value in this spectrum. I would personally prefer not only higher harmonics to be lowered, but also H2 - H6.
I am attaching a result of another soundcard by RMAA. This was a loopback test as well, and I must say that it is much better to use an external high quality generator.
Attachments
No Michael, your amplifier it is not horrible, i was hearing it rigth now.
Watching a good movie, and only the Symassym was switched on...and to full range operation.
Wonderfull sound Michael...has no conection with awfull amplifiers, the one that told you the amplifier was not good, made a mistake, and this is human...he do not know what he is talking about, as he did not heard Symassym.
So, do not mind, having not hearing it, The one cannot evaluate...graphics do not play music Michael.
In the middle of those guys, one or two are mistaken...there are the majority of good and experienced guys here too...but if they did not listened to Symassym, even those ones are impeached to evaluate... they have only simulations, theories and graphic tools...not enougth.... human evaluation is better..one man, one single man can produce evaluation better than hundreds of papers....because can tell you, how the amplifier sound!....hehe ...papers can't!
Your amplifier is guaranteed, and they are asking the spectrum analisers do guarantee..and they are not talking nothing. ahahahaha!
regards,
Carlos
Watching a good movie, and only the Symassym was switched on...and to full range operation.
Wonderfull sound Michael...has no conection with awfull amplifiers, the one that told you the amplifier was not good, made a mistake, and this is human...he do not know what he is talking about, as he did not heard Symassym.
So, do not mind, having not hearing it, The one cannot evaluate...graphics do not play music Michael.
In the middle of those guys, one or two are mistaken...there are the majority of good and experienced guys here too...but if they did not listened to Symassym, even those ones are impeached to evaluate... they have only simulations, theories and graphic tools...not enougth.... human evaluation is better..one man, one single man can produce evaluation better than hundreds of papers....because can tell you, how the amplifier sound!....hehe ...papers can't!
Your amplifier is guaranteed, and they are asking the spectrum analisers do guarantee..and they are not talking nothing. ahahahaha!
regards,
Carlos
PMA, can you show a spectrum of an amp of yours recorded with that
soundcard ? Can you record at 96khz ? Most of the disturbances
started at ~20khz with my recording.
Can you also show freq-response for this card ?
Or you can mail it to me if you'd like.
Mike
soundcard ? Can you record at 96khz ? Most of the disturbances
started at ~20khz with my recording.
Can you also show freq-response for this card ?
Or you can mail it to me if you'd like.
Mike
Hi Pavel,
I can live with that!! Once, back in 1962, I was wrong....... so I guess it could, with bad luck, happen again!!
However, we should not talk absolutes. How does 0.045% distortion at 20W and 20KHz seem, most of it H2/H3?
These figures obtained with an HP339A.
Would this muddy Mahler?
On another note, Pavel, do you know the input impedance of your PMA/NP circuit? Does it need much grunt to drive it?
Cheers,
Hugh
I can live with that!! Once, back in 1962, I was wrong....... so I guess it could, with bad luck, happen again!!
However, we should not talk absolutes. How does 0.045% distortion at 20W and 20KHz seem, most of it H2/H3?
These figures obtained with an HP339A.
Would this muddy Mahler?
On another note, Pavel, do you know the input impedance of your PMA/NP circuit? Does it need much grunt to drive it?
Cheers,
Hugh
- Status
- Not open for further replies.