Harmonics masked by what? Listening tests with what for speakers?
Here is spectrum of 375Hz fundamental with:
- H2 -22dB
- H3 -26dB
- H4 -30dB
- H5 -37dB
Attached is wave file with above construction on left channel; pure 375Hz on right. I'm sure HD850 will allow you to tell the difference.
Why settle for amplifiers and transducers that may unmask their crappy undersides when many options, and affordable ones, exist for most applications having distortion at or below the threshold of perception under any non-clipping signal conditions?
By the fundamental and the other harmonics. It's a function of the ear. I believe that tests were done at levels near 75-80dB SPL and at 400Hz, among others.
I don't know. But I would imagine they were not the same in 1924 and in 1960s.
When I find the original research we'll know more.That's one argument, however it's not always true of real circuits and real drivers. A circuit that is well behaved at one frequency, load or level may not be so well behaved at others. (or so I've measured and seen measured) Drivers are even less well behaved.
Your file is nice, but it's the original spectrum, not the final one produced by the amp and speakers/headphones. That's what counts.
The gist I have, from genaral reading, is that a particular ratio of harmonics tends to be less intrusive than the simple sum of its component parts. Only those harmonics that approximate closely to music notes count, so 7th and higher primes are always relatively conspicuous. It's possible that this is an argument dreamed up by valve afficionados/apologists. There are lots of variations.
The implication would then be that, if you have an objectionable amount of 5th harmonic, it could be made less so with the addition of the right amount of 2nd and 3rd.
Why the 4th should be different from the 2nd by this account, I don't know. Nor whether the program material makes any difference.
I guess there's been all sorts of hypotheses. Here we have contributors with the equipment and skills, and plenty of ears, to carry out a sensible test maybe.
But it's a horribly complicated issue really. Maybe that's why so many engineers get scared and opt for scorched earth, just in case.
The implication would then be that, if you have an objectionable amount of 5th harmonic, it could be made less so with the addition of the right amount of 2nd and 3rd.
Why the 4th should be different from the 2nd by this account, I don't know. Nor whether the program material makes any difference.
I guess there's been all sorts of hypotheses. Here we have contributors with the equipment and skills, and plenty of ears, to carry out a sensible test maybe.
But it's a horribly complicated issue really. Maybe that's why so many engineers get scared and opt for scorched earth, just in case.
Hi Everybody,
this is "Martina" ...
I have just by pure accident sent Nigel's half an hour reply to Internet Nirvana,
so I thought its only appropriate to apologize here...
I am with Nigel as I just went to England to get myself a new 6 cyl. XJ6 3,2 L engine for
my X300 Jaguar amd enjoy Tech Talk with Nigel ( my pistons did burst due to wrong air intake and throttle problems , so a replacement engine had to be found...)
What this to do with Audio you might ask...
Well everything!
.... And my car stereo in the Jag has the Best sounding cassette deck I have ever used!!!!
I share a bit about the genesis of the Priest:
The PSU is exactly as somebody said: dealing with inductance in the PCB
and the solution is exactly as you say.
So please continue to enjoy this thread.
Joachim Gerhard, Michael Cech, Nigel and I work together on projects.
We have one philosophy in common :
Anything is right if it sounds right ( like all is well that ends well ... and if it isn't well
it is not the end yet!)
So I give back the posting to Nigel!
Best Wishes to all of You!
Martina Schoener
Martina stopped last night before going to Wales to get the engine . She dumped my replay I was writing about op amp . She will introduce me to Jean . Martina does write on the forums under her own name . MPP thread she says . She is now 15 minutes late for Wales !!!! I doubt she writes as much as me .
this is "Martina" ...
I have just by pure accident sent Nigel's half an hour reply to Internet Nirvana,
so I thought its only appropriate to apologize here...
I am with Nigel as I just went to England to get myself a new 6 cyl. XJ6 3,2 L engine for
my X300 Jaguar amd enjoy Tech Talk with Nigel ( my pistons did burst due to wrong air intake and throttle problems , so a replacement engine had to be found...)
What this to do with Audio you might ask...
Well everything!
.... And my car stereo in the Jag has the Best sounding cassette deck I have ever used!!!!
I share a bit about the genesis of the Priest:
The PSU is exactly as somebody said: dealing with inductance in the PCB
and the solution is exactly as you say.
So please continue to enjoy this thread.
Joachim Gerhard, Michael Cech, Nigel and I work together on projects.
We have one philosophy in common :
Anything is right if it sounds right ( like all is well that ends well ... and if it isn't well
it is not the end yet!)
So I give back the posting to Nigel!
Best Wishes to all of You!
Martina Schoener
Martina stopped last night before going to Wales to get the engine . She dumped my replay I was writing about op amp . She will introduce me to Jean . Martina does write on the forums under her own name . MPP thread she says . She is now 15 minutes late for Wales !!!! I doubt she writes as much as me .
Thanks Pano .
If I get to meet Jean I will ask if he knows more on this . Martin Colloms also as I plan to meet with him soon on RFI via a friend .
The point people don't seem to get here is the Hiraga harmonics and the zero distortion sound correct . Does this simplify our goals ? No . After that it is whatever shade of grey you want . You can have white or paper as alternatives .
If I get to meet Jean I will ask if he knows more on this . Martin Colloms also as I plan to meet with him soon on RFI via a friend .
The point people don't seem to get here is the Hiraga harmonics and the zero distortion sound correct . Does this simplify our goals ? No . After that it is whatever shade of grey you want . You can have white or paper as alternatives .
One thing to say . It can not be harmonics alone as the speaker influences the outcome . It must be a manifestation of the cogs in the machine . It is perhaps what Sergey was saying ? It is evidence of an unsuspected mechanism ? Good enough to know it's signature I would say . I am more of a visual person . I can make a machine come alive in my head . Years of repairing things I guess ? To put words to it is hard . A machine in harmony is exactly that . I called it symmetry before . I meant harmony .
Paper is not exactly white and it has texture . We take it as white . The texture is OK . We still can say text on paper is superior to a computer screen ( perhaps ) . It was suggested to me we absorb information from paper better than the screen . Paper is LP , screen is visual CD ?
Одно дело сказать. Это не может быть гармоник только как динамик влияет на результат. Оно должно быть проявлением винтики в машине. Это, пожалуй, то, что Сергей говорил? Это свидетельствует о неожиданной механизм? Этого вполне достаточно знать, что это подпись я бы сказал. Я больше визуальный человек. Я могу сделать машину оживают в моей голове. Годы ремонта вещи я думаю? Говоря словами трудно.Машина в гармонии является именно это. Я назвал его симметрии раньше. Я имел в виду гармонию.
Бумага не совсем белый и имеет текстуру. Мы воспринимаем как белый.Текстуры в порядке. Мы все еще можем сказать, текст на бумаге превосходит экране компьютера (возможно). Было предложено мне, что мы впитываем информацию из бумаги лучше, чем на экране. Бумага LP, экран визуальной CD? Одно дело сказать. Это не может быть гармоник только как динамик влияет на результат. Оно должно быть проявлением винтики в машине. Это, пожалуй, то, что Сергей говорил? Это свидетельствует о неожиданной механизм? Этого вполне достаточно знать, что это подпись я бы сказал. Я больше визуальный человек. Я могу сделать машину оживают в моей голове. Годы ремонта вещи я думаю? Говоря словами трудно.Машина в гармонии является именно это. Я назвал его симметрии раньше. Я имел в виду гармонию.
Бумага не совсем белый и имеет текстуру. Мы воспринимаем как белый.Текстуры в порядке. Мы все еще можем сказать, текст на бумаге превосходит экране компьютера (возможно). Было предложено мне, что мы впитываем информацию из бумаги лучше, чем на экране. Бумага LP, экран визуальной CD?
Paper is not exactly white and it has texture . We take it as white . The texture is OK . We still can say text on paper is superior to a computer screen ( perhaps ) . It was suggested to me we absorb information from paper better than the screen . Paper is LP , screen is visual CD ?
Одно дело сказать. Это не может быть гармоник только как динамик влияет на результат. Оно должно быть проявлением винтики в машине. Это, пожалуй, то, что Сергей говорил? Это свидетельствует о неожиданной механизм? Этого вполне достаточно знать, что это подпись я бы сказал. Я больше визуальный человек. Я могу сделать машину оживают в моей голове. Годы ремонта вещи я думаю? Говоря словами трудно.Машина в гармонии является именно это. Я назвал его симметрии раньше. Я имел в виду гармонию.
Бумага не совсем белый и имеет текстуру. Мы воспринимаем как белый.Текстуры в порядке. Мы все еще можем сказать, текст на бумаге превосходит экране компьютера (возможно). Было предложено мне, что мы впитываем информацию из бумаги лучше, чем на экране. Бумага LP, экран визуальной CD? Одно дело сказать. Это не может быть гармоник только как динамик влияет на результат. Оно должно быть проявлением винтики в машине. Это, пожалуй, то, что Сергей говорил? Это свидетельствует о неожиданной механизм? Этого вполне достаточно знать, что это подпись я бы сказал. Я больше визуальный человек. Я могу сделать машину оживают в моей голове. Годы ремонта вещи я думаю? Говоря словами трудно.Машина в гармонии является именно это. Я назвал его симметрии раньше. Я имел в виду гармонию.
Бумага не совсем белый и имеет текстуру. Мы воспринимаем как белый.Текстуры в порядке. Мы все еще можем сказать, текст на бумаге превосходит экране компьютера (возможно). Было предложено мне, что мы впитываем информацию из бумаги лучше, чем на экране. Бумага LP, экран визуальной CD?
Last edited:
Very good observation.
I like it when a single explanation covers all the questions in a consistent way.
jan
Joachim Gerhard told me years ago about experiments on perception . He made me beleive zero distortion and zero noise is the goal . He also made me believe it should sound soft !!!
The surprise is there is another zero . If one could do it correctly you might be surprised how similar they sound . Between is all shades of taste .
The sterile sound is residuals . As I said before I think it is a byproduct of the mechanism . I don't think the residuals matter , they simple can't when - 90 dB . It is the machine that makes them that matters . Lack of harmony disguised by a feedback loop I would guess ? Did the music deceive the feedback loop ?
The surprise is there is another zero . If one could do it correctly you might be surprised how similar they sound . Between is all shades of taste .
The sterile sound is residuals . As I said before I think it is a byproduct of the mechanism . I don't think the residuals matter , they simple can't when - 90 dB . It is the machine that makes them that matters . Lack of harmony disguised by a feedback loop I would guess ? Did the music deceive the feedback loop ?
By the fundamental and the other harmonics. It's a function of the ear. I believe that tests were done at levels near 75-80dB SPL and at 400Hz, among others.
I don't know. But I would imagine they were not the same in 1924 and in 1960s.
That's one argument, however it's not always true of real circuits and real drivers. A circuit that is well behaved at one frequency, load or level may not be so well behaved at others. (or so I've measured and seen measured) Drivers are even less well behaved.
Your file is nice, but it's the original spectrum, not the final one produced by the amp and speakers/headphones. That's what counts.
Function of the ear? Mine apparently function quite well; The distorted version of the 375Hz tone sounds like crap. I changed phasing of harmonics; still sounds like crap. I changed fundamental between 200-1200Hz. All sound like crap.
Any real circuit that forms harmonic structure like this:
In any portion of its bandwidth is a crappy circuit.
If all the signals pass through driver with harmonic distortion as above, telling the difference between clean v dirty becomes very difficult. The majority of compression drivers exhibit distortion profiles not unlike the above.
If point of reference is high distortion amplifier playing through very good driver, then performance of high distortion speaker is perceived as less distorted than it really is.
Nigel;
If above harmonic structure I was given is anything close to Hiraga harmonics, they sound nowhere close to correct.
When IMD that is brought into play with non linear systems producing harmonic distortion as above the effects are far, far away from correct; and typically sound like total crap.
You'll need to read the article. Incomplete information is causing folks here to jump to all sorts of conclusions - I apologize for that. There may be a translation around somewhere, I don't know.
The levels referenced are acoustic, not electrical. The Hiraga article cites some FFT work done in June 1974 by the Japanese magazine "Radio Technology". They apparently tested 55 amps and compared subjective impressions of the amps to the distortion spectra. I've tried in vain to find the magazine. I'd love to learn more.
You can see by the fuzzy image of the charts attached below that the distortion of the amps themselves is not nearly as high as the masking figures given by the ear research. It's the spectra that is said to lead to different subjective impressions. I don't understand why there are differences in the numbers between ear and amp research, that's one thing I'd like to learn.
I can translate the captions, if you like. Will also look for some harmonics of Hiraga amps.
The levels referenced are acoustic, not electrical. The Hiraga article cites some FFT work done in June 1974 by the Japanese magazine "Radio Technology". They apparently tested 55 amps and compared subjective impressions of the amps to the distortion spectra. I've tried in vain to find the magazine.
I'd love to learn more.You can see by the fuzzy image of the charts attached below that the distortion of the amps themselves is not nearly as high as the masking figures given by the ear research. It's the spectra that is said to lead to different subjective impressions. I don't understand why there are differences in the numbers between ear and amp research, that's one thing I'd like to learn.
I can translate the captions, if you like. Will also look for some harmonics of Hiraga amps.
Attachments
Nigel, i remember our discussion well.
It was not about distortion.
It was about the noise.
I talked about " noise modulation".
I found when i measured the noise spectrum of tubes it was kind of "pink" and BJT transitors showed kind of " white" spectrum.
I could be then that the silicon had much less noise in the bass but over 1kHz the tube was better. I reasoned that transistor noise is more annoying then tube noise so should be as low as posible, ultimately close to zero.
The reasons for this are .....
About distortion there could also be setup a therory that tube distortion is " nice " and BJT distortion " not so nice ". what i know now it is not simply exponential decay of harmonics but more " unmusical certain change " over level and spectrum.
Continuos distortion even when high ( say 1% ) is more benigh than a rapid change over lever and frequency in amount and spectrum, even if the level of distortion is lower.
One example : fix a little bell on your woofer membranes and measure second and third harmonic distortion. With or without the bell it will measure very similar but you hear the bell ringing and buzzing along with the signal.
It was not about distortion.
It was about the noise.
I talked about " noise modulation".
I found when i measured the noise spectrum of tubes it was kind of "pink" and BJT transitors showed kind of " white" spectrum.
I could be then that the silicon had much less noise in the bass but over 1kHz the tube was better. I reasoned that transistor noise is more annoying then tube noise so should be as low as posible, ultimately close to zero.
The reasons for this are .....
About distortion there could also be setup a therory that tube distortion is " nice " and BJT distortion " not so nice ". what i know now it is not simply exponential decay of harmonics but more " unmusical certain change " over level and spectrum.
Continuos distortion even when high ( say 1% ) is more benigh than a rapid change over lever and frequency in amount and spectrum, even if the level of distortion is lower.
One example : fix a little bell on your woofer membranes and measure second and third harmonic distortion. With or without the bell it will measure very similar but you hear the bell ringing and buzzing along with the signal.
Last edited:
Dear Pano, Now I understand what you mean, but I think all of these graphs do not apply to masking. Аll amplifiers of different power and possibly different in structure, I could try to explain it ,but this is part of my topic, sorry. I do not think another way you can get an explanation.
I am not hindered by a digression, try and so.
I am not hindered by a digression, try and so.
Last edited:
Hi Joachim . That's me isn't it ? We were walking across the bridge at Blenheim Palace . I take noise to be distrotion . There was a mechanical analysis of damping factor circa 1955 that concluded 3 as a minimum . We must remember the cones had stiff surrounds them . I love when people ask the same questions in a different way .
Beethoven (1944 Stereo): Adagio from Concerto for piano and orchestra No.5 - Gieseking/RO Berlin - YouTube
This one to the day was rerecorded 12 years before my birth . Transmitted a few months later .
http://www.youtube.com/watch?v=hFgn3Nr9WUI
One thing to say . It can not be simple spectra . Good distorted amplifiers sound good through a variety of speakers that have different distortion themselves . The better valve amp quickly forces one to remove the feedback loop . If courageous levels of 1 % THD are possible with just 2 valves . If the speakers are suitable this sounds remarkable like classical music . So much so that one remembers the difference between a concert and good hi fi . It is the very strange interplay of instruments . Like freezing strawberry's one usually can not bring the orchestra to life at home. Beethoven is such . The true sound is most like 1938 78's capture , there are good technical reason why that would be true . By 1944 microphones were in some ways as good as they ever would be .
I would say what happens by accident is we find the truth . If we can make something which is on the edge of acceptable with good spectra we probably have made a good machine ? My suggestion is if we prefer to make damping factor a priority we should push on to " double zero " . I take noise to be distortion . However noise can be useful to make transistors switch nicely . So critical noise is important . Do not throw away a free lunch here . Equally class A transistor is not always ideal . I am sure with care AB can be proven to be better from the PSU point of view . Remember in a high damping factor amp the amplifier is just a PSU switching device . Class AB ( B ) is a very big free lunch . Quad 303 seems to show maintaining desirable harmonics will give a valve like signature even when 20 db below any sane assessment of requirements ? Now guys how many conversions we have here are sane when better than - 60 dB THD ? Like yourselves I do it because I can , not because it is better ? If the mechanism is important and feedback is for " fun " , then the open loop might be the answer ? Get it right open loop , then add feedback . Statements of the bleeding obvious ? I know the complexities . Often have a nasty amp that accepts feedback happily . That can work . Even so , look for the proper spectrum . I know it matters , I just don't know exactly why . The transistor are pistons if you like . Are they timing correctly ? Is not a valve a soft tune device that rolls more easily ?
Last edited:
All I can see is the pistors . It is not exactly spectra . It is harmony and balance . It can not be spectria for reasons given . What people forget is class AB is two halves . How the machine works is complicated . The speaker works as a mechanical integrator . If we help it do it's jobs all is happy . The best class B by comparison to SE class A sounds a bit like MP3 . Something is missing . It almost is unimportant ( 98% ? ) . Notes going into noise . They stop slightly short of zero . If the amplifier is not so good if the last part is modulated . This can be from a not obvious source . RFI is one . Some people earth the metal of the speaker to help that . It is cheap and might work for some .
The important thing is to see the machine in your head . You push the speaker , you pull the speaker . The amplifier is like a gearbox that connects the PSU to the wheels that is the speaker . How good is the gearbox . Not just power deliver .
My friend Martin Renwick makes esoteric bicycles . They have a battery to store energy . The bicycle is under minute computer control . It gives back the energy to make a non cyclist enjoy riding . As Martin says the bike seems to fly like a bird . Nothing like a motorcycle he says . My boss took me out in his Porsche ( above Boxster ) . It just doesn't fly like a bird . Point and squirt . I strongly disliked it . Mazda MX 5 is not like that . This is the mechanism I think is available in amplifiers . I have ridden Speedway bikes ( Jawa 2 valve ) . The Porsche is pale by comparison for 0 to 100 kph . Someone told me such cars are Fanny Magnets . In the USA I might get away with saying that !!! When I told a German guy that , he nearly peed himself ( his dog was called Fanny ) . He agreed , his 911 attracted a wife he ultimate did NOT get on with ( still together ) . He said he should have let his Beetle speak for him . He much preferred the Beetle . I drove a restored Beetle with Porsche bits fitted . The engine was standard , largest size used . It was a peach with none of the Beetle or 911 problems . Just right , the sweet spot . It cost $25 000 in 1980 to build . His friend got some Italians in Italy to rebuild a MGB chrome bumper convertible . They had no idea how and were annoyed he asked them . They made it into an Italian car . It was 100 % correct , only better than new . Oh how I wanted it .
This is how crazy ideas start . Why not make it work above 200 kph ? It is Martina's new old engine . I said why not direct drive above 200 kph ? Below that switched off . Class G . Before you ask it is just the engine will fit in the boot . Martina is rebuilding the engine . When I gave her the job title engineer I said she must be one . Her father says girls can not be mechanical engineers . The 3.2 litre 6 has the sweet spot ( fuel )
The important thing is to see the machine in your head . You push the speaker , you pull the speaker . The amplifier is like a gearbox that connects the PSU to the wheels that is the speaker . How good is the gearbox . Not just power deliver .
My friend Martin Renwick makes esoteric bicycles . They have a battery to store energy . The bicycle is under minute computer control . It gives back the energy to make a non cyclist enjoy riding . As Martin says the bike seems to fly like a bird . Nothing like a motorcycle he says . My boss took me out in his Porsche ( above Boxster ) . It just doesn't fly like a bird . Point and squirt . I strongly disliked it . Mazda MX 5 is not like that . This is the mechanism I think is available in amplifiers . I have ridden Speedway bikes ( Jawa 2 valve ) . The Porsche is pale by comparison for 0 to 100 kph . Someone told me such cars are Fanny Magnets . In the USA I might get away with saying that !!! When I told a German guy that , he nearly peed himself ( his dog was called Fanny ) . He agreed , his 911 attracted a wife he ultimate did NOT get on with ( still together ) . He said he should have let his Beetle speak for him . He much preferred the Beetle . I drove a restored Beetle with Porsche bits fitted . The engine was standard , largest size used . It was a peach with none of the Beetle or 911 problems . Just right , the sweet spot . It cost $25 000 in 1980 to build . His friend got some Italians in Italy to rebuild a MGB chrome bumper convertible . They had no idea how and were annoyed he asked them . They made it into an Italian car . It was 100 % correct , only better than new . Oh how I wanted it .
This is how crazy ideas start . Why not make it work above 200 kph ? It is Martina's new old engine . I said why not direct drive above 200 kph ? Below that switched off . Class G . Before you ask it is just the engine will fit in the boot . Martina is rebuilding the engine . When I gave her the job title engineer I said she must be one . Her father says girls can not be mechanical engineers . The 3.2 litre 6 has the sweet spot ( fuel )
Last edited:
I don't love Porsche, mine 320i BMW SE 2012, me quite arranges.
You read on Italian?
Vecchie riviste
Here happened?
You read on Italian?
Vecchie riviste
Here happened?
I 've found different things. It harash but warm(enjoyable), when loaded with loudspeaker. And has less effect while used as VAS or non speaker loaded preamp.
That is why sony ta1055 sounds so warm at bass with extra diode used at output.
Some type of tubes has bad sound, some others are very nice, like 300B it really has something, looks like it is not caused by distortion figures, figures are similar one and another.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.