Newbie post.
There have been studies about how much distortion one can hear by well trained listeners. A sensible thing for normal listeners would be to be perfectly happy with an amplifier that surpasses these specifications even by small margin.
For subjective sound quality people hear differently, have different hearing capacity, different rooms, placement etc. so it should be left to small number of people who discuss it in forum or in an audio meet. If one intends to buy or want to experience other persons subjective experience one can only take cues from such discussions. But it should be done without having any beliefs, biases, power of suggestion and by being rational.
regds.
There have been studies about how much distortion one can hear by well trained listeners. A sensible thing for normal listeners would be to be perfectly happy with an amplifier that surpasses these specifications even by small margin.
For subjective sound quality people hear differently, have different hearing capacity, different rooms, placement etc. so it should be left to small number of people who discuss it in forum or in an audio meet. If one intends to buy or want to experience other persons subjective experience one can only take cues from such discussions. But it should be done without having any beliefs, biases, power of suggestion and by being rational.
regds.
It is certainly true that high order harmonics are offensive.
2nd (octave) and 3rd (octave and perfect 5th) harmonic distortions, although used as a metric of amplifier performance are intervals that appear in all major and minor scales. But higher odd order harmonics such as 5th and 7th are much more sensitive to perception because they bear no harmonic relation to the music.
I seem to recall, although I cannot now find a reference on line, that concert grand pianos are designed in a way that suppresses high order harmonics from parts of the piano structure.
2nd (octave) and 3rd (octave and perfect 5th) harmonic distortions, although used as a metric of amplifier performance are intervals that appear in all major and minor scales. But higher odd order harmonics such as 5th and 7th are much more sensitive to perception because they bear no harmonic relation to the music.
I seem to recall, although I cannot now find a reference on line, that concert grand pianos are designed in a way that suppresses high order harmonics from parts of the piano structure.
So both H2 and H3 added together?
Or each one separately?
Did it sound "better" or "worse" (although that is probably difficult to say with sine tones)?
I would rather say that THD predicts bad sound quality.
But there are other factors that can predict bad sound quality.
And in my experience it is the top value that sadly enough seems to make us react.
I made an experiment with my pupils about 40 years ago.
I made my class of furture electronic servicemen listen individually to music played with an amplifier there i could switch on and off idle current to make crossover distortion.
They all reacted for 0,1% top value. No difference between the pupils. Some were Hifi interested others not at all.
The rms value of the distortion was probably under 0,01%.
The music was Dire Straits Brothers in arms CD. And of course the song where they are carrying color TV:s.
I have a friend hearing data reduction artefacts that i dont hear. I have a problem with bad damping in speaker boxes and room acoustic in many Tv crimis. It is possible to train to hear special things in music.
I once saw a quire rehersal on tv where the leader stopped, pointed at one of more than 50 persons and said you were little high on the C.
Out of 50 persons singing equal load he heard wich one was a little bit of key.
But there are other factors that can predict bad sound quality.
And in my experience it is the top value that sadly enough seems to make us react.
I made an experiment with my pupils about 40 years ago.
I made my class of furture electronic servicemen listen individually to music played with an amplifier there i could switch on and off idle current to make crossover distortion.
They all reacted for 0,1% top value. No difference between the pupils. Some were Hifi interested others not at all.
The rms value of the distortion was probably under 0,01%.
The music was Dire Straits Brothers in arms CD. And of course the song where they are carrying color TV:s.
I have a friend hearing data reduction artefacts that i dont hear. I have a problem with bad damping in speaker boxes and room acoustic in many Tv crimis. It is possible to train to hear special things in music.
I once saw a quire rehersal on tv where the leader stopped, pointed at one of more than 50 persons and said you were little high on the C.
Out of 50 persons singing equal load he heard wich one was a little bit of key.
This is kind of a silly question. It is well known that there is more that matters than THD as typically measured, or even SINAD for that matter. Bob Cordell's 2nd edition of his power amp book has a whole chapter on "other sources of distortion" and its not only stuff that only can produce THD as an audible effect. Then there is also stuff like asymmetrical crosstalk between channels, ability to drive a difficult speaker load (say, ranging from 2 ohms to 20ohms from moment to moment), channel independence at very high power levels, EMI incursion suseptibility, blah, blah, blah.
To focus on THD or HD profiles alone (often only measured one channel at time) should be something every person here knows is not all there is that can matter.
To focus on THD or HD profiles alone (often only measured one channel at time) should be something every person here knows is not all there is that can matter.
I am much happier listening to great sounding effect processor with pleasing musical distortion than boring amplifier with barely measurable distortion giving me headache and listening fatigue.
^ At least a few. I don't know that any of them made attempts to settle what seem to be the never-ending discussions around...
Can objective / instrument-based measurements predict:
Can objective / instrument-based measurements predict:
- any single human's (nevertheless a group's) enjoyment of a music playback system or component?
- what any majority (or even a plurality) of people might characterize as "good" sound quality?
Ok, found it. This was the thread which promissed to come up with parameters to fully characterize amps. I stopped following it. But here it is:
Thread 'AFOM: An attempt at an objective assessment of overall amplifier quality' https://www.diyaudio.com/community/...sessment-of-overall-amplifier-quality.407222/
I believe that @gedlee created a weighted distortion metric that could best distill the audibility of harmonic distortion down to a single number. It may not be perfect, it appears to me to be a much better metric than THD.
Finally, "sound quality metric" is subjective, varies with different support equipment, people, time of day, and glasses of beer consumed. It cannot be reliably quantified and therefore should always be a secondary assessment.
THD just predicts the amount of harmonics an amp will add to the signal that will be amplified. This is a very objective information.
All amplifiers add harmonics. The minimum level you should accept depends on what you are able to hear. Pay or build what is enough for you.
You have other parameters to look at such as noise, power, damp factor, impedance, slew rate, frequency response etc.
Subjective sound quality is only valid for you. Most modern marketing "reviews" are based on subjective SQ for convinience since they don't have to show numbers or any test results. They can say anything they want to promote the product. And the reviewer doesn't need to be a qualified person.
On the other hand, if YOU listen to an amp and you like it, that's enough to decide for you and subjective was useful.
For me, any amp with THD less than 0.3% (-48dB) is ok since I'm no longer able to detect less THD levels.
I made this test yesterday creating wav files in Audacity and adding 2nd, 3rd harmonics.
Others may require less THD cause they can detect it.
Snake oil salesmen always did that: They convince customers that quantitative criteria arent relevant, you just have to believe and snake oil will work. They are professionals, they know how to bias you into "halfwitt" mode.
“Ignorance more frequently begets confidence than does knowledge: it is those who know little, not those who know much, who so positively assert that this or that problem will never be solved by science."
(Charles Darwin-The Descent of Man)
Last edited:
I'm in the ultra low THD+N camp now but have owned VTL (amongst other) alternates. Mr. AKSA are you suggesting there is something transmitted in the electrical signal that can't be measured?
Frequency response with a loudspeaker load is more important, as the non-linear distortions of any half-decent amplifier are negligible anyway (as long as you don't drive it into clipping).
How distortion changes with level is important. There are a few brands out there that use a conventional and decent measurement voltage amplifier, either solid state or valved and an open loop follower output stage, either bipolar or mosfet. These tend to have very high crossover distortion - extra "detail", but measure low distortion at full power.
People like it.
People like it.