THD has a much bigger effect on sound than you think

[MakeMineVinyl]

Even if true, If the total level (THD+N) is loud enough it shouldn't be an audible issue for most. As to even vs odd low order harmonics, take a look of Marantz vs Denon AVRs on the ASR chart and one may be surprised to see that Marantz's pretty much always have more 3rd than Denon's relatively speaking, yet we keep seeing posts about Marantz sound is warm and musical, relative to Denon's, good figure! I bet in a controlled DBT, using analog inputs and in direct mode even the die hard Marantz warm/musical sound believers wouldn't be able to tell that little difference in the 2nd and 3rd but then who knows, it not proven until it is, I can bet all I want..

I am not suggesting Nelson Pass was right when he said the following, but I bet it may surprise many who admire him:

Audio distortion and feedback - Pass Labs

Audiophiles seem to revel in minor controversies – vinyl vs CD’s, tubes versus solid state, capacitor, wires, magic dots… and negative feedback. At one extreme, the position is that “feedback makes amplifiers perfect”. At the other extreme, “feedback is a … Continue reading →

www.passlabs.com

So much for "facts".., perhaps the real "fact" is, it's more like opinions, preference..

So many people gush about the sound of analog tape, but the distortion produced is mostly odd order, and the more 'perfect' the machine is, the less there is of even order products. While I absolutely love my tape machines, I find that the distortion produced - which is about 0.6% - takes a definite hit on clarity vs the original input signal. Some people obviously like that effect.

Vinyl playback distortion tends to resemble that produced by a single ended triode amplifier with the strongest harmonic being the 2nd, followed by the 3rd, 4th and 5th, each in progressively lower levels.

 

[audio2design]

Without reading the whole thread, I am going to make a leap of faith with:

p.s. Loudness countour weighted harmonics is not a new concept.
pps Most solid state amps have their lowest distortions at low frequencies due to highest loop gain on feedback.
ppps It would be exceptionally rare (some poorly designed single ended amp perhaps) where the distortion in the speaker would not vastly exceed the solid state amp at <=200Hz

The op has poorly stated the perceived problem and coupled that with a poor knowledge of electronics and perhaps speakers.

Conceptually, if not clear, the op believes, I think, that since there is a lot more energy near 50-120Hz (approx), versus say 500-1000Hz. He suggested 20-40db (without knowing it). Hence believing that the "sound" at 50-120Hz would create much more audible distortion since the audible distortion is in a frequency range where there is less power .... perhaps with a helping of poorly understood equal loudness countours.

• At typical listening levels, there is closer to 10 db difference in equal loudness between 100Hz and 1KHz.
• Bass is rarely low in power in music, tending to be used for higher power "events". At high power the loudness countour from 100Hz to 1Khz is <5db
• Because the bass is loud, and the harmonics less so, you don't here the harmonics, they are masked (see previous two points). In music instruments, the harmonics are usually pretty substantial by the way.
• See points above about amplifiers and speakers
• Due to loudness countours, this could raise an eyebrow at low volume levels, fortunately speakers are most linear in this range. Fortunately not a lot of power normally at low frequencies when listening at average/typical levels

 

[Pdxwayne]

whole song.
Yes, with shorter pieces you see a bigger dynamic difference.

Let's analyze this plot and assume the amp used has 0.01% THD (-80 dB) where 2nd, 3rd, 4th and 5th are present.
We see bass notes 28Hz (A0) and 32.7Hz (C1)
There will be the following harmonics in the instrument (levels will differ)
56, 65.4, 84, 98.1, 112, 13.8, 140, 163.5 of course there will be more frequencies higher up as well all the way up to kHz range in increasingly small values.
-80dB is for the total harm energy so the individual harmonics will be lower in amplitude. Let's say -85dB each.
The level of the fundamental is -18dB and -20dB.
So the harmonics generated by this (not that great performing) amp will be around -103 and -105dB.
Now lets look at the plot in the range between 50Hz and 200Hz.
We see -40dB as the lowest signal. This is the musical content.
The distortion is 60dB below those signals so masked by the harmonics of the instrument itself.
The distortion signals of a 0.01% THD amp thus is 1000x lower than the music content.
Nope... you are not going to hear that.
Note that those lines will be dancing up and down.
When you would mix in other frequencies (or noise) not related and the dancing 'poles' are absent at those extra frequencies or are constant.. yes you can probably hear those quite far down.

Besides what OP claims is nonsense for electronics unless you move into the tube amp or amp territory where intentionally huge amounts of distortion is produced.
In that case it is not the HD but the IM distortion that becomes an issue.
The comment OP made about the furnace is not relevant to the amplifier. When you can hear the furnace when no music is playing it has already reached audible levels which are much higher than any generated (and masked) distortion even a half decent amp can reach.

Hmm, you were using the whole song earlier to make an argument about balance between bass and treble, which I think you agreed is not appropriate.

You seem to go the other extreme this time to argue about audibility of distortions. Don't you think you are using too little information this time?

Let's use OP's 1khz example. How many fundamental tones can produce harmonics to "around" that 1khz tone? You can't simply use one fundamental tone as an example and call it a day, right? You have to add all possible harmonics at around 1khz, right? Music is not just one fundamental tone.

Let's say speaker itself already near the edge of harshness at around 1khz (due to speaker's own distortions) when play music loud. When being pushed by higher distortion amp, would all the extra harmonics from amp make the speaker to start sounding harsh? What would be the tipping point?

 

[MRC01]

There are always many different tones playing in music, whether they're fundamentals or harmonics. ... The real harmonics are at similar levels to what is generated by amplifier distortion when an amplifier with high THD is used (say 0.2% to be very safe). ...

This may be true in some specific cases but is not generally true. The level of harmonics relative to fundamental of musical instruments vary quite a bit but are commonly much more than 0.2%. In fact, with some instruments it can exceed 100% - that is, the first harmonic or two can actually be louder than the fundamental!

What I am trying to say is that low notes, while physically not able to be "fast" like higher notes tend to be very punchy (bass drum), whereas higher notes tend to come with more sustain in actual music ...

This is true in some specific cases but is not generally true in either case, for bass or for treble. The bass could be a drum hit, or it could be a pipe organ, or a bass violin bowed (not plucked). The highs could be a sustained note, or it could be percussive hits (sticks on wood, bells, etc.).

 

[solderdude]

You seem to go the other extreme this time to argue about audibility of distortions. Don't you think you are using too little information this time?

Of course. Together with HD there is also IM (as mentioned earlier). What should not be forgotten is that music consists of fundamentals and harmonics that vary in time and harmonic spread and the tone defining ones will be magnitudes higher than those generated by a half decent amp.

Let's say speaker itself already near the edge of harshness at around 1khz (due to speaker's own distortions) when play music loud. When being pushed by higher distortion amp, would all the extra harmonics from amp make the speaker to start sounding harsh? What would be the tipping point?

I don't understand anything what you are talking about nor the question about tipping points.

 

[Pdxwayne]

I don't understand anything what you are talking about nor a question about tipping points.

Let's say a speaker is by default a bit bright sounding. Would using an amp with higher hd cause the speakers to sound harsher as compared to using amp with lower hd?

 

[audio2design]

No

 

[solderdude]

Bright sounding as in elevated a few dB above 3kHz ?
Why would an amp make any difference ?
For an amp to sound 'harsher' you are probably already clipping it.

 

[Pdxwayne]

Bright sounding as in elevated a few dB above 3kHz ?
Why would an amp make any difference ?
For an amp to sound 'harsher' you are probably already clipping it.

Elevated dB is partially the issue. Also about distortions level when playing louder, like at 90+dB vs at 80db or lower. If you check measurements for speakers, you will always see much higher distortions when go louder. Some more than others.

See https://www.audiosciencereview.com/...a100-review-powered-speaker.27425/post-944929 for 96db test.

So, let's say that that speakers in the link is passive and are playing near high distortion volume territory already, would an amp's HD make a difference in how soon/easy the speakers started to sound harsh?

 

[audio2design]

No

 

[Pdxwayne]

No

Please post your results for this test?

Blind Testing a 0.5 dB Level Difference

Discover the smallest difference in level (dB) you can reliably hear.

www.audiocheck.net

 

[audio2design]

Please post your results for this test?

Blind Testing a 0.5 dB Level Difference

Discover the smallest difference in level (dB) you can reliably hear.

www.audiocheck.net

No

 

[solderdude]

So, let's say that that speakers in the link is passive and are playing near high distortion volume territory already, would an amp's HD make a difference in how soon/easy the speakers started to sound harsh?

Let's go with your theory. At 86dB we see distortion at 400Hz we see distortion just below 50dB SPL. We see predominantly 3rd harmonic (so clipping alike) and is the harm. of 133Hz which also is 50dB SPL (so at . 2nd harm (200Hz) is at 44dB SPL. Lets assume the speaker is 86dB @1W.

Now lets take an audiophile universally loved and excellent sounding amp. We all assume this is the worst of the worst.
Fortunately Amir measured it at a realistic 1W level as well. 2nd harm = -50dB. 3rd harm = -72dB. Lets agree that below -95dB we cannot hear diddly squad because at 85dB SPL and -95dB we are already below a hearing threshold.

Lets agree, for argument sake, that the 50dB line is the limit where we start to hear ****** sound and it is linear with frequency.
86dB - 50dB = 36dB = 1.6% distortion (so not 0.1 and certainly not 0.01%) and for 2nd harm = 86dB - 44dB = 42dB = 0.8% distortion.

2nd arm. at 2.83V (=1W) -44dB = 0.017V (equivalent)
3rd harm. at 2.83V (=1W) -36dB = 0.045V (equivalent)

Why do I calculate this ? Because in order to add the dB's you have to calculate back to actual levels, then add the levels and convert back to dB.

Now the amp at this level has 2.83V out (1W/8ohm).
2nd harm = -50dB = 0.009V = 0.3%
3rd harm = -72dB = 0.0007V = 0.03%

Now the big trick... add levels of the rather huge distortion of the amp to that of the speaker.

2nd harm = 17mV + 9mV = 26mV
3rd harm = 45mV + 0.7 = 45.7 mV
Of course this is opposite 2.83V so we can now calculate the dB's SPL we end up with for the 2nd and 3rd harm. Higher harm are pointless to add as the amp will be below the hearing threshold for actual human beings (audiophools exempt).

2nd harm = -41dB (was -44dB on a good amp) so 2nd harm. distortion increased from 0.6% to 0.9% and is still below the magic 1.6%.
3rd harm = -35.8dB (was -36dB) so 2nd harm. distortion increased from 1.58% to 1.62% and is slightly above the magic 1.6%.

Now... you could hear 0.5dB increase in levels.
This means we now have to calculate how much the 400Hz increased in level (in dB).

2nd harm = increase of 0.3% opposite a good amp. = 0.026dB increase of level at 400Hz from a 200Hz fundamental. Remember this is the worst amp Amir had on his test bench till now.
3rd harm = 0.04% increase = 0.004dB increase.
Do you think you can actually hear this as a change in tonal balance ?

Of course... we haven't talked about IM products here which aren't harmonically related and alas... with HD also comes IM distortion.
IM distortion is also amplitude related and is only present when more tones.
Each note already has harmonics which could even be higher than the fundamental b.t.w.
The audibility of those IM products is another thing though because at very small levels (quiet passages) the HD and IM is also much lower so IM is also music (frequency and amplitude) dependent.

 

[Hayabusa]

Let's go with your theory. At 86dB we see distortion at 400Hz we see distortion just below 50dB SPL. We see predominantly 3rd harmonic (so clipping alike) and is the harm. of 133Hz which also is 50dB SPL (so at . 2nd harm (200Hz) is at 44dB SPL. Lets assume the speaker is 86dB @1W.

Now lets take an audiophile universally loved and excellent sounding amp. We all assume this is the worst of the worst.
Fortunately Amir measured it at a realistic 1W level as well. 2nd harm = -50dB. 3rd harm = -72dB. Lets agree that below -95dB we cannot hear diddly squad because at 85dB SPL and -95dB we are already below a hearing threshold.

Lets agree, for argument sake, that the 50dB line is the limit where we start to hear ****** sound and it is linear with frequency.
86dB - 50dB = 36dB = 1.6% distortion (so not 0.1 and certainly not 0.01%) and for 2nd harm = 86dB - 44dB = 42dB = 0.8% distortion.

2nd arm. at 2.83V (=1W) -44dB = 0.017V (equivalent)
3rd harm. at 2.83V (=1W) -36dB = 0.045V (equivalent)

Why do I calculate this ? Because in order to add the dB's you have to calculate back to actual levels, then add the levels and convert back to dB.

Now the amp at this level has 2.83V out (1W/8ohm).
2nd harm = -50dB = 0.009V = 0.3%
3rd harm = -72dB = 0.0007V = 0.03%

Now the big trick... add levels of the rather huge distortion of the amp to that of the speaker.

2nd harm = 17mV + 9mV = 26mV
3rd harm = 45mV + 0.7 = 45.7 mV
Of course this is opposite 2.83V so we can now calculate the dB's SPL we end up with for the 2nd and 3rd harm. Higher harm are pointless to add as the amp will be below the hearing threshold for actual human beings (audiophools exempt).

2nd harm = -41dB (was -44dB on a good amp) so 2nd harm. distortion increased from 0.6% to 0.9% and is still below the magic 1.6%.
3rd harm = -35.8dB (was -36dB) so 2nd harm. distortion increased from 1.58% to 1.62% and is slightly above the magic 1.6%.

Now... you could hear 0.5dB increase in levels.
This means we now have to calculate how much the 400Hz increased in level (in dB).

2nd harm = increase of 0.3% opposite a good amp. = 0.026dB increase of level at 400Hz from a 200Hz fundamental. Remember this is the worst amp Amir had on his test bench till now.
3rd harm = 0.04% increase = 0.004dB increase.
Do you think you can actually hear this as a change in tonal balance ?

Of course... we haven't talked about IM products here which aren't harmonically related and alas... with HD also comes IM distortion.
IM distortion is also amplitude related and is only present when more tones.
Each note already has harmonics which could even be higher than the fundamental b.t.w.
The audibility of those IM products is another thing though because at very small levels (quiet passages) the HD and IM is also much lower so IM is also music (frequency and amplitude) dependent.

After reading so many conclusions about audibility of distortion I decided to test my own ears wrt harmonic distortion.

This is what I did:
REW had a sine wave generator that allows for adding harmonic distortion.
For each harmonic (e.g. 2nd 3rd .. etc) you can mix it with percentage of the fundamental.
I used level calibrated headphones so I know at what SPL I am actually listening.
I limited the search for lowest audible distortion to only H2 H3.

I stepped tru fundamental frequencies, SPL levels and varied H2 and H3 to find the lowest distortion level I could still hear.

The results:
for H2 the lowest audible level when playing 800Hz at 60dB SPL was -54dB (0.2%)
for H3 this was also for 800Hz and 65dB SPL at -65dB (0.06%)

What surprised me is that SPL levels must be quite low for my ears to have the highest sensitivity for hearing this distortion.

Taking the H3 figure at 65dB SPL and a typical speaker sensitivity of 85dB/1w this means an amp should have less than 0.06% H3 @800Hz at only 10 milliwatts!

Maybe we should measure HD also at these low power levels?

 

[Pdxwayne]

Let's go with your theory. At 86dB we see distortion at 400Hz we see distortion just below 50dB SPL. We see predominantly 3rd harmonic (so clipping alike) and is the harm. of 133Hz which also is 50dB SPL (so at . 2nd harm (200Hz) is at 44dB SPL. Lets assume the speaker is 86dB @1W.

Now lets take an audiophile universally loved and excellent sounding amp. We all assume this is the worst of the worst.
Fortunately Amir measured it at a realistic 1W level as well. 2nd harm = -50dB. 3rd harm = -72dB. Lets agree that below -95dB we cannot hear diddly squad because at 85dB SPL and -95dB we are already below a hearing threshold.

Lets agree, for argument sake, that the 50dB line is the limit where we start to hear ****** sound and it is linear with frequency.
86dB - 50dB = 36dB = 1.6% distortion (so not 0.1 and certainly not 0.01%) and for 2nd harm = 86dB - 44dB = 42dB = 0.8% distortion.

2nd arm. at 2.83V (=1W) -44dB = 0.017V (equivalent)
3rd harm. at 2.83V (=1W) -36dB = 0.045V (equivalent)

Why do I calculate this ? Because in order to add the dB's you have to calculate back to actual levels, then add the levels and convert back to dB.

Now the amp at this level has 2.83V out (1W/8ohm).
2nd harm = -50dB = 0.009V = 0.3%
3rd harm = -72dB = 0.0007V = 0.03%

Now the big trick... add levels of the rather huge distortion of the amp to that of the speaker.

2nd harm = 17mV + 9mV = 26mV
3rd harm = 45mV + 0.7 = 45.7 mV
Of course this is opposite 2.83V so we can now calculate the dB's SPL we end up with for the 2nd and 3rd harm. Higher harm are pointless to add as the amp will be below the hearing threshold for actual human beings (audiophools exempt).

2nd harm = -41dB (was -44dB on a good amp) so 2nd harm. distortion increased from 0.6% to 0.9% and is still below the magic 1.6%.
3rd harm = -35.8dB (was -36dB) so 2nd harm. distortion increased from 1.58% to 1.62% and is slightly above the magic 1.6%.

Now... you could hear 0.5dB increase in levels.
This means we now have to calculate how much the 400Hz increased in level (in dB).

2nd harm = increase of 0.3% opposite a good amp. = 0.026dB increase of level at 400Hz from a 200Hz fundamental. Remember this is the worst amp Amir had on his test bench till now.
3rd harm = 0.04% increase = 0.004dB increase.
Do you think you can actually hear this as a change in tonal balance ?

Of course... we haven't talked about IM products here which aren't harmonically related and alas... with HD also comes IM distortion.
IM distortion is also amplitude related and is only present when more tones.
Each note already has harmonics which could even be higher than the fundamental b.t.w.
The audibility of those IM products is another thing though because at very small levels (quiet passages) the HD and IM is also much lower so IM is also music (frequency and amplitude) dependent.

Thank you for spending time explaining.

Yeah, there are IM too.

I assume that you are using 1 Watt amp output for your calculations. What about 5 watts+? Let's say, for a certain music passage, bass is taking most of the 5+ watts and it is at 90+db and thus producing the most harmonics. Let's assume the treble level is 65db. Would the cumulative loud bass harmonics add up to much more than 1db to the treble?

 

[MRC01]

...
Taking the H3 figure at 65dB SPL and a typical speaker sensitivity of 85dB/1w this means an amp should have less than 0.06% H3 @800Hz at only 10 milliwatts!
Maybe we should measure HD also at these low power levels?

Incidentally, this is one the arguments for audibility of crossover distortion: low level signals.

 

[solderdude]

What surprised me is that SPL levels must be quite low for my ears to have the highest sensitivity for hearing this distortion.

For the expanation see post #33 by @Blumlein 88

Note that hearing distortion in a constant single pure tone is much easier (and thus much lower values) than in music which does not consist of single constant frequencies.

Maybe we should measure HD also at these low power levels?

Amir also shows these plots but the lower you go the lower the distortion and the noise increases.
The distortion products thus drown in the noise.

Note that both humans and measurement equipment can detect constant tones buried below the noise floor. Unfortunately it doesn't help with music below the noise floor. You can't hear that.

This is the difficulty of determining audibility thresholds. You can do that with music and pure tones. For audio, in the end, only the numbers using music count.
For 'safety reasons' in audio engineering we look at audibility thresholds using constant sine waves as worst case. Everything better than that simply means it does not matter.

 

[solderdude]

I assume that you are using 1 Watt amp output for your calculations. What about 5 watts+? Let's say, for a certain music passage, bass is taking most of the 5+ watts and it is at 90+db and thus producing the most harmonics. Let's assume the treble level is 65db. Would the cumulative loud bass harmonics add up to much more than 1db to the treble?

Nope. You saw an example of the worst measuring amp. At 5W the distortion of the speaker increases more than that of the amp.
Assume bass is 100Hz and looking at the performance of the amp we see H5 (500Hz) to be below any audibility levels. H10 (1kHz) is 130dB down.
For that to become audible you would have to play at 140dB SPL in an anechoic room. The 140dB 100Hz would make you put your fingers in your ears and makes you want to get out of the room assuming the speaker could reproduce that without any distortion.
With all but the 1kHz harmonic playing and the rest removed your ears would have to get used to the silence and then you can very faintly hear a faint 'beeeeeeep' but certainly not when 100Hz is played at the same time at 140dB.

 

[Pdxwayne]

Nope. You saw an example of the worst measuring amp. At 5W the distortion of the speaker increases more than that of the amp.
Assume bass is 100Hz and looking at the performance of the amp we see H5 (500Hz) to be below any audibility levels. H10 (1kHz) is 130dB down.
For that to become audible you would have to play at 140dB SPL in an anechoic room. The 140dB 100Hz would make you put your fingers in your ears and makes you want to get out of the room assuming the speaker could reproduce that without any distortion.
With all but the 1kHz harmonic playing and the rest removed your ears would have to get used to the silence and then you can very faintly hear a faint 'beeeeeeep' but certainly not when 100Hz is played at the same time at 140dB.

I am talking about "cumulative" bass distortions. Not a single tone distortion. For example, add all distortions from the bass notes, for example 20 to 200 hz,

 

[Flak]

Note that hearing distortion in a constant single pure tone is much easier (and thus much lower values) than in music which does not consist of single constant frequencies.

Amir also shows these plots but the lower you go the lower the distortion and the noise increases.
The distortion products thus drown in the noise.

Note that both humans and measurement equipment can detect constant tones buried below the noise floor. Unfortunately it doesn't help with music below the noise floor. You can't hear that.

This is the difficulty of determining audibility thresholds. You can do that with music and pure tones. For audio, in the end, only the numbers using music count.
For 'safety reasons' in audio engineering we look at audibility thresholds using constant sine waves as worst case. Everything better than that simply means it does not matter.

Let me quote the above which leads me to think the opposite of the thread's title... maybe

THD has a much smaller effect on sound than you think ​

as music is made of transients and my understanding is that even a 10% THD cannot be identified by the listener if only present during peaks that last a few milliseconds.
Tests about the audibility of distortion with music, instead of test signals, have been done for example by Archimago but I think that the same amount of distortion had been injected during the whole playback while in real life the distortion is related to the output level.
In a few words determining the thresholds of THD's audibility during playback of real music probably isn't simple, please correct me if I'm wrong!