• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Revel M106 Bookshelf Speaker Review

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
Well, Linkwitz has written the following to which I agree.
Take into account the equal loudness contour and then you'll find that for the 2nd harmonic of for instance 20Hz where it falls is 40Hz and our ear is more sensitive there. So much more sensitive that for a 90dB SPL 20Hz tone, if the 2nd HD is 10% that is 70dB SPL our ear hears it at the same volume as the fundamental. And if the 2nd HD is higher than 10% for 20Hz we hear the 2nd HD as louder than the fundamental.
For the 3rd harmonic it is even worse, a 90dB SPL 20Hz tone can only have 3% 3rd HD before the 3rd harmonic becomes louder to the ear than the 20Hz fundamental.
Our ears/brain can do a trick though, they can fill in the missing fundamental in such a case. But don't think you're actually hearing good deep bass under the crazy distortion figures most 2-way speakers measure. You're not.
https://en.wikipedia.org/wiki/Equal-loudness_contour
So yes, agreed that the prevailing opinion on not being able to hear those crazy HD figures in the bass is complete bs. (also from personal experience)

I thought I had read most of what Linkwitz provided about speaker design, on his site, but I don't recall seeing this. And it is an angle that had not occurred to me, even though it seems obvious once made aware of it. To say it succinctly, since our hearing is more sensitive at the harmonics of a deep bass note than at the deep bass note itself (the fundamental), harmonic distortion does not have to be especially great in order for the distortion to sound louder to us than the fundamental.
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
I don't think a poorly integrated sub is that bad. It's just not that good either. A well-integrated sub with room correction is amazing, though.

By "happy place", do you mean the men's restroom at the Minneapolis–St. Paul International Airport?

Uh, no, I've never been there, but because of the way you mentioned it, my inference is that it is not my kind of place.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU
Well, Linkwitz has written the following to which I agree.
Take into account the equal loudness contour and then you'll find that for the 2nd harmonic of for instance 20Hz where it falls is 40Hz and our ear is more sensitive there. So much more sensitive that for a 90dB SPL 20Hz tone, if the 2nd HD is 10% that is 70dB SPL our ear hears it at the same volume as the fundamental. And if the 2nd HD is higher than 10% for 20Hz we hear the 2nd HD as louder than the fundamental.
For the 3rd harmonic it is even worse, a 90dB SPL 20Hz tone can only have 3% 3rd HD before the 3rd harmonic becomes louder to the ear than the 20Hz fundamental.
Our ears/brain can do a trick though, they can fill in the missing fundamental in such a case. But don't think you're actually hearing good deep bass under the crazy distortion figures most 2-way speakers measure. You're not.
https://en.wikipedia.org/wiki/Equal-loudness_contour
So yes, agreed that the prevailing opinion on not being able to hear those crazy HD figures in the bass is complete bs. (also from personal experience)
These are very clever trains of thought!
For normal two-way speakers, however, these considerations should be made for 40, 50 and 60Hz (20Hz is not realistic).

So much more sensitive that for a 90dB SPL 20Hz tone, if the 2nd HD is 10% that is 70dB SPL...
If Linkwitz chose this example, it was very manipulative, because these values refer to a basic sound level of 20dB@1kHz.
Hear very rarely with this volume, since a normal living room has a "base level" of 30-50dB of noise.

You should also note that the "equal-loudness contour" describes the perceived equal loudness of two individual tones (AIUI).
This means that a possible masking effects of e.g. a 40Hz basic tone and music/white noise as masker on the resulting HD2 (at 80Hz) is not included.

In addition, it should be noted that 90% of the instruments do not produce a pure tone, but produce overtones themselves, which "mask" harmonic distortions.

Nevertheless, if we make the same conclusions for the fundamental at 40Hz, but for realistic sound pressure levels of 80dB and 100dB (not so realistic :oops:) @1kHz, we get the following:
100dB@1kHz: HD2 equally loud at -11dB -> 28%, half as loud -21dB -> 9%
100dB@1kHz: HD3 equally loud at -13dB -> 22%, half as loud -23dB -> 7%

80dB@1kHz: HD2 equally loud at -10dB -> 31%, half as loud -20dB -> 10%
80dB@1kHz: HD3 equally loud at -15dB -> 18%, half as loud -25dB -> 6%

1593602942328.png
 

JustIntonation

Senior Member
Joined
Jun 20, 2018
Messages
480
Likes
293
These are very clever trains of thought!
For normal two-way speakers, however, these considerations should be made for 40, 50 and 60Hz (20Hz is not realistic).


If Linkwitz chose this example, it was very manipulative, because these values refer to a basic sound level of 20dB@1kHz.
Hear very rarely with this volume, since a normal living room has a "base level" of 30-50dB of noise.

You should also note that the "equal-loudness contour" describes the perceived equal loudness of two individual tones (AIUI).
This means that a possible masking effects of e.g. a 40Hz basic tone and music/white noise as masker on the resulting HD2 (at 80Hz) is not included.

In addition, it should be noted that 90% of the instruments do not produce a pure tone, but produce overtones themselves, which "mask" harmonic distortions.

Nevertheless, if we make the same conclusions for the fundamental at 40Hz, but for realistic sound pressure levels of 80dB and 100dB (not so realistic :oops:) @1kHz, we get the following:
100dB@1kHz: HD2 equally loud at -11dB -> 28%, half as loud -21dB -> 9%
100dB@1kHz: HD3 equally loud at -13dB -> 22%, half as loud -23dB -> 7%

80dB@1kHz: HD2 equally loud at -10dB -> 31%, half as loud -20dB -> 10%
80dB@1kHz: HD3 equally loud at -15dB -> 18%, half as loud -25dB -> 6%

View attachment 71483

I did not directly quote Linkwitz, what I wrote was from myself. But I read Linktwitz reference this train of thought once, didn't look it up again before I posted.

The equal loudness curve is not in any way about 1kHz.
We can easily talk about 20Hz or 40Hz or 50Hz at 90dB SPL. And 90dB SPL in the bass is not a crazy volume and subjectively much much less loud than 90dB at 1kHz (as you can see in the equal loudness contour).
The lines you've drawn in the equal loudness graph are not correctly labeled therefore.
If you want to see what a 40Hz tone at 80dB compares to then you simply look at 80dB at 40Hz and follow the red line from there on to 80Hz for H2 and look at what SPL that line is at 80Hz and that SPL is where 80Hz is percieved as equally loud as 40Hz at 80dB SPL.
The way I explained it in my previous post is correct and not manipulative.

edit: To write it more clearly. The red lines in the equal loudness contour represent the lines at which tones of various frequencies are perceived as the same loudness. So you can take any of those lines and simply follow it. So for instance 20Hz at 90dB SPL is perceived as the same loudness as 40Hz at 70dB SPL, 1kHz at 20dB SPL, 3kHz at 15dB SPL and 10kHz at 35dB SPL. (those points all lie on the same red line in the equal loudness contour graph)
 
Last edited:

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU
If you want to see what a 40Hz tone at 80dB compares to then you simply look at 80dB at 40Hz and follow the red line from there on to 80Hz for H2 and look at what SPL that line is at 80Hz and that SPL is where 80Hz is percieved as equally loud as 40Hz at 80dB SPL.
The way I explained it in my previous post is correct and not manipulative.
Your example with 90dB@20Hz applies to a perceived equal volume of 20 phone, which corresponds to a reference tone of 20dB@1kHz.
No one can perceive this sound pressure in a normal environment.

Therefore you have to choose a "reasonable" volume for the observation. Therefore, a medium sound level of 80dBA is certainly reasonable, which corresponds approximately to the 80phone curve.

UPDATE: Now understand what you mean, you only want to compare the sound pressure in the low bass, not the whole frequency band.
 
Last edited:

JustIntonation

Senior Member
Joined
Jun 20, 2018
Messages
480
Likes
293
Your example with 90dB@20Hz applies to a perceived equal volume of 20 phone, which corresponds to a reference tone of 20dB@1kHz.
No one can perceive this sound pressure in a normal environment.

Therefore you have to choose a "reasonable" volume for the observation. Therefore, a medium sound level of 80dBA is certainly reasonable, which corresponds approximately to the 80phone curve.

UPDATE: Now understand what you mean, you only want to compare the sound pressure in the low bass, not the whole frequency band.

Aah I understand what you mean now too :) (I was still editing my post before reading yours)
I do think 20 phone / 90dB SPL at 20Hz is perceivable, it is not masked by noise nobody has noise levels like that in the bass, and it is a good 16dB or so above detection treshold.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU
We can easily talk about 20Hz or 40Hz or 50Hz at 90dB SPL. And 90dB SPL in the bass is not a crazy volume and subjectively much much less loud than 90dB at 1kHz (as you can see in the equal loudness contour).
This is even more complicated than described in post#307, because now music plays at 80dB (dBA rated), but we look at the difference in equal-loudness-contour for 40Hz at e.g. 20phone - Aren't there effects like "remote masking" then?

Anyway, let's have a look at the values for 40Hz@83dB:
Fundamental 40Hz@83dB HD2 equally loud at -14dB -> 22%, half as loud -24dB -> 10%
Fundamental 40Hz@83dB HD3 equally loud at -22dB -> 8%, half as loud -32dB -> 2.5%

There is much more room for the typical 2-way loudspeaker than with the not so realistic 20Hz.
Staying below the same volume with HD2 and HD3 is barely possible for good two-way speakers with 6'' woofer (HD3 rather not).

However, from 8'' woofers upwards it should not be a problem. Unfortunately I do not have a better distortion measurement of a 2-way speaker with 8'' woofer at hand - but the tendency can be seen (about 86dB, with cheap 8'' woofer)
1593611515956.png


With a really large diaphragm area, you don't have to worry at all - 15'' woofer @85dB:
1593611810158.png
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
These are very clever trains of thought!
For normal two-way speakers, however, these considerations should be made for 40, 50 and 60Hz (20Hz is not realistic).


If Linkwitz chose this example, it was very manipulative, because these values refer to a basic sound level of 20dB@1kHz.
Hear very rarely with this volume, since a normal living room has a "base level" of 30-50dB of noise.

You should also note that the "equal-loudness contour" describes the perceived equal loudness of two individual tones (AIUI).
This means that a possible masking effects of e.g. a 40Hz basic tone and music/white noise as masker on the resulting HD2 (at 80Hz) is not included.

In addition, it should be noted that 90% of the instruments do not produce a pure tone, but produce overtones themselves, which "mask" harmonic distortions.

Nevertheless, if we make the same conclusions for the fundamental at 40Hz, but for realistic sound pressure levels of 80dB and 100dB (not so realistic :oops:) @1kHz, we get the following:
100dB@1kHz: HD2 equally loud at -11dB -> 28%, half as loud -21dB -> 9%
100dB@1kHz: HD3 equally loud at -13dB -> 22%, half as loud -23dB -> 7%

80dB@1kHz: HD2 equally loud at -10dB -> 31%, half as loud -20dB -> 10%
80dB@1kHz: HD3 equally loud at -15dB -> 18%, half as loud -25dB -> 6%

View attachment 71483

I spent about fifteen minutes reading this but eventually I gave up. If English is not your original language and you still are not comfortable with English, then please accept my apologies. But if it happens that English is your original language then I cannot help but ask you to make more effort toward writing complete, grammatically correct sentences that properly say what it is you want to so. I strongly suspect that you are making a valid technical point, but to my way of thinking, unless I understand it completely such that I would be able to rephrase what you are saying, then I don't really understand it at all. I gave it the old college try but gave up.
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,781
Location
Oxfordshire
In addition, it should be noted that 90% of the instruments do not produce a pure tone, but produce overtones themselves, which "mask" harmonic distortions.
I don't think the overtones mask distortion but rather the distortion changes the timbre of the instrument, which may be more evident to a musician used to playing it than a hifi enthusiast listening to it.
 

SECA_alan

Member
Joined
Feb 8, 2020
Messages
32
Likes
29
Thank you @amirm for this review, which caught my eye as a Revel F-208 user. It's very interesting to see the design philosophy examined even when it's not my own Loudspeaker. There is precious little brand recognition and representation for Harman (and especially Revel) here in the UK; I bought mine almost blind EOL from a stockist in Scotland (I'm on the south coast)! I remain delighted with them, BTW.

Thank you everyone else for the wonderful 'rabbit hole' this thread disappeared into. It was a most interesting read, broadening my understanding and encouraging me to learn more. Much appreciated!

Best regards,

Alan Brown
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
I did not directly quote Linkwitz, what I wrote was from myself. But I read Linktwitz reference this train of thought once, didn't look it up again before I posted.

The equal loudness curve is not in any way about 1kHz.
We can easily talk about 20Hz or 40Hz or 50Hz at 90dB SPL. And 90dB SPL in the bass is not a crazy volume and subjectively much much less loud than 90dB at 1kHz (as you can see in the equal loudness contour).
The lines you've drawn in the equal loudness graph are not correctly labeled therefore.
If you want to see what a 40Hz tone at 80dB compares to then you simply look at 80dB at 40Hz and follow the red line from there on to 80Hz for H2 and look at what SPL that line is at 80Hz and that SPL is where 80Hz is percieved as equally loud as 40Hz at 80dB SPL.
The way I explained it in my previous post is correct and not manipulative.

edit: To write it more clearly. The red lines in the equal loudness contour represent the lines at which tones of various frequencies are perceived as the same loudness. So you can take any of those lines and simply follow it. So for instance 20Hz at 90dB SPL is perceived as the same loudness as 40Hz at 70dB SPL, 1kHz at 20dB SPL, 3kHz at 15dB SPL and 10kHz at 35dB SPL. (those points all lie on the same red line in the equal loudness contour graph)

Yes.

I was not able to understand what ctrl was trying to say.

As for the interpretation of the Fletcher-Munson curves, the lowest curve corresponds to the threshold of audibility. At 1 kHz the threshold of audibility is 0 dB. This is not a coincidence. In essence we affix the label of 0 dB to the sound pressure at which a 1 kHz tone will be barely audible, and then we use this as a reference point for all the other values. All of the other values obtain from that same curve or any of the other curves are referenced to the pressure at which a 1 kHz tone will be barely audible. For example, looking at that same curve, the one at the bottom, a 20 Hz tone becomes audible when the sound pressure is about 70 dB greater than the sound pressure at which a 1 kHz tone becomes audible. The curves are spaced at 20 dB intervals, and while it doesn't really matter much how the spacing is determined, it is worth noting that it is not done on the basis of a 20 dB increase in perceived volume, because this notion, that it would make sense to say that one perceived loudness level is exactly 20 dB greater than another perceived volume level, is silly. It is thus apparent that the spacing is based on actual sound pressure, but again the more important point is that it doesn't much matter what the spacing is or how it is determined. A given curve shows, in a relative way using dB, the different actual sound pressure values that are required in order for all the points on that curve to be perceived as having the same loudness. Note that while it is not meaningful to say that one tone sounds 20 dB louder than another, it is perfectly meaningful to say that two different notes sound the same in loudness, or that one simply sounds louder than the other, or not as loud as the other.
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
I don't think the overtones mask distortion but rather the distortion changes the timbre of the instrument, which may be more evident to a musician used to playing it than a hifi enthusiast listening to it.

YES, YES, YES!!!!! The masking effect is an effect of the fundamental itself, and it is a stronger effect for harmonics that are close to the fundamental than for harmonics that are distant from the fundamental. And it is not specific to distortion per se. The first harmonic in a musical note made by most any musical instrument will be strongly masked by the fundamental. As you said so perfectly, the overtones in a musical note do not mask the distortion components. Rather, the distortion components bolster the overtones and alter the timbre. And certainly a musician accustomed to the sound of the instrument is going to be more sensitive to this effect compared to hifi enthusiasts that do not have the same familiarity with the sound of the live instrument. This is the reason that I have been saying for many, many years that the ideal listener for loudspeaker evaluations is not the trained listener in the sense of someone who has undergone training in a synthetic environment, but is rather the professional musician who has been trained over a very long period of time to be intimately familiar with how the instrument is supposed to sound. This is so patently obvious to me that I struggle to understand why so many people don't get it.
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,781
Location
Oxfordshire
YES, YES, YES!!!!! The masking effect is an effect of the fundamental itself, and it is a stronger effect for harmonics that are close to the fundamental than for harmonics that are distant from the fundamental. And it is not specific to distortion per se. The first harmonic in a musical note made by most any musical instrument will be strongly masked by the fundamental. As you said so perfectly, the overtones in a musical note do not mask the distortion components. Rather, the distortion components bolster the overtones and alter the timbre. And certainly a musician accustomed to the sound of the instrument is going to be more sensitive to this effect compared to hifi enthusiasts that do not have the same familiarity with the sound of the live instrument. This is the reason that I have been saying for many, many years that the ideal listener for loudspeaker evaluations is not the trained listener in the sense of someone who has undergone training in a synthetic environment, but is rather the professional musician who has been trained over a very long period of time to be intimately familiar with how the instrument is supposed to sound. This is so patently obvious to me that I struggle to understand why so many people don't get it.
By a country mile the most important aspect of a speaker to me, when I have been looking to change, is accuracy of instrumental timbre.
I don't even notice stereo image or spatiality or any of that guff if the instrumental timbre is unconvincing.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU
I don't think the overtones mask distortion but rather the distortion changes the timbre of the instrument, which may be more evident to a musician used to playing it than a hifi enthusiast listening to it.
Of course, depending on the strength of the harmonic distortions, both take place.
If the harmonic distortions are significantly less than the harmonics, this will hardly change the timbre.

If the harmonic distortions become too extreme, the harmonics are changed in their sound level, which naturally changes the timbre.


Here in the example a piano keystroke F5 is shown. The first harmonic is -15dB, the second harmonic -25dB below the fundamental.
If HD2 (1400Hz) is now at -20dB and HD3 (2100Hz) at -30dB, the harmonics are only marginally changed in their sound pressure.
The sound will not change noticeably, but the harmonic distortion is "masked" by the harmonics.

1593616952200.png
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
Your example with 90dB@20Hz applies to a perceived equal volume of 20 phone, which corresponds to a reference tone of 20dB@1kHz.
No one can perceive this sound pressure in a normal environment.

Therefore you have to choose a "reasonable" volume for the observation. Therefore, a medium sound level of 80dBA is certainly reasonable, which corresponds approximately to the 80phone curve.

UPDATE: Now understand what you mean, you only want to compare the sound pressure in the low bass, not the whole frequency band.

A 20 phon tone is a tone that, whatever its frequency may be, is perceived to be of the same loudness as a 1 kHz tone where actual sound pressure (SPL) is +20 dB relative to the actual sound pressure at which a 1 kHz tone will barely be audible.

So how loud exactly is this tone? If you play a 1 kHz tone and turn it up or down until you are just barely able to hear it, then increase the SPL by 20 dB, this is how loud it will be. Not especially loud, for sure, but still very audible, since after all the SPL is 20 dB greater than the SPL at which that same tone would be audible. What is the basis for the statement that no one can perceive this sound in a normal environment? It is obviously dependent on your definition of a "normal environment", and I cannot help but view the claim as being somewhat circular, since on the face of it, it seems that your ad hoc definition of a normal environment is one where a 20 phon sound would not be audible. I would not hazard a guess at how many people listen to their music in very quiet environments, but I expect that it is not insignificant.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,088
Location
.de, DE, DEU
What is the basis for the statement that no one can perceive this sound in a normal environment?
The background noise in a normal living room is given with 30-50dB.
1593618236210.png

http://www.sengpielaudio.com/TableOfSoundPressureLevels.htm


UPDATE: At my seat in front of the screen I just measured 36dBA (I live in the city, near a busy street, house has soundproof windows)

UPDATE2: But this will not be enough to completely mask a 1kHz@20dB sine wave. So you are right, it will probably still be perceivable.
 
Last edited:

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
This is even more complicated than described in post#307, because now music plays at 80dB (dBA rated), but we look at the difference in equal-loudness-contour for 40Hz at e.g. 20phone - Aren't there effects like "remote masking" then?

Anyway, let's have a look at the values for 40Hz@83dB:
Fundamental 40Hz@83dB HD2 equally loud at -14dB -> 22%, half as loud -24dB -> 10%
Fundamental 40Hz@83dB HD3 equally loud at -22dB -> 8%, half as loud -32dB -> 2.5%

There is much more room for the typical 2-way loudspeaker than with the not so realistic 20Hz.
Staying below the same volume with HD2 and HD3 is barely possible for good two-way speakers with 6'' woofer (HD3 rather not).

However, from 8'' woofers upwards it should not be a problem. Unfortunately I do not have a better distortion measurement of a 2-way speaker with 8'' woofer at hand - but the tendency can be seen (about 86dB, with cheap 8'' woofer)
View attachment 71489

With a really large diaphragm area, you don't have to worry at all - 15'' woofer @85dB:
View attachment 71490

I really, really, really, really, REALLY wish that people would stop talking about one sound being "half as loud" as another. I realize that there is a popular belief that a 10 dB difference in SPL corresponds to a doubling or halving of perceived volume level, and this notion is the basis for what you wrote. But it just doesn't make a whit of good sense, no matter how popular the notion may be. It is meaningful to say that two tones sound equally loud, or to say that one sounds louder than the other. But it does NOT make good sense to say that the perceived loudness of one sound is twice greater than the perceived loudness of another sound. There isn't anything in the Fletcher-Munson curves that corresponds to this notion, or assumes it, or whatever. In fact when you study these curves and come to understand what they really mean and what Fletcher & Munson did, you cannot avoid the realization that they eschewed the notion that it would make sense to try and force-fit different levels of perceived volume into some quantitative scheme. They did not do anything of this sort. Any single one of the curves is concerned purely with sounds that are EQUAL in perceived volume, and the only quantitative relationship between any two curves is with respect to SPL, not perceived volume. For example, for the 2nd curve from the bottom, a 1 kHz tone is first played at +20 dB actual SPL relative to the actual SPL at which a 1 kHz tone is just barely audible. The perceived loudness of that 1 kHz tone at that actual SPL is then used to obtain the other points falling on the same curve. For the 3rd curve from the bottom, a 1 kHz tone is played at +40 dB actual SPL relative to the actual SPL at which a 1 kHz tone is just barely audible. Thus, the spacing between the curves has no direct relationship to quantitative intervals in perceived volume. The notion that differences in perceived volume are quantifiable in a meaningful way, e.g., that one sound is "twice as loud" as another in terms of perceived volume, is simply not a meaningful notion. Regardless of the unfortunate fact that you encounter this notion fairly often, it is a meaningless, silly notion.
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,781
Location
Oxfordshire
The sound will not change noticeably, but the harmonic distortion is "masked" by the harmonics.
I disagree. Adding a tone at -20dB to the existing one at -15dB, plus a bit to all the other harmonics too, including the harmonic distortion of the harmonics themselves will change the timbre of the piano, not much, maybe, but it will change the timbre since the timbre of the instrument is defined by the number and intensity of the harmonics.
 
Top Bottom