• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Harmonic and Intermodulation Distortion

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
2,803
Likes
4,758
Location
North-East
IMD is what THD looks like with multiple tones. I don't know why they are treated as being intrinsically different, when all that's changed is the input signal.

Non-linearity of the system transfer function is the underlying, common cause. THD and IMD is what it looks like when passing one or multiple tones, respectively, through the same non-linearity.
 

Holmz

Senior Member
Joined
Oct 3, 2021
Messages
476
Likes
258
Location
Australia
IMD is what THD looks like with multiple tones. I don't know why they are treated as being intrinsically different, when all that's changed is the input signal.

Often because “things” can have harmonics… like violins, door knocking sounds, etc.

IMD is often thought of as products between different things.

And harmonics can be more pleasing than IMD.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
2,262
Likes
2,410
Location
Pacific Northwest
Often because “things” can have harmonics… like violins, door knocking sounds, etc.
IMD is often thought of as products between different things. ...
True, but most "things" are not a single frequency, but a bunch of frequencies, a fundamental and a bunch of harmonics. Those harmonics are what makes a chainsaw revving at 622 Hz sound different from a clarinet playing E flat. And when that signal passes through an amp, those harmonics of even a single "thing" trigger IMD difference tones.
 

Holmz

Senior Member
Joined
Oct 3, 2021
Messages
476
Likes
258
Location
Australia
True, but most "things" are not a single frequency, but a bunch of frequencies, a fundamental and a bunch of harmonics. Those harmonics are what makes a chainsaw revving at 622 Hz sound different from a clarinet playing E flat. And when that signal passes through an amp, those harmonics of even a single "thing" trigger IMD difference tones.

That reminds we I set my TIG welder up at a pulse frequency of 440 Hz.
My welds are not grade “A” but at least it has the right sound for an “A” :cool:

If we know what causes the distortions, then one could, in theory, ameliorate them.
That seems to me to be is easier to do for a speaker than for an amplifier.
Whether it sounds better or worse, is subjective; but having less distortion would be higher fidelity.
 

thewas

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,626
Likes
7,653

Holmz

Senior Member
Joined
Oct 3, 2021
Messages
476
Likes
258
Location
Australia
You mean by redesigning the driver or by manipulating the electric signal send to them?
The second usually only works well in the lower frequency range when the behaviour is purely pistonic https://www.audiosciencereview.com/...lly-lower-distortion.11216/page-3#post-321065

The later case.
If the BL curve was flat then the motor is linear.
Or if one knows the BL, then one could make the “realised position“ end up correct.

The cone break up (and pistonics), I thought… was largely solved with those Carbon Kentos?
 

thewas

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,626
Likes
7,653
The later case.
If the BL curve was flat then the motor is linear.
Or if one knows the BL, then one could make the “realised position“ end up correct.
Or you can use motional feedback sensors and control like some do, but again this all works in the purely pistonic region.

The cone break up (and pistonics), I thought… was largely solved with those Carbon Kentos?
Cone break up is never fully solved, just some materials can bring it to higher frequencies with other problems though like for example higher Q of break up.
 
OP
DonH56

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
5,787
Likes
10,464
Location
Monument, CO
Music and other signals can have any number of frequency components, harmonically related and not, that is not an issue. What we do not want is our equipment adding additional distortion products of their own.
 

pozz

Data Ordinator
Forum Donor
Editor
Joined
May 21, 2019
Messages
3,984
Likes
6,384
Often because “things” can have harmonics… like violins, door knocking sounds, etc.

IMD is often thought of as products between different things.

And harmonics can be more pleasing than IMD.
Pleasing? THD and IMD are separate measurements meant to characterize the same nonlinear system. The only thing that changes is the input signal!
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
2,262
Likes
2,410
Location
Pacific Northwest
Pleasing? THD and IMD are separate measurements meant to characterize the same nonlinear system. The only thing that changes is the input signal!
Separate measurements because their audibility is different. If you play tones at 9 kHz and 10 kHz you won't hear any HD even if it exists, because they're outside the range of hearing. But you will hear IMD if it is there.
 

pozz

Data Ordinator
Forum Donor
Editor
Joined
May 21, 2019
Messages
3,984
Likes
6,384
Separate measurements because their audibility is different. If you play tones at 9 kHz and 10 kHz you won't hear any HD even if it exists, because they're outside the range of hearing. But you will hear IMD if it is there.
I think you've changed topics, maybe without noticing. @Holmz said "harmonics can be more pleasing than IMD".

The issue is not the audibility of nonlinear distortion for particular input signals. What you said is true. But what I'm drawing attention to is a more fundamental misunderstanding. Gear will produce nonlinearity no matter what you feed into it. How can you regard one battery of tests, producing HD, as pleasing and then take same device, same nonlinearity, but under different a test call it displeasing? The same problem is at the root.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
2,262
Likes
2,410
Location
Pacific Northwest
I think you've changed topics, maybe without noticing. @Holmz said "harmonics can be more pleasing than IMD".

The issue is not the audibility of nonlinear distortion for particular input signals. What you said is true. But what I'm drawing attention to is a more fundamental misunderstanding. Gear will produce nonlinearity no matter what you feed into it. How can you regard one battery of tests, producing HD, as pleasing and then take same device, same nonlinearity, but under different a test call it displeasing? The same problem is at the root.
Back to that point: One manifestation of the nonlinearity, HD, at least resembles the harmonic structure of every sound we hear. The other manifestation, IM, doesn't follow those frequency ratio patterns and is less "natural".
 
  • Like
Reactions: Ata

pozz

Data Ordinator
Forum Donor
Editor
Joined
May 21, 2019
Messages
3,984
Likes
6,384
Back to that point: One manifestation of the nonlinearity, HD, at least resembles the harmonic structure of every sound we hear. The other manifestation, IM, doesn't follow those frequency ratio patterns and is less "natural".
Only some sounds have harmonic structures. Most natural sounds are noise, and animal sounds (not just those of mammals—but of bugs, fish and so on) produce structured frequency relationships based on biomechanical resonances, which aren't necessary harmonic. Clicks, screeches, purrs, barks, whatever. The envelope only allows certain spectral features to be heard. Overwhelmingly, structured frequency relationships, of which harmonics are one type, occur only for sustained sounds.

Our ear has 7% IMD according to Mead Killion of Etymotic. Complex, dense sum/difference sounds are inherent to the hearing system. You get it in any resonant structure.

There is no equating harmonic structures to what happens with nonlinearity. You don't get extra "music" with HD, and HD is not necessarily pleasant. That pleasure is as context and signal dependent as audibility.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
2,262
Likes
2,410
Location
Pacific Northwest
... Our ear has 7% IMD according to Mead Killion of Etymotic. Complex, dense sum/difference sounds are inherent to the hearing system. You get it in any resonant structure. ...
Yes, some IMD is built into our hearing. As a flute player, I learned this as a kid. Flute has a fairly clean/simple harmonic structure especially in the top registers - the tone is closer to a pure sin wave than most other instruments. We play flute duet etudes, where the notes are carefully selected so they they sound harmonious, while also the difference tones become part of the music itself. But it only sounds right if each player has perfect intonation. If your intonation is dead-on, the IM tones created by your ear make a nice harmonic buzz. If you're even slightly off, they disappear or sound horribly dissonant.

Musicians also use this (the build-in IMD of our ears) for getting in tune. If you're slightly off, you hear "beats" or amplitude pulses that are difference tones. That is, if you're off by 4 Hz the sound pulses 4 times per second.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,085
Likes
1,382
Musicians also use this (the build-in IMD of our ears) for getting in tune. If you're slightly off, you hear "beats" or amplitude pulses that are difference tones. That is, if you're off by 4 Hz the sound pulses 4 times per second.

Beat frequencies (as described here) are due to interference effects.
Where does a dependency on human auditory IMD enter?
 

pozz

Data Ordinator
Forum Donor
Editor
Joined
May 21, 2019
Messages
3,984
Likes
6,384
Beat frequencies (as described here) are due to interference effects.
Where does a dependency on human auditory IMD enter?
Seems fair to call combination tones a form of IMD. Two real notes and third synthesized one.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,085
Likes
1,382
Seems fair to call combination tones a form of IMD.

It isn't, really. But certainly not worth squabbling about.
Also, I was just observing that beat frequencies exist irrespective of human ears and their innate IMD.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
2,262
Likes
2,410
Location
Pacific Northwest
It isn't, really. But certainly not worth squabbling about.
Also, I was just observing that beat frequencies exist irrespective of human ears and their innate IMD.
That's true and an important difference. I thought beats were related to the IM of our hearing, but I did a test, superimposed 2 frequencies 500 Hz and 502 Hz, and saw the beats. So beats are amplitude pulses that result from summing waves together, not created by our hearing like difference tones are. Or put differently, if you made recording, the beats would be in the recording itself but the difference tones would not be, they are a perceptual anomaly that happens when you listen to it.
 

pozz

Data Ordinator
Forum Donor
Editor
Joined
May 21, 2019
Messages
3,984
Likes
6,384
Both phenomena exist. Beats happen physically and in our hearing.
 
Top Bottom