Doppler distortion doesn't make sense (Edit: It actually does!)

[gnarly]

Here's a snip from a Paul Klipsch paper http://www.readresearch.co.uk/loudspeaker_papers/klipsch_modulation_distortion_article_1.pdf

I really like the term "mud index"....i wish it had caught hold

 

[fpitas]

Here's a snip from a Paul Klipsch paper http://www.readresearch.co.uk/loudspeaker_papers/klipsch_modulation_distortion_article_1.pdf

I really like the term "mud index"....i wish it had caught hold

View attachment 280833

Really makes one appreciate why you almost have to port boxes for little woofers.

 

[DonH56]

Paul Klipsch argued for less excursion, achieving that by using horn-loaded drivers. Others used multiple parallel drivers again to reduce the excursion of any single driver. In addition to various AES articles, there was an article in Audio magazine about FMD (FM distortion) that presented a speaker using an array of drivers to minimize the effect.

Edit per @gnarly: The usual solution today is multiple drivers covering different frequency bands.

A reminder that, as mentioned by myself and others (and per the equations presented by @NTK), this distortion is the result of a moving source. It is not a simple addition (summation) of signals; the signals actually modulate each other to create new frequencies (distortion).

 

[gnarly]

Paul Klipsch argued for less excursion, achieving that by using horn-loaded drivers. Others used multiple parallel drivers again to reduce the excursion of any single driver. In addition to various AES articles, there was an article in Audio magazine about FMD (FM distortion) that presented a speaker using an array of drivers to minimize the effect.

A reminder that, as mentioned by myself and others (and per the equations presented by @NTK), this distortion is the result of a moving source. It is not a simple addition (summation) of signals; the signals actually modulate each other to create new frequencies (distortion).

Yeppers.

I think less excursion is one method.
The other, perhaps more common, is less bandwidth covered per driver section. Good ole multi-ways.

Both so simple in concept; yet both so difficult to measure in a way that correlates with the universe of widely varying perceptions/preferences , huh?
Personally, I'm currently of the opinion that some form of modulation distortion ranks higher than harmonic distortion as a suspect for reducing SQ.

 

[fpitas]

Yeppers.

I think less excursion is one method.
The other, perhaps more common, is less bandwidth covered per driver section. Good ole multi-ways.

Both so simple in concept; yet both so difficult to measure in a way that correlates with the universe of widely varying perceptions/preferences , huh?
Personally, I'm currently of the opinion that some form of modulation distortion ranks higher than harmonic distortion as a suspect for reducing SQ.

It's kind of a skeleton in the closet for the small speaker crowd.

 

[gnarly]

It's kind of a skeleton in the closet for the small speaker crowd.

Yeah, a T-Rex skeleton Lol

btw, agree with your comment about ports for woofers.
I've come to view ports as essentially adding another driver section, to relieve the woofer of excursion demands down low.
(much akin to a passive radiator i guess)

 

[fpitas]

Yeah, a T-Rex skeleton Lol

btw, agree with your comment about ports for woofers.
I've come to view ports as essentially adding another driver section, to relieve the woofer of excursion demands down low.
(much akin to a passive radiator i guess)

Yes, if you just have to keep excursion down, ported is the only good solution other than a big bass horn ala VOTT. I designed my horn system as a three-way to avoid that and have a sealed woofer.

 

[DonH56]

I think less excursion is one method.
The other, perhaps more common, is less bandwidth covered per driver section. Good ole multi-ways.

Yes, I should have mentioned that. The obvious and most-used solution today.

Both so simple in concept; yet both so difficult to measure in a way that correlates with the universe of widely varying perceptions/preferences , huh?
Personally, I'm currently of the opinion that some form of modulation distortion ranks higher than harmonic distortion as a suspect for reducing SQ.

What data I have read points to harmonic distortion (and thus IMD) being much more significant than FMD (Doppler distortion). Do you have, or have a reference for, FMD levels compared to HD/IMD in modern speaker designs? I do not know, but since the references I have are very old (1960's through the early 1980's) I sort of thought modern multi-way designs had pretty much solved the FMD problem, or at least sufficiently reduced it below the HD/IMD levels. I am not a speaker designer so it is not something I track or know with any certainty.

Curious - Don

 

[gnarly]

What data I have read points to harmonic distortion (and thus IMD) being much more significant than FMD (Doppler distortion). Do you have, or have a reference for, FMD levels compared to HD/IMD in modern speaker designs? I do not know, but since the references I have are very old (1960's through the early 1980's) I sort of thought modern multi-way designs had pretty much solved the FMD problem, or at least sufficiently reduced it below the HD/IMD levels. I am not a speaker designer so it is not something I track or know with any certainty.

Curious - Don

Same here.
I've never really put much stock into FMD, but that is as much as from a point of ignorance about it, as not.

And I almost feel like a babe in the woods trying to learn how/if IMD measurements can be shown to correlate with what I sense.
Here's a post from another thread as an example of some of the measurements I've been trying.

Loudspeaker resolution measurement.

I doubt very much anyone would run a CEA-2010 test indoors. I think i remember, the specs call for measurements made at 2m, ground plane outdoors. JRS put a link in #89, explaining CEA-2010......well worth a read. @JRS, very cool thanks. If I understand correctly, the "waterfall" fall shown in...

www.audiosciencereview.com

 

[fpitas]

FWIW, listen for yourself:

Doppler distortion vs IMD? - PURIFI

Let’s get this out of the way first: Doppler distortion is a form of IMD. This is interesting because IMD in general tends to get little interest on loudspeaker forums, whereas Doppler distortion is often made a big deal of. Any distortion process will cause both harmonic and intermodulation...

purifi-audio.com

The phase modulation sounds like crap to me, but it is taken with extreme excursion, +/-12.5mm.

 

[gnarly]

FWIW, listen for yourself:

Doppler distortion vs IMD? - PURIFI

Let’s get this out of the way first: Doppler distortion is a form of IMD. This is interesting because IMD in general tends to get little interest on loudspeaker forums, whereas Doppler distortion is often made a big deal of. Any distortion process will cause both harmonic and intermodulation...

purifi-audio.com

The phase modulation sounds like crap to me, but it is taken with extreme excursion, +/-12.5mm.

Yes, I've seen that article and compared the two.
To my ears, the PureAM modulation is way, way worse sounding, than the PureFM modulation which i thought corresponds with Doppler distortion. No?

Doesn't the AM sound worse to you?
Or are you saying the FM sounds like crap too?
Confused i am

 

[fpitas]

Yes, I've seen that article and compared the two.
To my ears, the PureAM modulation is way, way worse sounding, than the PureFM modulation which i thought corresponds with Doppler distortion. No?

Doesn't the AM sound worse to you?
Or are you saying the FM sounds like crap too?
Confused i am

The FM (actually, it's PM) sounds just as bad to me. Both are rather intolerable. Maybe different brains process them differently.

 

[fpitas]

Rod makes the case for it mainly being phase modulation:

Doppler Distortion in loudspeakers

Doppler distortion in loudspeakers - Real or Imaginary?

sound-au.com

 

[Cbdb2]

What data I have read points to harmonic distortion (and thus IMD) being much more significant than FMD (Doppler distortion). Do you have, or have a reference for, FMD levels compared to HD/IMD in modern speaker designs? I do not know, but since the references I have are very old (1960's through the early 1980's) I sort of thought modern multi-way designs had pretty much solved the FMD problem, or at least sufficiently reduced it below the HD/IMD levels. I am not a speaker designer so it is not something I track or know with any certainty.

Curious - Don

Doppler distortion is IMD. Its not harmonicaly related to the signal so its more obvious and detrimental. Most music already has lots of harmonics so much of the HD gets burried in the music. IMD adds freqs between the music harmonics.

 

[Cbdb2]

Just something to think about. Relating a fft with the time domain. Something the freq spectrum dosnt show is whats physically happening. That is a pitch shift. Play a 1khz tone and move the speaker back and forth a couple feet at 5hz. What you hear is a tone vary from 995hz to 1005hz (maybe not exactly but you get the point) vibrato. Same as doing it on a synth. In real time the fft would move the tone up and down 5hz. You don't hear side bands. At what point do you start hearing the sidebands?
You can look at AM the same way. If I grab a volume pot and turn it up and down fast what do you hear, the volume changing or sidebands? In real time the fft would change the level of the tone. So take the fft without the phase info with a grain of salt, as shown above where the ffts of AM and FM look identical.
At some point we start hearing sidebands, the physics doesn't change at higher modulation freqs, so it must be our hearing.

 

[DonH56]

Doppler distortion is IMD. Its not harmonicaly related to the signal so its more obvious and detrimental. Most music already has lots of harmonics so much of the HD gets burried in the music. IMD adds freqs between the music harmonics.

The mechanism and derivation for FMD (Doppler distortion; I generally do not use that term) is different than IMD so I treat IMD and FMD separately. I do understand what they are, but in my world intermodulation and frequency modulation are different things. YMMV. I agree both IMD and FMD generate tones not harmonically related to the signal and thus sound worse than HD.

Edit: With a little more time to spend... HD and IMD are generated by nonlinearities in the transfer function, e.g. https://www.audiosciencereview.com/...armonic-and-intermodulation-distortion.25436/ (the derivation is in post #31). FMD or Doppler distortion is caused by changing physical placement, basically a dynamic shift in source position that translates to a shift in frequency and (in general) phase that changes the signal frequency perceived (seen, heard) by the observer (listener). The shift can be defined as variation in time or velocity (e.g. radar equations). For loudspeakers, the driver (cone, panel) can itself be perfectly linear but whenever more than one frequency is present the amplitude modulation (excursion) of the cone means doppler effect and thus frequency modulation distortion (FMD) is occurring.

FWIWFM - Don