tuga
Major Contributor
KLIPPEL LIVE Series1_Part9 - Intermodulation Distortion - Music is More than a Single Tone
-
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!
Neumann KH420 Review (Studio Monitor)
- Thread starter amirm
- Start date
Thanks tuga !
This was very interesting reading .
IMD distortion seems to be very audible according to this paper.
Last edited:
KSTR
Major Contributor
This is not correct.
IMD, like HD, is a result of a true nonlinearity and this shows as additional tones in the spectra which are not present in the stimulus. It occurs in all speaker/driver types and is not strictly related to excursion, oftentimes it is more related to current. Lots of IMD really corrupt music playback.
Doppler "distortion" is a linear phenomenon (does not produce new tones, rather the original ones are slightly modulated in frequency/phase and sometimes even level) and is only a function of excursion, a lower tone modulating the source position or geometry for a higher tone (and itself, in extreme cases). It happens in all speakers/drivers, even theoretically ideal ones, once there enough excursion but it is a rather benign type of distortion, the sound being slightly "rougher" than without but not cluttered with IMD components hash.
Ideal test signal for IMD is a twin-tone, the first tone at the lower cutoff of the used range for the driver and the second at the higher cutoff. Other drivers switched off, if any possible.
Maybe I should have said only occurs to a significant degree in coaxial type designs. Regardless it doesn't really seem benign to me in the few cases I've seen it measured in 2-way designs like this one from Erin's Q100 review. It does seem basically irrelevant in 3-way designs.
KSTR
Major Contributor
Care to share a link etc for info on this? I've never heard or seen any claim like this.
How would you combine separate 60Hz and 7kHz HD measurements for a 60Hz+7kHz IMD? Genuinely interested...
With a cellin at 2.5m, i have celling reflections with a listening distance at 2.4 m. Greater is the distance, greater in the dispersion cone.
With the floor, i use japanese tatamis so i reduce the level fo reflection.
it means what is write : BS marketing
KSTR
Major Contributor
I'd call this geometry distortion.Maybe I should have said only occurs to a significant degree in coaxial type designs. Regardless it doesn't really seem benign to me in the few cases I've seen it measured in 2-way designs like this one from Erin's Q100 review. It does seem basically irrelevant in 3-way designs.
Not to be confused with what Purifi calls it SD modulation: https://purifi-audio.com/2021/10/14/some-speaker-problems-that-needed-solving/
The main
Sanctus - in this picture you see differences in amplitude of about 10 dB at 7,5 KHz .
If you are playing bass and cymbals at the same time, even if the volume where lower and the difference between the cone placement where half - you still have a difference at 7,5 KHz at maybe 5 dB modulation which is very audible in real music. It will cause listening fatigue.
If you are playing bass and cymbals at the same time, even if the volume where lower and the difference between the cone placement where half - you still have a difference at 7,5 KHz at maybe 5 dB modulation which is very audible in real music. It will cause listening fatigue.
Last edited:
But the horizontal beam is wider than the vertical one, isn't it?
OP
- Thread Starter
- #590
Did you read this summary slide?
What I have highlighted in red is exactly what I post above. The context of Klippel presentation here is for a designer who is diagnosing specific problems and can choose those frequencies appropriately. That notion cannot be applied to a universal set of dual-tones used to compare speakers. What brings out IMD in one driver/speaker is NOT what brings it out in another. This is a showstopper.
Don't confuse what is a design tool with a benchmark metric across multiple products.
yes like all speakers well designed.
It's very clearly called doppler distortion in the Genelec Ones white paper, so I dunno. Maybe different sources are referring to different things when they say doppler distortion.
tuga
Major Contributor
My post was not a reply to yours. When I clicked Post Reply yours was already there.
Like coax genelecs you mean?
yes in orange color.Like coax genelecs you mean?
This is absolutely the correct way of going about things for both you and Amir given the intentions of provided measurements.
Even for those who's technically apt enough to understand such measurements and caveats there's a questionable amount of value such measurements can provide.
Unless there's a way to make such measurements in a reliable, reproducible and transferable way between speakers it might only provide noise.
Since you're thinking about it, how do you imagine it could work on a Klippel in a small room? Bandwidth-limited IMD?
Might be valuable for two-ways at least?
KSTR
Major Contributor
Yes, nomenclature may not be all that clear.
Doppler distortion is usually referred to as a form of phase modulation (amplitude stays unaffected) and I'm quite sure this is not the case when the geometry (the acoustic loading) is modified. Doppler distortion has been successfully eliminated by a dynamic phase pre-distortion in active 2-way speakers with motional feedback (which keeps regular IMD low enough that Doppler distortion becomes any relevant): https://www.silbersand.de/Doppler.html
Silbersand does it in the analog domain but it has been realized in the digital domain, too.
Alas, for me the personal threshold is lower than -53% with the Klippel self-test, sigh!
But it depends on the frequency range a lot. Interestingly the most vulnerable regions seems to be the lower mids. In the higer mids, say above 1,5kHz the shrill of the rubbing tones alone is so annoying, that additional IM doesn't make a difference for me. => Go 4-way!
This is only anecdotal. But feel encouraged to do it yourself with yourselves. It might help to understand the nature of IM also.
Doppler is IM, as it can be transformed mathematically. It is said that the special phase relation in Doppler makes it less bad. For headphones that seems to be true (!?). But with speakers in room the phase relation is destroyed by crossover and mostly by room reflections. No way out here. Doppler easily, independently from the driver's quality, will reach up to 1% with only 1..2..3 mm of cone excursion in a 2-way.
HarmonicTHD
Major Contributor
- Joined
- Mar 18, 2022
- Messages
- 3,326
- Likes
- 4,838
Interesting. thx.
Wouldn’t that effect also show up during the usual sweeps in the crossover region btw. mid/tweeter when both are active at the same time? Yet, I don’t remember significant distortion for the UniQs at crossover, might be wrong though. Or does the effect, to become significant, require a higher difference in the two frequencies so the mid shows significant cone movements?
Wouldn’t it also be great to ask KEF and or Genelec. Given the time they have been using these coaxials, I would guess they have looked into it, so they wouldn’t bank on a possible problematic technology down the road. Might be wrong here too. But asking wouldn’t hurt ? (Kefs head of development is on this forum isn’t he).
thx.
Last edited:
In Genelecs white papers of ”the one” they mention the theoretical dissadvantage with coaxials because of this, but they didnt write that there was any compromise in sound quality. The crossover point for ”the one ” is at 320 Hz for 8351 with a step active filter ( probably 24 dB/oct or more ) so maybe the problem goes away then .
Similar threads
- Poll
