• 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.

Evidence-based Speaker Designs

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
2,589
Likes
2,031
Almost. If you need to play the signal from a file generated by REW open the generator and select the same signal so REW knows what to calculate. The generator doesn't need to be running, just showing.
That's a neat trick, thank you!

What distance and what SPL would you recommend when measuring IMD in a room?
 

JohnPM

Technical Expert
Technical Expert
Joined
Apr 9, 2018
Messages
159
Likes
450
Location
UK
What distance and what SPL would you recommend when measuring IMD in a room?
Close to the speaker, e.g. 1 metre, preferably with the speaker in the middle of the room. Level depends on what playback level you want to see the figures for, since distortion will increase with level.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
2,589
Likes
2,031
Close to the speaker, e.g. 1 metre, preferably with the speaker in the middle of the room. Level depends on what playback level you want to see the figures for, since distortion will increase with level.
Ok, I understand. What would you say is a good value for high quality speakers? Would it be, for example, <0.5% north of 200Hz?
 

tuga

Major Contributor
Forum Donor
Joined
Feb 5, 2020
Messages
1,473
Likes
960
Location
Oxfordshire
I'm not finished, point to another science where one single study has been held up as fact? That's not how science works if you want to look at science you can pretty much prove anything or try to prove anything science is a bunch of people trying to prove things it's a battle but eventually there's a consensus.

it's not and never has been one source of information held above all else without any challenge.


The harman research is a factor in audio science but it's not audio science.

Claiming otherwise is basically to form a religion around harming research.
 

Jon AA

Active Member
Forum Donor
Joined
Feb 5, 2020
Messages
287
Likes
556
Location
Seattle Area
Out of interest though, could you be more specific about the kind of felt underlay you're talking about?
This from his book:

CarpetAbsorbption.jpg



Obviously, clipped-pile carpet by itself will have most of its effect in the mid-high frequencies. But with the underlay, you do get some usable absorption in the lower-mid/upper bass frequencies.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
2,589
Likes
2,031
I just noticed that the speaker is only 0.32m deep. This means that the driver offset is probably 0.2m instead of the previously assumed 0.28m.

If the simulation is corrected with this, the result might reflect your room measurements a bit better.
View attachment 64787
@ctrl , @andreasmaaan , I found the uncorrected measurement, this is where the dip is:

Capture.JPG
 

andreasmaaan

Major Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
4,418
Likes
4,215
The Harman search was all about selling speakers into home environments.

That's the dollar they were chasing, not ultimate performance if that was the case the enclosures would have been different and it all would have been different.
I reckon you're right as to their intentions @Thomas savage, but then I also think that if that's the case, they arguably went about it the wrong way, as they did the tests in quite nicely set-up, well-treated rooms, with speakers placed in pretty optimal positions (at least for conventional designs).

If they were looking at speakers that were preferred in typical home environments, wouldn't it have made more sense to set them up in more difficult positions, in less optimal test rooms? More like what Bruel & Kjaer did back in the 70s?

Given what Harman actually did, whether they intended it or not, I think their findings hold weight for people who have well treated, well set-up rooms, i.e. people who have a serious listening room, a home theatre, etc. (That's not to say necessarily they don't also hold for less optimal rooms, too.)

EDIT: removed pics after realising they were of the of the wrong room :facepalm:
 
Last edited:

ctrl

Active Member
Joined
Jan 24, 2020
Messages
236
Likes
796
Location
DE
@ctrl , @andreasmaaan , I found the uncorrected measurement, this is where the dip is:
You really want to know! ;)

My guess would be that the dip at 750Hz is caused by cancellation due to phase shifting of the rearward offset chassis - that fits the simulation pretty well.

If your measurement without EQ is again valid for 4m listening distance, the dip at 400Hz might be the floor bounce. To be really sure, you would have to change the mirror point of the floor bounce - either with absorbing material or a reflecting surface (e.g. a wooden board).
If the dip disappears or shifts, you have the proof.
1590223591727.png
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
2,589
Likes
2,031
You really want to know! ;)

My guess would be that the dip at 750Hz is caused by cancellation due to phase shifting of the rearward offset chassis - that fits the simulation pretty well.

If your measurement without EQ is again valid for 4m listening distance, the dip at 400Hz might be the floor bounce. To be really sure, you would have to change the mirror point of the floor bounce - either with absorbing material or a reflecting surface (e.g. a wooden board).
If the dip disappears or shifts, you have the proof.
View attachment 64927
Yes, this is the same MMM measurement without EQ done at LP, so 4m.
It may be indeed floor bounce as right channel (blue line) has thick carpet at the mirror point while left channel doesn't, hence the difference.

Once again I thank you for your effort!

I will try to do 40:600Hz IMD distortion from small distance as you suggested. Is there any other frequency pair that is interesting for measuring IMD? Maybe 400:4000Hz?
 

ctrl

Active Member
Joined
Jan 24, 2020
Messages
236
Likes
796
Location
DE
I will try to do 40:600Hz IMD distortion from small distance as you suggested. Is there any other frequency pair that is interesting for measuring IMD? Maybe 400:4000Hz?
First of all, I am not an expert in IMD measurements.
With my DIY speakers I measure IMD with a multitone signal at different sound pressures (85, 95, 105 dB) to see if and at what sound level something unusual happens.
In the current beta version of REW, the signal generator has probably been extended considerably, so that it can generate different multitone signals.
http://www.roomeqwizard.com/betahelp/help_en-GB/html/siggen.html#multitone
So REW should be able to provide the complete range of IMD measurements in the next major version.

With 2-way speakers I also use two-tone measurements to see what happens e.g. with high excursion of the chassis in the area of the chassis surround resonance - depending on the chassis this could happen with e.g. 40Hz/1200Hz excitation.

To find abnormalities I usually use impedance measurement (which the normal hifi enthusiast does not measure) and THD measurements at different sound pressures. If a THD at 800Hz shows an increased HD3, then it's a good idea to look at it with a larger excursion of the chassis and take a two-tone measurement with e.g. 40/800Hz.
 
Top Bottom