Measurements: When, What, How, Why

[D OB G]

Zaph:

All the answers are in the measurements. And I mean ALL the answers. Some people don't know how to interpet the measurements. Some don't want to know all the answers.

Geddes:

These things are all measureable. Just because you don't know how to do it does not mean that it cannot be done.

Has anyone systematically addressed:

What to measure
How to measure it
How to interpret the measurement
What does it correlate with?
What matters anyway?

David

[graaf]

Originally Posted by D OB G

Has anyone systematically addressed:

anyone? why ask anyone?

when Zaph and Dr Geddes have adressed!

therefore they make the best loudspeakers

don't think, don't bother, just buy

best,
graaf

[D OB G]

Very good graaf !!

Zaph and Geddes have indeed alluded to these things ...

[wintermute]

hehehe I think it very interesting that the two quotes are from two people with IMO VERY different views on what is important

If I were to go on zaphs tests of my chosen midbass drivers (the morel MW144), I should have thrown them out before I started! However the fact that they are used in some very reputable and expensive commercial speakers (and the fact my ears tell me that they are fine) makes me very pleased I had not read zaph's tests of them before I purchased. Zaph and I actually had a conversation about them in this thread His measurements and mine bore absolutely no resemblance, mine were pretty close to factory specs, whereas his were most definitely not!!.

Edit: and I would also say that the measurements cannot tell you the whole story, afterall I do not believe the measurements we make are measuring the true ability of a speaker to reproduce multiple different possibly conflicting sounds all at the same time and produce an image that sounds realistic. Unless someone develops a test that can do that I don't believe that any tests can tell you the full story

I would think that the various intermodulations that could occur with the reproduction of real music would be far more complex than any form of impulse response, chirp, warble tone or sine wave could hope to measure.

Tony.

[graaf]

Originally Posted by wintermute

hehehe I think it very interesting that the two quotes are from two people with IMO VERY different views on what is important

I think that the question of what is relevant measurement/interpretation ultimately follows the question of what I like or how I like it (for example I like it loud), especially when someone is self-declared hi-fi enthusiast

in my opinion there is a method of objectivization of all this, of finding correlation between subjective impressions and measurements graphs and any kind of objectivity that is beyond what we statistically like

I mean the method of controlled comparative (with different audio chains and with real sound sources) listening tests with participants statistically selected from consumer target group (Dr Toole's style) in a selection of standarized listening rooms

because I seriously believe that as Linkwitz put it unbiased listeners have no difficulty recognizing (more) accurate sound reproduction

Originally Posted by wintermute

Edit: and I would also say that the measurements cannot tell you the whole story, afterall I do not believe the measurements we make are measuring the true ability of a speaker to reproduce multiple different possibly conflicting sounds all at the same time and produce an image that sounds realistic. Unless someone develops a test that can do that I don't believe that any tests can tell you the full story

I would think that the various intermodulations that could occur with the reproduction of real music would be far more complex than any form of impulse response, chirp, warble tone or sine wave could hope to measure.

I agree
to overcome such problems Weber Rehde conducted His own measurements for example what He called comparative harmonics measurements of real musical sounds vs reproduced

then - how about using spectrograms of real music samples vs reproduced?
Spectrogram - Wikipedia, the free encyclopedia

best regards,
graaf
ps.
image attached is from a historical book by Hartley and it shows an example of a problem that seems to be overlooked nowadays

[Loren42]

Originally Posted by wintermute

Edit: and I would also say that the measurements cannot tell you the whole story, afterall I do not believe the measurements we make are measuring the true ability of a speaker to reproduce multiple different possibly conflicting sounds all at the same time and produce an image that sounds realistic. Unless someone develops a test that can do that I don't believe that any tests can tell you the full story
Tony.

I tend to disagree. However, I can see why you might say that.

Testing is one the hardest aspect of the project and few of us have the resources or the knowledge to test effectively. Given the shortcomings we encounter it doesn't surprise me that you can't depend on the results.

That doesn't mean that a loudspeaker's attributes are not testable or that the "nuances" you propose are not quantifiable.

The biggest problem I have with listening tests are that they are not quantifiable. You can't pin a quanta on to some subjective feeling and your ears and perception change over time. Just look at the way people (and experts) describe what they hear with terms like warmth, raspy, muddy, colored, etc. They are very subjective ratings.

[Robh3606]

Really, how about freedom from dynamic compression. How about good imaging and feeling of space and sound stage some systems have that others don't. Care to show me where I can see that in the measurements??

My original question that prompted Earls response. I see some people saying it can all be measured yet so far no recommended measurement set has been forth coming.

The idea is to be able to compare 2 speakers and by looking at the measurements be able to see which one would be superior as far as dynamic/power compression and imaging. Now this is using the standard tools we have like Clio, Arta, Holm Impulse ect.

Rob

[skeptic43]

There is Joseph D'Appolito's "Testing Loudspeakers" book from Audio Amateur Press. I've read thru it and Joe gives a lot of detail about the process - formulas and charts galore. However, what's lacking big time is how to interpret the measurements.

That's like getting the perfect ingredient list for the perfect recipe and no instructions on how to prepare it (e.g. order of addition, how much of each, cooking method, time, etc...). You get the picture (I hope).

[soongsc]

The important issue is one looks for different things in driver measurements and system measurements. Then one also looks for different things depending on system design or dirver design. Most software assume the user is going to design a system.

[SY]

The idea is to be able to compare 2 speakers and by looking at the measurements be able to see which one would be superior as far as dynamic/power compression and imaging.

The first (power compression) is pretty easy to measure and is done routinely by every competent driver and speaker manufacturer.

Imaging is a different matter- it's the "illusion" part, a trick your brain plays on you. There IS no-one standing between your speakers, singing. That quality is going to be ranked differently by different people- for example, my wife has very asymmetrical hearing between left and right, I don't, so my speakers don't image that well for her, yet they image superbly for me. So do they have good imaging or not?

The separate issue are the things that cause the illusion, and as has been pointed out, those are also easy to measure.