2. tech-talk
  3. 1332143340

How much can be measured -- and how much cannot?

There has been a lot of discussion over the years on Audiogon regarding the measurement of components and other audio products. Some people claim everything is either measurable now or will one day become measurable with more sophisticated measuring equipment. But others say there are things in high end audio that will never me measurable and that measurements are really not that important.

Here is a typical example -- a quote taken from the Stereophile forum regarding their review of the Playback Designs MPS-5:

"JA 2/17/10 Review Measurements of Playback Designs MPS-5
Posted: April 13, 2011 - 8:42am

John Atkinson's 2/17/10 review measurements of the Playback Designs MPS-5 revealed less than stellar technical performance even though Michael Fremer really liked the player. I've included JA's closing measurement remarks below followed by the manufacturer's comments.

To my knowledge there was never any followup in Stereophile regarding the manufacturers reply the MPS-5 could not be adequately measured with traditional measurement techniques.

I believe Stereophile should respond to this reply in the interests of its own measurements credibility.

Len"

How important do you think measurements are? Are the ears really the only true arbiter?
sabai

Showing 7 responses by tbg

tbg
5,263 posts
 
And around and around we go. In all cases where JA didn't like much in measuring and Fremer liked the component I have always agreed with Fremer. A classic example was the WAVAC SH-833 monoblock amps. Fremer loved it as did I on risking hearing something I could not afford as CES. Atkinson panned it and it does look pretty awful. My conclusion is that we can measure what is not important to what we hear.
tbg
5,263 posts
 
Unsound, you are right, speaker frequency response and sensitivity matter. All the cases I was alluding to were electronic components.
tbg
5,263 posts
 
Unsound, I have certainly heard the impact of some of these, but I don't know of much concern with any of these save what is audible. A tuner or an amp that fails to deal with such issues probably will not sell given how it performs or sounds.
tbg
5,263 posts
 
Newbee and Unsound, I certainly would take no exception to your centering on specs that you find closely associated to your tastes, but I will continue to trust my ears. I have never found any measures other than whether the unit is on or not that are associated with quality music reproduction. Michael Fremer and I certainly are in total agreement about the WAVAC SH-833 monoblock amps. He is the only reviewer in Stereophile or TAS, that I would trust in recommending a component.

At one RMAF, John Atkinson presented a seminar where he had a Boulder amp and an unnamed amp hooked to a THD meter. Under various loads the unnamed amp was horrible and the Boulder exemplary. He went on and on. He had no capability to listen to the two amps. Having heard the Boulder and not liking it, I asked whether he thought the designer of the unnamed amp thought THD was a major concern in design. He was flabbergasted as were most in the audience. I left.
tbg
5,263 posts
 
Newbee, you are right that listening entails tinkering with speaker placement, compatible electronics, etc. and the tastes of the reviewer. While measurements are more stable, but also they ignore the subtleties of tweaking with placement, etc. But obviously many of us with some experience with measurement perfect electronics, know that listening by the customer is the only real guide.
tbg
5,263 posts
 
Unsound, when measurements were added to reviews I applauded it. However, I have seen little correlation between them. Nor have I heard good measuring components to sound better nor good reviewing components to sound better. Related to this is double blind testing. Having participated in several of these and having read critiques of the use of 30 sec. same/different tests in psychology, I find them invalid.

I really think that components should be sent to a known room where we have made a recording and had the performance digitized. Then we compare that component inserted into a known system whose measurements are known. We would then compare the change noted from both the existing system and the actual recording. The closer to the initial recording the better.

Screw all other measurements.
tbg
5,263 posts
 
Unsound, I can hear frequency response and sense dispersion also. I also think I can hear phase problems in speakers. I cannot recall ever getting such measurements before deciding on a speaker. But I cannot really understand a speaker designer who wouldn't use such measurements. I must say that I have sought and appreciate single driver speaker systems and two way systems.
More to discover