How important is low W & F performance anyway?


I recently completed work on a direct drive motor controller for a turntable mfr with IMHO, rather impressive results (0.004% 2 sigma method, 0.002% RMS).  In measuring other tables actual performance (vs published specs) I was shocked at the rave reviews two tables received that have rather lousy measured performance (but impressive specs).  It made me wonder whether the goal of ultra low W&F performance was really necessary?  I trust the measurements as they were verified by several methods and software tools and they correlated rather closely, yet the reviewers almost universally praise these tables.  It made me wonder if the reviewers even know what they are hearing or listening for and not to put to fine a point on it, does it even matter? 

phoenixengr

Elliott nailed it, if you can't hear it, it is not important.

I would tend to agree, but some of the numbers are high enough it's difficult to believe that no one could hear it.  Depending on the amount and band of the frequency deviation, I wonder if even reviewers (and certainly the average listener) don't realize that they are hearing it or perhaps it is perceived as some sort of euphonics or "voicing".

A lot of the tools used to measure this at the retail level are inadequate IMO.  Most of them high pass filter the signal to remove record eccentricity effects & tone arm contributions and are upper BW limited, both of which will mask serious problems with the drive.

The other table in question sells for more than $50K and has a laughable W&F spec of 1PPM (0.0001%) but actually measured 0.16%.  Again, nothing but praise from the listening tests.  One would have to conclude that specs are irrelevant since mfrs can claim almost anything they want with impunity and the fact that reviewers and owners of these tables have nothing but good things to say about the sound leads me to believe that even tables with lousy W&F performance apparently sound good to just about everyone.

This is a very broad statement that makes a lot of assumptions. One assumption it makes is that all tables of the same make and model perform identically. That is not necessarily true.

Another is that your statement assumes that there’s no confirmation bias. People listening or reviewing, say, a $25,000 turntable with fantastic published specs have a confirmation bias that they probably should love it.

Your conclusion that specs probably doesn’t matter is therefore not correct.

Condidering potential misrepresentation of true specifications, some natural variability between units (for a number of reasons including issues during transportation and the environment they are used in), and confirmation bias all play a major role in how things are perceived, and this goes not only for turntables, of course.

You make a lot of false assumptions:

Production variations and tolerances can account for small discrepancies from published specs, but the table in the quote you cited was off by more than 3 orders of magnitude (1600x).

My statements did not exclude the possibility of confirmation bias and I certainly believe that is a factor.  If a reviewer is evaluating a piece of equipment and the specs are fairly normal or at least reasonable, there should be no reason to doubt them or scrutinize more carefully.  In the case of DD2, a W&F spec of 1PPM begs for additional scrutiny, confirmation bias be damned.

I have no reason to believe that either table was not representative of a typical production unit.  While DD2 was not new, I saw no evidence of abuse or neglect or any other reason why it should perform as it did other than what I perceive as a design flaw (using an iron core motor for DD).  DD1 was purchased new from an authorized dealer and the packaging was substantial and bore no signs of distress or mishandling in shipping.  I sold it at a discount after the evaluation was complete and did not hear of any complaints from the buyer which would only further support my hypothesis.  You have no first hand knowledge of the units yet you speculate (assume?) that the poor measurements are due to transportation or environment?

@phoenixengr Bill, I have to agree with you that some of the numbers are too high to be inaudible.  Clearly either the listener was not paying attention or was deceiving him or herself on the basis of expectation.