From
http://www.diyaudio.com/forums/multi...ml#post1506291
by Lynn Olson
"There's a little story I heard a long time ago. Los Angeles, home of Altec and JBL, was deeply resistant to time-based measurement methodologies. It was swept-sinewave or nothing for these guys.
Just to stir things up, Richard Heyser sprang a little surprise at a Los Angeles AES meeting. It was a little black box that had very low distortion, measured completely flat using Altec and JBL's swept-sinewave methodologies, yet was so grossly distorted that speech was unlistenable. Much head-scratching ensued - and Dick already had a well-established reputation as one of those Caltech wise guys.
At the end of the talk, he revealed the secret of the black box: it was a very high-order allpass filter, that was randomly switched in and out of the signal path. It completely garbled the time domain in the fashion of an especially bad shortwave signal.
This is one of those apocryphal stories I heard back when I lived in The City of the Angels, but I knew several Caltech guys, and it sure sounded in character to me - they were always up to some kind of complex intellectual stunt.
I look back at the numerous articles Richard Heyser published in Audio magazine in the early Seventies, and the resounding silence that greeted them from Altec and JBL. No time domain for us, no sir! There's some irony in the fact that both companies were turning their back on the (time-aligned) Shearer Theater Horn that they owed their existence to. "
Also please check "The Heyser Box" by Robert Harley
http://www.theabsolutesound.com/arti...sounds-like-1/
The story is not trying to say the measurement itself is not important. On the contrary, this tells us knowing how and what to measure is the most important thing or you will get the wrong conclusion otherwise. Just my 0.02.