Originally Posted by R Harkness
I'm so with you on this John!
I've been back in to auditioning all sorts of different speakers again recently and getting those elements of string sound right is always the Achilles heal of a sound system. I've heard countless sound systems, and own many different speaker designs (including the Paul Barton NRC-designed waveform speakers) and nothing ever truly captures the combination of presence and silkiness of a real string section. Even the most revealing speaker systems sound more like canned, thin, slightly strident samples vs the real thing. (That was true of when I auditioned the older Salon speakers as well).
And that is a specific symptom of the general difference I find with live instruments/voices vs reproduced: the way live sound has such presence and clarity yet is so relaxed and luxurious in it's detail and tonality.
Given no sound system (that I'm aware of) truly reproduces the full quality of live sound sources - especially given the variability of the source material we have to begin with - we tend to chase one or another aspect of sound we want to get most right. One of the reasons I've favored certain types of tube amplification over the years is because, while it may be a distortion introduced into the chain, it is a distortion that nudges the sound in the right direction to my ears in terms of what I'm trying to re-create - it can relax the sound, fill it out a bit, while not simply rolling it off and making the sound "dark." So to my ears vocalists sound more human, string sections a bit more silky yet present. Just one of many ways to skin a cat, but that is where you might say I may depart a bit from your approach. Like you I want accuracy, and I prefer neutral sounding speakers overall. But given the still-there compromises in sound systems, and the variability in sources, I feel a bit of coloration can be judiciously introduced to achieve another aspect of "subjective accuracy" - accuracy to how real sounds sound to me.
Fascinating that no one here seems ever to have listened to Allison loudspeakers, all of which have a uniquely 'silky' tweeter that puts out as much sound sideways as forward.
(Quite apart from their smoother fill in the lower midrange, around and the octave below middle C.)
That (unpatented) tweeter alone usually carries the day in blind and sighted testing alike.
I am not at all surprised that the speaker here which puts more treble into the reverberant field (even if by only a little) is preferred. This was widely known, appreciated, and made a goal, with improving technologies, from the 1940s on, from Jensen pronouncements then to Villchur, Olson, Allison, even Bose in his sloppy way, and on and on. The Boston Audio Society starting in 1972, notably via the speaker researcher Dr Mark Davis, went on at length about the paramount important of horizontal radiation pattern. I mean paramount in the literal sense. Not driver material, not 'time stuff, not distortion: horizontal radpat.
Davis wrote two important but today clearly unknown papers for High Fidelity and Stereo Review about what is important to our hearing in speaker playback, complete with testing.
Toole and Olive's subsequent work quantifying and further testing and refining all of this vein of thought, from Jensen through Davis, can hardly be overstated in its importance. You can take it to the bank.
(MDavis went on to design a line of several uniform-beamforming speakers for dbx.)
Aside from how much fun it would be to test the Revel against say an Allison Nine (much less the BL90, BL5, and new Kii, all with their wide and less-wide controlled beamforming, and the Keele CBT36K, and there are others of some interest, from Ohm to Linkwitz), it would be cool to see how it does against the very modest Infinity P362 or P363, which 'won' earlier Olive-Toole blind tests. Its spinorama is impressively, 'superiorly' uniform