Originally Posted by Floyd Toole
You gotta take some time off and read a bit. Over the years it has been shown repeatedly that trained listeners and untrained listeners agree on their preferences. Sean Olive has done it again, and again, most recently with headphones all over the world, and there are no noticeable national, age or cultural biases to what people tell us they like. The results are in AES papers. Good sound is good sound. The advantage of trained listeners is that they arrive at statistically significant ratings quickly and they describe what they hear in terms that are useful for design engineers. If you are in the business of designing loudspeakers time cannot be wasted.
Many reviewers give rave reviews to superb speakers and also to mediocre speakers. Customers in informal sighted listening evaluations are undoubtedly making biased judgments. Nobody can change that. If you come to the conclusion that sound quality does not matter, you could be forgiven. Several brands over the years have done well employing that philosophy.
However, manufacturers have a choice. Why not make neutral loudspeakers? It does not cost any more!!! A decision and competent engineering make the difference. At least those customers who care to listen carefully will be rewarded. All else is a lottery, leading to endless forum discussions.
Thank you for your contributions both to this thread and the industry in general. I have really
enjoyed reading your comments.
If you have the time, could you comment on the role listener preference plays in determining the meaning of speaker "accuracy" as Harmon uses the term? I am having a hard time getting my head wrapped around how listener preference is used to determine whether a speaker is faithfully reproducing the sound of the source.
From what I have gleaned from this thread, Harmon tests for listener preference and then designs speakers that objectively measure to meet this preference. And, it appears to me the purpose of the "spins" are to measure whether a speaker meets this tested-for preference. And it also appears that speakers that measure pretty
close to flat in a chamber are generally preferred in controlled tests, but not necessarily exactly flat. But this also seems to me different
from stating Harmon designs speakers to be what I always thought an "accurate" speaker meant, i.e
., that the speaker is faithfully reproducing the source, regardless of preference.
Jumping forward a step, I have also gleaned that it may be impossible for a speaker to, in fact, be meaningfully "accurate" in any given room, because of the impact any given room has on the speaker's response. Further, it may be impossible to even define "accurate" in a meaningful way outside of a chamber because of the interaction between speaker and room.
All this said, is Harmon using listener preference to help define whether a speaker is generally "accurate" for sound reproduction in small rooms through the spins because there is simply no better way to judge speaker accuracy outside a chamber
, and, in grey areas, Harmon wants to define
accuracy in a way that people generally prefer
? Or is there something else going on that I am missing and I have completely missed the boat?
I hope my question is not completely ridiculous to the pros here; I am not a sound professional or scientist, but I love buying gear, watching movies, and continuing to try to improve my listening experience, and I am trying to understand the theory behind Harmon's preference-testing process a little better.