spkey wrote:laird wrote:I'll reiterate: companies make up these numbers. there is no standard for measurement or reporting. So don't trust these numbers AT ALL.
Of course there is. There are a number of ways to measure a speakers' frequency response...(edited for bevity by laird]... That's what the frequency response graph displays. I do not remember at what intensity measurements are done but the information should be there on the manual when you buy a good speaker.
There are ways of measuring, yes. I agree with you.
but that's not what i said, now, was it.
there is no
standard, like there is for say cars and gasoline consumption. GM can't roll their cars down hills and say they get 200 miles per gallon. Logitech can.
A company can choose any way it wants to measure their speakers' frequency response.
the white noise test you mention is good. Obviously, the more data points the better.
generally, only good speakers, just like you mention, actually list that sort of information somewhere.
Every brand of cheap speaker I've seen
doesn't even label their graphs, let alone tell you how or what they measured.
which is why I say I could sell my speakers with a perfectly flat response by only measuring two frequencies. When a company says their freq range is 20hz-20,000hz, what does that mean? they get to choose their cutoff points. A good company wouldn't list the ability of their speakers to produce a 20hz tone if its 100dB below a 400hz tone of similar singal amplitude. A bad company might.
So I say again... don't trust those "numbers", they have no meaning.
I agree with you that there are ways of making meaningful measurements, but there are no laws saying companies have to do this.
Finally,
if you ever see a graph that has no error bars, ignore it.
taking a single measurement is like never backing up your computer data: stupid!
