speakers: was AW: [sdiy] DIY help needed

[Scott Evans, Gen Mgr]

See below,

Scott

Czech Martin wrote:
> 
> Well, this is another can of worms ;->
Yup

> on one hand you want to divide the audio spectrum in a lot of bands, usually
> up to 4, since you want to avoid these problems:
> 
> -partial resonances in membranes (@1kHz the wavelength is about 30cm,
>  so standing waves can fit into a woofer cone as large as that)
Good reason for watching how high the woofer goes (see also doppler
below)

> -focussed radiation beam if wavelength comes close to cone diameter
> -large low frequency excursion (2kHz is LOW for a tweeter)
agreed

> -doppler FM distortion due to large velocity
Limiting the upper range of drivers mitigates these effects,
particularly in the low end driver, since the ear is less sensitive to
pitch changes in the low end.

> It is simple to show that a three way speaker can never
> have linear phase @ all points of the listening room.
Yes, but with adequate dispersion, the sweet spot can be large enough to
accommodate a greater area.

> with additional phase distortion. Crossover networks
> will make all this worse, regardless what manufacturers claim.
Herein lies the key. Crossovers make a marked change on the phase
alignment of the signal, even to the extent of interfering with physical
time alignments of the drivers. Crossover design cannot be ignored to
this extent. KEF did some extensive research in this area in the 1970's
with pulsed noise and fast fourier time analysis. They measured not only
the transient response of integrated systems, but how the frequency
response of the system changed over time. Quite interesting, and
accounted for a number of fine reference speakers being produced, unlike
the KEF of today.

> Funny enough that all these inventions come from 1970 or so.
> These years seem to be very innovative for audio.
> Hey, isn't there any progress today?
Good question.

> The question is of course if all this matters, if we
> have the reproduction of pure electronic music in mind.
Yes! Even though we don't know what our electronic "sound" is supposed
to sound like, and it may be claimed that the loudspeaker is part of the
musical instrument for live performances, I want to know if the
reproducer is giving me at least a reasonable facsimile of the waveform
presented to it.