John Fields said:
On Thu, 10 Mar 2005 07:27:49 -0800, "Larry Brasfield"
---
Superstitious minds also exist among those who prefer "semiconductor
sound" and I'm sure that, among that fragment, you'll also find many
with magnetic shoe inserts, copper bracelets, or a cupboard containing
many items with sea salt in the ingredient list, so I can't see why
you're singling out those who prefer "the tube sound" as a population
more likely to be superstitious than another unless it's for the
purpose of denigrating them because they have preferenes different
from yours.
You are mistaken about my preferences. I cringe
when I hear hard-limited sound and would prefer
the softer limiting that tube amps can provide.
My "singling out" is directed not to those who may
prefer a given type of distortion or limiting, but to
those who believe only tube amps can deliver that
performance. I have seen no studies of how tube
amp preference correlates to superstition, so my
above statement is based on a small sample and
may well be unrepresentative.
Well, prepending 'not' to what I posted can certainly
be reasonably interpreted as a contradiction.
The frequency response of a transformer is determined by using a
constant voltage, constant impedance source with a sinusioidal output
to sweep the input of the transformer (the "primary") through a band
of frequencies while measuring the output voltage from the resistively
loaded secondary and taking note of the output amplitude variations.
The transient response, on the other hand, deals with the behavior of
the transformer when subjected to complex input signals and is
affected by, among other things, the leakage reactance and
interwinding capacitance in ways which wouldn't be readily apparent
when exciting the transformer with a single spectral line.
For a linear time-invariant system, the transient response
is perfectly predictable from the frequency response. So
your distinction is somewhat puzzling in this context.
As for the properties you attribute to the transient response
that are not visible in a magnitude versus frequency plot, my
understanding is that humans are insensitive to the phase
relationships among components of an audio signal.
For
example, do you think that if you had a transformer with a flat
frequency response from 10Hz to 20kHz at 10 watts and you fed it with,
say, 100Hz and 8372Hz simultaneously that its output would exactly
mimic its input?
Such fidelity would surpise me.
I don't, and I don't believe that if you fed the
output of that transformer into a loudspeaker side-by-side with an
identical loudspeaker being fed with the output of a solid-state amp
with a treble control that the inputs to both loudspeakers could be
made indentical no matter how the treble control was adjusted.
I agree with that. But I doubt anybody could
hear the difference provided that there was no
wild difference in the phase response of the
two systems, such as delay in excess of a mS.
What triggers my skepticism is the suggestion
that transformer transient response uniformly
tends to smooth output waveforms. I expect
that in the case of soft limiting, but I see little
reason to predict that the uneven phase response
of a transformer, together with the phase response
of whatever is done to preceeding stages to get
the magnitude response straightened out, will act
to produce smoother outputs.