[GreenKeys] Incredible . . .

[Craig Sawyers]

> > Now OK - the tube guys say that the carbon comp resistors are
> the ones that
> > sound better, and that is where my scientific approach breaks down.....
>
> And that is PRECISELY what he claimed!

I simply don't have a good technical argument why a carbon comp resistor
should sound better.  They are prone to long term drift in value (most of my
old AB stock is out by +20% or so, particularly high value ones), and they
have a horrendous temperature coefficient.  They are much better at
surviving medium term current surges than film resistors, at least in part
because the heat is dissipated throughout the bulk of the carbon rod, but I
can't see that being a critical parameter for audio use.

But, there is this odd thing about sound.  The ear/brain processing system
has evolved over millions of years, and is exquisitely sensitive to spotting
things that shouldn't be there.  An optical analogy, funnily enough used by
a pro-audio friend just today.  Imagine a large brick wall, as large as you
like, pretty much perfect in all regards.  Somewhere on it is a spider web -
and your attention imediately fixes on the web, even though it is only a
tiny object on the wall.  Or, you are at a party with a hundred people.
Amid the general loud sound of conversation and background music, someone
whispers your name at the other side of the room.  You immediately hear
that.  In both cases, you recognise something that is exceptionally
difficult to measure - the signal that you latch onto so effortlessly is way
below -100dB below the noise.

I guess what I am saying is that, implausible though it may sound that
carbon resistors give a more "realistic" (however that is defined)
presentation of the sound, it is just possible that they do.  But I am open
to be persuaded - by a suitable double blind test.  Amplifier X with metal
film (or whatever), and amplifier X but with carbon comp.  Which sounds
better?

Here's a waffly sort of pseudo-argument.  Harmonic distortion comes in even
and odd orders.  Even harmonics fall at octave intervals, and are thus in
tune with the sound (assuming that we are using a "western" musical system
based on 12 notes per octave, and an octave is a doubling of frequency).
Also, the third harmonic, and possibly the fifth, are also musically
acceptible (I believe).  The 7th and 9th harmonics are the real problem,
because they are not consonantly related to the music.  So there may be some
sort of interaction between the amplifier's active elements (tubes,
transistors etc) and a particular combination of passive components which
gives a more acceptable balance of harmonics in the distortion signature.

Of course this may all be different with different musical systems, like
Chinese.  Or the Gamelan from Java.  Or potentially oriental music in
general.

> There is of course a whole world of parameters not measured

And I think that makes my point.

Craig