Paul Harris wrote:

anybody-but-bush wrote:

An age old fuel source for flame wars.
1) the recorded signals of musical instruments,etc are not symmetrical.(fact)
2) some people claim to be able to tell which is true to the source and vice verse.
3) some people claim not to be able to tell is was is true to the source and vice verse.
4) multitracked recordings where a source sound is picked up by more than one mic obscure this
phenomenon .

Thanks to everyone who has shed light on this and apologies for fanning
old flames.


A legitimate question, none-the-less (although its flamed past
may have relegated it a spot in the FAQ :-)


I had not considered symmetry. The "distortion on high volume" got me
wondering if there are speakers that oscillate around a voltage other
than zero i.e. the natural rest position on the cone is different from
the powered but silent position in which case when switched on, the cone
would be seen to jump forward (or back) and hold due to a non-zero
voltage being applied.


The only place that I have seen this is when an amplifier has a
DC component on its output (not good). This not only has the
speaker dissipating power when there is no sound being produced
(hey! That would make it a "class A" speaker! :-), it now puts
the operating point on the speaker such that it can limit in one
direction befor it limits in the other.


Since nobody has mentioned this, I guess there would be no advantage in
such a design - there is no need to "park" a speaker when powered off
like a hard disk drive head.


There might be such an advantage on more exotic speaker designs
that are attempting to utilize some other principles in
operation, but the speaker would need its own (matched) power
circuitry to ensure that the zero point of the transducer element
was matched to the zero point of the driving electronics. It
would not be so much a need to "park" the transducer when off as
the need to optimize its zero point when in operation. I don't
know of any transducer types that do this but that doesn't mean
there aren't any.

On a normal speaker, the unpowered reference point naturally
falls in the middle of the speaker's normal travel range (like a
spring that has neither been compressed or stretched). No current
applied, speaker is at reference point. Positive current applied,
speaker moves out from reference. Negative current applied,
speaker moves in from reference (sorta like a "class B" operation :-)

Since the farther you move away from the reference point, the
closer you approach the speaker's physical limits and the greater
the possibility of operating in non-linear regions of the
speaker's travel. If the non-linear areas were not symmetrical
about the rest position, then conceptually applying a DC offset
to "center" the operation between the non-symmetric areas might
seem feasible. However, it would appear to me that the complexity
of doing this could be orders of magnitude above that of just
designing the transducer so that it is symmetric and linear in
both directions (i.e., the current phylosophy of most speaker
driver design)

- Jeff