Larry Troxler wrote:
> 
> On Sat, 18 Apr 1998, tolve wrote:
> > concerned that subsonic sideband frequencies that result from fm might
> > damage speakers. tried to check out some files with sndinfo, but base Frq
> > doesn't show up in the report on ppc running csound 3.48. how do i
> > ascertain lowest frequency?

sndinfo will not give you this information. To get it, you would have to do a
fairly detailed analysis of the sound itself either with an lpc or an fft.


> Hi, although I don't have a specific answer to this question, the general
> question of avoiding speaker/headphone/ear damage is one that has always
> been on my mind, and surprisingly I haven't seen this obvious
> high-priority topic discussed much.
>
> First, regarding subsonic frequencies; is it really true that these are
> bad for speakers? Aren't they blocked by the amplifier? If not, what is
> the mechanism by whitch they are damaging?

Good quality modern power amps typically accept signals down to DC. Sub-audio
frequencies are no more damaging to speakers than other frequencies if they are
within the dynamic range of the electroacoustic chain. Where there might be
damage however is with strong DC-shifted sounds (a typical side-effect of
delayed feedback loops) where the onset of the sound crushes the speaker cone
outward (or inwards) by a sudden and large displacement.
 
> 
> Diverging to the more general topic of avoiding speaker damage, I think it
> would be interesting to hear what techniques people on the list tend to
> use. I'm speaking of the harzards involved when generating sound-files, as
> computer-music composer do, and not general techniques for avoiding
> damage from known-good sound sources.

In itself, digital distortion does not damage speakers. It is a belief based on
the harshness and unplesantness of the sound it produces. If the signal that
gets to the amp and speakers is within range, no damage is done. Damaging your
hearing is another matter... In other words, when in doubt, keep the level down!

> 
> Currently, I use the "first listen with cheap headphones" technique.
> 
> Along these lines, I've always thought it would be neat to have some way
> do detect mismatches sound formats (like littl/big endian mixups) at the
> back end, for example in the sound driver itself. I suppose things like
> this, and simlar problems like amplitude overflow, are the things I most
> worry about.

It would be nice to have a quick way to prevent overflow. Some users here aree
fans of floating opint computations with scaling at the end. That tends to work
well but involves an extra step. Predicting overflow howerver would probably be
more time-consuming than computing the soundfile itself...

-- 
________________________________________________________
Jean Piche
Universite de Montreal
http://mistral.ere.umontreal.ca/~pichej
http://www.musique.umontreal.ca/electro/CEC/