[LAU] Pro Audio? OT rant.

Chris Bannister cbannister at slingshot.co.nz
Mon Dec 24 03:06:32 UTC 2012


On Sun, Dec 23, 2012 at 10:05:22PM +0000, Fons Adriaensen wrote:
> On Sun, Dec 23, 2012 at 02:48:09PM -0500, Gene Heskett wrote:
> 
> > Filters that sharp ring unavoidably. 
> 
> That 'ringing' is a complete misnomer. It's *not* the same
> thing as would be produced by e.g. a high Q resonance. 
> 
> Even if the waveform looks like oscillating before and 
> after the transient, that does not mean that any energy
> at that frequency is actually present. And in in fact it 
> is *not* present. The 'ringing' you see is the difference
> between the full-bandwidth signal and the filtered one.
> In other words, it's the part of the signal that was
> ***removed*** by the filter.

So artifacts (transient peaks) on the trailing or leading edges of a
square wave is not a form of ringing?

I remember years ago when I was working in Telecom (then Post Office)
and me and another guy were interested in checking out some VHS VCR's,
so we "borrowed" some models which we would return on Monday.

This other guy knew what he was doing and we also had access to quite a
bit of test equipment. One test I remember we did was to feed a square
wave (which was available as a reference signal from the actual o'scope we
were using) into the VCR and view the output on the o'scope. On one
model (Sharp, I think) it was distorted to hell, it looked more like a
bad sine wave than anything resembling a square wave. The weird thing
was that it sounded clearer than the other models, you could hear paper
rustling and clothes brushing against things very clearly.

On another model, the output was not a bad representation of the input
signal, but there were noticeable peaks on the trailing edge of the square
wave. If my memory serves me correctly, he referred to this as ringing,
caused by the filtering stages in the circuitry. Also I remember that he
said that if the peak was on the leading edge it was filtering on the
input stage of the circuitry and if it was on the trailing edge it was
caused by filtering on the output stage of the circuitry.

That test alone made me realise that a simple listening test is not
sufficient. I'm guessing that the Sharp model would produce "listening
fatigue" long before the other model which had minimal distortion on the
output squarewave.

-- 
"If you're not careful, the newspapers will have you hating the people
who are being oppressed, and loving the people who are doing the 
oppressing." --- Malcolm X


More information about the Linux-audio-user mailing list