In article ,
Keith G wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in this
software!!)

When you add two mono signals - like say left and right of a stereo signal
which are identical, you get an increase of 6 dB. You can get near this if
a genuine stereo signal contains the loudest part in effectively mono. So
to avoid any chance of over modulation many simple progs will attenuate
things by 6 dB when you switch to mono.

--
*Geeks shall inherit the earth *

Dave Plowman London SW
To e-mail, change noise into sound.

On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Keith G wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in this
software!!)

When you add two mono signals - like say left and right of a stereo signal
which are identical, you get an increase of 6 dB. You can get near this if
a genuine stereo signal contains the loudest part in effectively mono. So
to avoid any chance of over modulation many simple progs will attenuate
things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

d

Ian Iveson wrote:
Ian Bell wrote:

That's not distortion that's clipping.

And clipping isn't distortion? Symmetrical clipping is a sum
of odd harmonics. Asymmetric clipping is a mixture of even
and odd. Clipping also introduces severe IM.


Don't be such a pedantic arse.

If both gain stages produce only even harmonics, then they
will tend to cancel and the remainder will still be even. If
both produce lots of even and some odd, then the evens will
tend to cancel and the odds sum when combined. The
consequence is that a range of combinations of odd and even
are possible, depending on amplitude and the relative gain
contributed by each of the two stages.


Rubbish, the proportion of distortion in each stage is primarily
determined by the signal level in that stage, the last stage thereby
swamping that produced by earlier ones.

Cheers

Ian
Ian

Don Pearce wrote:
On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Keith G wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in this
software!!)
When you add two mono signals - like say left and right of a stereo signal
which are identical, you get an increase of 6 dB. You can get near this if
a genuine stereo signal contains the loudest part in effectively mono. So
to avoid any chance of over modulation many simple progs will attenuate
things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

d


Which prog is it?

Cheers

Ian

In article , Don Pearce
wrote:
On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article , Keith G
wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in
this software!!)

When you add two mono signals - like say left and right of a stereo
signal which are identical, you get an increase of 6 dB. You can get
near this if a genuine stereo signal contains the loudest part in
effectively mono. So to avoid any chance of over modulation many simple
progs will attenuate things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

Does it flat-top clip or invert the sign? Wondering if the internal values
are misusing signed/unsigned int values.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

In article ,
Don Pearce wrote:
On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Keith G wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in
this software!!)

When you add two mono signals - like say left and right of a stereo
signal which are identical, you get an increase of 6 dB. You can get
near this if a genuine stereo signal contains the loudest part in
effectively mono. So to avoid any chance of over modulation many simple
progs will attenuate things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

Best avoided, then. An accurate way of measuring levels is essential for
any sound work.

d

--
*i souport publik edekashun.

Dave Plowman London SW
To e-mail, change noise into sound.

On Sun, 07 Feb 2010 11:28:56 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Don Pearce wrote:
On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Keith G wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in
this software!!)

When you add two mono signals - like say left and right of a stereo
signal which are identical, you get an increase of 6 dB. You can get
near this if a genuine stereo signal contains the loudest part in
effectively mono. So to avoid any chance of over modulation many simple
progs will attenuate things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

Best avoided, then. An accurate way of measuring levels is essential for
any sound work.


Sure, but once you are aware of the quirk it is easy enough to
sidestep the problem.

d

On Sun, 07 Feb 2010 11:02:05 +0000 (GMT), Jim Lesurf
wrote:

In article , Don Pearce
wrote:
On Sun, 07 Feb 2010 10:19:41 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article , Keith G
wrote:
(Wish I knew what caused the -6 dB 'ceiling' when recording mono in
this software!!)

When you add two mono signals - like say left and right of a stereo
signal which are identical, you get an increase of 6 dB. You can get
near this if a genuine stereo signal contains the loudest part in
effectively mono. So to avoid any chance of over modulation many simple
progs will attenuate things by 6 dB when you switch to mono.

No, I've seen the effect with this prog. It actually clips at the -6dB
line. Very strange - presumably a software bug that hasn't been sorted
out yet.

Does it flat-top clip or invert the sign? Wondering if the internal values
are misusing signed/unsigned int values.

Flat top, just like any old clipping.

d

"Ian Iveson" wrote in
message
Ian Bell wrote:

That's not distortion that's clipping.

And clipping isn't distortion?

Of course it is distortion. It is a mixture of even and odd distoriton
depending on how symmetrical it is.

Clip exactly half the waveform, and you get only even order distoriton.

Clip both top and bottom equally, and you get only even order distortion.

But, either are special cases. If you clip exactly half the wave, then the
amount of distortion is the same, regardless of level.

If you clip symmetrically, there is no distortion until clipping commences.

Symmetrical clipping is a sum of odd harmonics.

Agreed.

Asymmetric clipping is a mixture of
even and odd.

Agreed, except for the case of exactly symmetrical clipping, in which case
there is no even order distortion.

Clipping also introduces severe IM.

As does all other forms of nonlinear distortion. IM distortion is often
more prevelent than harmonic generation when music is being distorted.

If both gain stages produce only even harmonics, then they
will tend to cancel and the remainder will still be even.
If both produce lots of even and some odd, then the evens
will tend to cancel and the odds sum when combined. The
consequence is that a range of combinations of odd and
even are possible, depending on amplitude and the
relative gain contributed by each of the two stages.

The problem with clipping is that it degenerates into two special cases -
either equal distortion at all levels, or no distortion until a certain
threshold is crossed.

"Arny Krueger" wrote in message

"Ian Iveson" wrote in

Correction:

Clip both top and bottom equally, and you get only odd
order distortion.