"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim

Got me curious...

Good question...someone pls respond!

~S~


"Greg" wrote in message
om...
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim

The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim


--
Stephen Anderson

~At the end of the day, it's all about
the music

Further, DC offsets anywhere in the audio path have the effect of
lowering headroom on the side the offset is toward, thus potentially
resulting in premature clipping of one side of the waveform, not usually
a desirable effect in either the analog or digital world.

Stephen Anderson wrote:

The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:

"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim




--
Stephen Anderson

~At the end of the day, it's all about
the music

On Sun, 13 Jun 2004 18:05:55 GMT, Stephen Anderson
wrote:

The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

If anyone can actually HEAR a DC offset, either positive OR
negative, I would be very, very surprised. Hearing the EFFECT of it
(such as one polarity of a signal clipping before the other) may be
another matter.

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

I can't imagine that anyone would ever intentionally add a DC
offset.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording.

Recordings made from LP's may have a lot of low-frequency content
around 10Hz from the cartridge compliance/tone arm mass resonance
(especially if you can see the cartridge moving up and down over the
record). This can "approach" a DC offset and cause some of the same
effects, but it's something different. But this can be easily
eliminated in a DAW/wave editor with a 20Hz or so high-pass filter
(which also eleminates DC offset).

During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

The program surely reads sample values and does the calculations
correctly, else lots of people would have complained on Usenet.. What
amount of time are you scanning? A whole song (several minutes) or
more? This would be the "correct" way. If you're scaning a very short
time, such as 1/60th of a second, it would be easy to catch two
positive half-cycles and one negative half-cycle of a 40 Hz bass note,
and this would be detected as a positive DC offset.
If a whole track has DC offset, the DC offset should be eliminated

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion.

No. A low-frequency noise may look like a DC offset if you're
scanning less than several seconds.

Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets

Yes. I can't think of why you wouldn't want to do it. As someone
else responded, a LARGE DC offset can cause a significant lowering of
headroom. The problem with small DC offsets is when putting tracks
together, there will be a click or a pop at the transition from one to
the next, because the level changes from one DC offset value to
another, or to none if the next track has no DC offset. If all offsets
are removed, then all the DC levels are the same, so there's no "step"
or pop when going from one track to the next.
Using short fade envelopes will "fix" this, but there may still be
a thump. Eliminating DC offsets on each track before putting them all
together is the best solution.

Perhaps the most common cause of DC offsets is A/D converters that
are not adjusted correctly (presuming it has this adjustment), or that
have leaky input capacitors or some such. If this is fixed, then DC
offsets usually won't appear later on in the chain.

or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim


-----
http://mindspring.com/~benbradley

Actually, I have considered burning CDs with very slow but large DC ramps
and pulsed full power 20 kHz tones embedded into them and handing them out
to the owners of those boom cars that always seem to be driving along
beside you. That might accelerate the natural selection process.

On 13 Jun 2004 07:12:04 -0700, (Greg) wrote:

"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim


-- At the most of digital recordings, I consider anything down, say,
14 Hz as a LF junk and I get rid of it by applying multiple
simultanous high pass filters. These things only steal the headroom.
At a really low going bass loudspeaker, one could easily see sometimes
erratic movements and excursions at lowest frequencies caused not by
the DC offset itself but by low end modulation caused by ie. phono
tone arm resonances, LF leakages, vehicles passing by, foottaps, you
name it. A DC offsett steals the bandwidth here too: if you put a
silenced track, just with an amount of DC at these woofers, you'd
notice the membrane would go steady out or in, depending of the DC
offset polarity.
I consider removing DC offset as a sort of a "orthodigital hygiene"
and I appreciate any recording not showing significant DC offset when
put into an editor and checked for it. But again, numerous factors can
add a certain DC offset subsequently during a production.

But also, today, with those roaring thunder effects, one should remove
the DC only to preserve that booms. Or do a smart trick to enhance the
lowest end by adjusting some higher low frequencies.

Edi Zubovic, Crikvenica, Croatia

In the early days of digital recording, small DC offsets were often used to
intentionally shift the operating point of the ADC and/or DAC away from the
exact center of the digital code range. Early converters tended to have
their biggest non-linearity at exact mid-scale, distorting very low level
signals. With modern over-sampling converters, this is rarely an issue.

But old ideas, especially if their true intent was not understood, die very
slowly.

Just my 2-cents worth ...

Bill Whitlock

"Ben Bradley" wrote in message
...
On Sun, 13 Jun 2004 18:05:55 GMT, Stephen Anderson
wrote:

The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

If anyone can actually HEAR a DC offset, either positive OR
negative, I would be very, very surprised. Hearing the EFFECT of it
(such as one polarity of a signal clipping before the other) may be
another matter.

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

I can't imagine that anyone would ever intentionally add a DC
offset.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording.

Recordings made from LP's may have a lot of low-frequency content
around 10Hz from the cartridge compliance/tone arm mass resonance
(especially if you can see the cartridge moving up and down over the
record). This can "approach" a DC offset and cause some of the same
effects, but it's something different. But this can be easily
eliminated in a DAW/wave editor with a 20Hz or so high-pass filter
(which also eleminates DC offset).

During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

The program surely reads sample values and does the calculations
correctly, else lots of people would have complained on Usenet.. What
amount of time are you scanning? A whole song (several minutes) or
more? This would be the "correct" way. If you're scaning a very short
time, such as 1/60th of a second, it would be easy to catch two
positive half-cycles and one negative half-cycle of a 40 Hz bass note,
and this would be detected as a positive DC offset.
If a whole track has DC offset, the DC offset should be eliminated

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion.

No. A low-frequency noise may look like a DC offset if you're
scanning less than several seconds.

Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets

Yes. I can't think of why you wouldn't want to do it. As someone
else responded, a LARGE DC offset can cause a significant lowering of
headroom. The problem with small DC offsets is when putting tracks
together, there will be a click or a pop at the transition from one to
the next, because the level changes from one DC offset value to
another, or to none if the next track has no DC offset. If all offsets
are removed, then all the DC levels are the same, so there's no "step"
or pop when going from one track to the next.
Using short fade envelopes will "fix" this, but there may still be
a thump. Eliminating DC offsets on each track before putting them all
together is the best solution.

Perhaps the most common cause of DC offsets is A/D converters that
are not adjusted correctly (presuming it has this adjustment), or that
have leaky input capacitors or some such. If this is fixed, then DC
offsets usually won't appear later on in the chain.

or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim


-----
http://mindspring.com/~benbradley

Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occasionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

Normal editing adds DC offsets.

Consider a file composed of one full cycle of a sine wave. It will have no
DC offset due to its symmetry along its time base. The positive half wave
exactly cancels out the negative half wave. The wave sums to zero and there
is no DC offset.

However, if you remove any single portion of the wave, except a portion
along the time line that is symmetric about the mid-point of the wave, the
remainder of the wave will have a DC offset. There will be more of the
positive part of the wave than the negative part of the wave, or
vice-versa. The wave will no longer sum to zero and there will be a DC
offset.

Just extrapolate this to real world sound files that are composed of many
sine wave components that aren't and generally can't be edited so that all
waves have only symmetrical pieces removed, or changed.

In article writes:

I can't imagine that anyone would ever intentionally add a DC
offset.

No reason for it, but it's possible that non-symmetrical clipping
could be interpreted by a program as being a DC offset since the
average level would always be greater on one side of zero than the
other.


--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

Something to consider... Percussive Acoustic phenomenon occur due to either
something hitting something or plucking something. If it Hits, the polarity
of the initial wavefront will push.. Asymmetrically. conceiveably,
introducing a negative offset could result in a statistically 'hotter'
sounding mix if the peak signal level is established by assymetrical
material.
Just my .22 cents
GH

"Stephen Anderson" wrote in message
k.net...
Further, DC offsets anywhere in the audio path have the effect of
lowering headroom on the side the offset is toward, thus potentially
resulting in premature clipping of one side of the waveform, not usually
a desirable effect in either the analog or digital world.

Stephen Anderson wrote:

The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:

"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim




--
Stephen Anderson

~At the end of the day, it's all about
the music

Actually, no. The amplitude of the signal will still be the same
relative to its starting point, whether it's OVDC or some other DC voltage.

GHANKS77 wrote:
Something to consider... Percussive Acoustic phenomenon occur due to either
something hitting something or plucking something. If it Hits, the polarity
of the initial wavefront will push.. Asymmetrically. conceiveably,
introducing a negative offset could result in a statistically 'hotter'
sounding mix if the peak signal level is established by assymetrical
material.
Just my .22 cents
GH

"Stephen Anderson" wrote in message
k.net...

Further, DC offsets anywhere in the audio path have the effect of
lowering headroom on the side the offset is toward, thus potentially
resulting in premature clipping of one side of the waveform, not usually
a desirable effect in either the analog or digital world.

Stephen Anderson wrote:


The human ear can not differentiate absolute polarity except under very
specific conditions (see papers by Lip****z & Vanderkooey, I'll try and
find the reference). Thus, there is no perceivable difference between +
DC and - DC offsets. (Of course polarity differences between 2 or more
channels can be heard, that's not what this is about.)

Having worked in world class studios, including one associated with a
major label with world class mastering rooms, doing both analog as well
as various types of digital mastering (Sonic Solutions, et al) I have
personally never seen anyone add an offset to a track. But that's just
me, other may have had different experiences.

But, I have seen DC offsets in ProTools sessions, and have wondered WTF?

Greg wrote:


"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim



--
Stephen Anderson

~At the end of the day, it's all about
the music





--
Stephen Anderson

~At the end of the day, it's all about
the music

Sorry, no. The speaker cone would move either in or out to a certain
position and stay there, with audio movement superimposed upon it. DC
causes a speaker to stay in one place, as well as cause much distress to
the voice coil in terms of heat dissipation. The speaker motor depends
on a microscopic bit of rest at the zero crossing point, where no power
is being dissipated in the coil winding.

And this presumes a DC coupled audio path, from program all the way to
loudspeakers; possible, but not very likely. While certain consoles are
DC coupled, (SSL9K for example) I can't think of too many commonly used
power amplifiers whose input stages are.

Edi Zubovic wrote:

if you put a
silenced track, just with an amount of DC at these woofers, you'd
notice the membrane would go steady out or in, depending of the DC
offset polarity.


Edi Zubovic, Crikvenica, Croatia

--
Stephen Anderson

~At the end of the day, it's all about
the music

In the early days of digital recording, small DC offsets were often used to
intentionally shift the operating point of the ADC and/or DAC away from the
exact center of the digital code range. Early converters tended to have
their biggest non-linearity at exact mid-scale, distorting very low level
signals. With modern over-sampling converters, this is rarely an issue.

But old ideas, especially if their true intent was not understood, die very
slowly.

Just my 2-cents worth ...

Bill Whitlock

"Greg" wrote in message
om...
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion. Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)? I've
experimented with applying negative offsets on problematic songs, and
it seems to sometimes help reduce muddiness and boom, make things
clearer, brighten high-frequencies and correct uneven noise-floor and
channel-balance. But then sometimes it will sound too tinty or one
channel will sound either too fast or have a jerky tempo.

Should I always correct DC Offsets or are there times when adding an
offset would actually help or be desireable? If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be. And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

- Greg Yim

On Mon, 14 Jun 2004 06:52:26 GMT, Stephen Anderson
wrote:

Sorry, no. The speaker cone would move either in or out to a certain
position and stay there, with audio movement superimposed upon it.
--------8--------------

Yes, that's what I tried to mention but obviously used some wrong
words ) -- Oh well.

Edo Zubovic, Crikvenica, Croatia

PS As to removal of the DC, I remove everything what is up from abt.
19,5 or 20 kHz too (when working at 44100).

No worries, your english is really good! By the way, do you know a town
called Matije Gupca (spelling may be really bad, sorry)

Edi Zubovic wrote:
On Mon, 14 Jun 2004 06:52:26 GMT, Stephen Anderson
wrote:


Sorry, no. The speaker cone would move either in or out to a certain
position and stay there, with audio movement superimposed upon it.

--------8--------------

Yes, that's what I tried to mention but obviously used some wrong
words ) -- Oh well.

Edo Zubovic, Crikvenica, Croatia

PS As to removal of the DC, I remove everything what is up from abt.
19,5 or 20 kHz too (when working at 44100).

--
Stephen Anderson

~At the end of the day, it's all about
the music

Greg wrote:

"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound
Forge for awhile, and I've always corrected DC Offsets after
recording.

After? - with Cool Edit you can do it during.

During editing/mastering, I occassionally will run statistics
on a CD track and analyze it for signal levels, peaks, etc. to
see how the professionals do it.

Hmm ...

In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

You are confusing the natural asymmetry of some waveforms - generally
caused by the presence of second harmonics - and DC offset.

I was under the immpression that DC Offsets equal low-frequency noise
and harmonic distortion.

A DC offset is just that, nothing more.

Would a negative offset then actually cancel
out excessive boom on problem songs or act as a substitute for
noise-reduction (which always takes out too much of the music)?

No.

Should I always correct DC Offsets

Yes, but only when they are DC offset and you might as well do it during
recording/transfer. You should not correct natural asymmetry of a
waveform, trying to do that will cause a DC offset.

or are there times when adding an
offset would actually help or be desireable?

No.

If anyone knows the
answer to this, could you let me know and also briefly explain what
the effects of adding DC Offests - both positive and negative - would
be.

None.

And if anybody knows of any techniques, secrets, or figures they
could tell me, I'd greatly appreciate it. Thanks in advance!

Absolute polarity is a different issue. Asymmetry naturally occurs as a
consequence of the acoustic properties of air causing second harmonic
distortion. The naturally occuring asymmetry will _generally_ be such
that the largest excursion is the positive one. at least near the source
of sound. The issue of the audibility of absolute polarity is able to
carry a thread for at least 6 weeks to no conclusion other than that it
is sometimes audible, it is more audible to some than to others, and the
amount of second harmonic distortion in the playback system can act as a
bias that enhances audibility.

- Greg Yim



Kind regards

Peter Larsen

--
*******************************************
* My site is at: http://www.muyiovatki.dk *
*******************************************

On Mon, 14 Jun 2004 09:13:12 GMT, Stephen Anderson
wrote:

No worries, your english is really good! By the way, do you know a town
called Matije Gupca (spelling may be really bad, sorry)
---------------(-----------
Oh, it's so-so. Thanks for the compliment, anyway
-Well, the spelling is good but I think it may be rather a name of a
street; Matija Gubec has been a medieval peasant leader in a local
upstanding against the feudals near a town called Stubica in the
northern part of Croatia, he has been captured and killed at the end
and in Croatia, he's recognized as a hero of the past. There are many
streets, in various places, named after him. But I don't know for a
town bearing his name.

Edi Zubovic, Crikvenica, Croatia

Arny Krueger wrote:
Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge
for awhile, and I've always corrected DC Offsets after recording.
During editing/mastering, I occasionally will run statistics on a CD
track and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks
have large, uneven DC Offsets, often negative. Does anybody know if
this is a result of the programs reading the sample values wrong or
are they actually DC Offsets intentionally put there? If so, what
is the reason?

Normal editing adds DC offsets.

Consider a file composed of one full cycle of a sine wave. It will
have no DC offset due to its symmetry along its time base. The
positive half wave exactly cancels out the negative half wave. The
wave sums to zero and there is no DC offset.

No, DC implies a steady state, as opposed to a VLF 'average' as result of an
edit..


However, if you remove any single portion of the wave, except a
portion along the time line that is symmetric about the mid-point of
the wave, the remainder of the wave will have a DC offset. There will
be more of the positive part of the wave than the negative part of
the wave, or vice-versa. The wave will no longer sum to zero and
there will be a DC offset.

No, this will not give a steady DC voltage.

Just extrapolate this to real world sound files that are composed of
many sine wave components that aren't and generally can't be edited
so that all waves have only symmetrical pieces removed, or changed.

An average that is not zero due to asymetrical signals is not the same
thing as a DC offset. A DC offset will be there even with no signal.

geoff

Geoff Wood wrote:
Arny Krueger wrote:
Greg wrote:
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge
for awhile, and I've always corrected DC Offsets after recording.
During editing/mastering, I occasionally will run statistics on a CD
track and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks
have large, uneven DC Offsets, often negative. Does anybody know if
this is a result of the programs reading the sample values wrong or
are they actually DC Offsets intentionally put there? If so, what
is the reason?

Normal editing adds DC offsets.

Consider a file composed of one full cycle of a sine wave. It will
have no DC offset due to its symmetry along its time base. The
positive half wave exactly cancels out the negative half wave. The
wave sums to zero and there is no DC offset.

No, DC implies a steady state, as opposed to a VLF 'average' as
result of an edit..

Sure, DC in the larger context means steady state, from T= negative infinity
to T = positive infinity. However, I don't know any regular human who has
ever observed that happening, do you?

Might there be a practical definition of DC that allows us to think of
voltages that we've actually observed as being DC?

In this context, DC relates to a wave that does not average to zero over the
time span for which it is defined.

Arny Krueger wrote:

Sure, DC in the larger context means steady state, from T= negative
infinity to T = positive infinity. However, I don't know any regular
human who has ever observed that happening, do you?

Might there be a practical definition of DC that allows us to think of
voltages that we've actually observed as being DC?

No. A LF modulation or component though, maybe.

In this context, DC relates to a wave that does not average to zero
over the time span for which it is defined.

You mean that any aysmetric signal when averaged over time could be
considered to have a DC component ?

**** - just as well my speaker crossovers have a series C then.

geoff

In article ,
"Geoff Wood" -nospam wrote:

Arny Krueger wrote:

Sure, DC in the larger context means steady state, from T= negative
infinity to T = positive infinity. However, I don't know any regular
human who has ever observed that happening, do you?

Might there be a practical definition of DC that allows us to think of
voltages that we've actually observed as being DC?

No. A LF modulation or component though, maybe.

In this context, DC relates to a wave that does not average to zero
over the time span for which it is defined.

You mean that any aysmetric signal when averaged over time could be
considered to have a DC component ?


Yes, and that's a problem when you send it to a magnetic recording head. That's
why digital sum value is an important factor in coding for digital magnetic
recording.

**** - just as well my speaker crossovers have a series C then.

It's OK, your speakers are already magnetized anyway.

-Jay
--
x------- Jay Kadis ------- x---- Jay's Attic Studio ------x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x-------- http://ccrma-www.stanford.edu/~jay/ ----------x

"Greg" wrote in message
om...
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge for
awhile, and I've always corrected DC Offsets after recording. During
editing/mastering, I occassionally will run statistics on a CD track
and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks have
large, uneven DC Offsets, often negative. Does anybody know if this
is a result of the programs reading the sample values wrong or are
they actually DC Offsets intentionally put there? If so, what is the
reason?

I hadn't seen that before. Then I read this thread. An hour later, I mixed
down one of my Cakewalk projects, loaded the whole thing into Cool Edit for
some final tweeking -- there it was -- in the space I'd added some flange
was a clear negative DC offset.

There was plenty of headroom and it went away when I normalized the track.

Just wierd, ya know?

--
dt king
www.thoughtdog.com
Mellow New Age Music and More!

dt king wrote:
"Greg" wrote in message
om...
"QUESTION ABOUT INTENTIONAL DC OFFSETS"

I've been recording music from LPs using Cool Edit and Sound Forge
for awhile, and I've always corrected DC Offsets after recording.
During editing/mastering, I occassionally will run statistics on a
CD track and analyze it for signal levels, peaks, etc. to see how the
professionals do it. In doing so, I've noticed alot of CD tracks
have large, uneven DC Offsets, often negative. Does anybody know if
this is a result of the programs reading the sample values wrong or
are they actually DC Offsets intentionally put there? If so, what
is the reason?

I hadn't seen that before. Then I read this thread. An hour later,
I mixed down one of my Cakewalk projects, loaded the whole thing into
Cool Edit for some final tweeking -- there it was -- in the space I'd
added some flange was a clear negative DC offset.

There was plenty of headroom and it went away when I normalized the
track.

Just wierd, ya know?

Not weird at all, if you follow my explanation in another post. In general,
editing a file with a zero DC offset will result in a file with a non-zero
DC offset.

It's almost easier to list the operations on a file with a 0 DC offset that
Won't lead to file with a non-zero DC offset. Here's a few:

(1) Ideal gain change applied to the entire file
(2) Invert polarity for the entire file
(3) Interchange channels for the entire file
(4) High pass filter the entire file
(5) Ideally convert entire stereo file to mono

Any of these operations if not implemented ideally, can lead to a non-zero
DC offset.

Jay Kadis wrote:

**** - just as well my speaker crossovers have a series C then.

It's OK, your speakers are already magnetized anyway.


I was more worried about my voice-coil actually, and linearity and power
handling ...

geoff