[[email protected]]

Or is my understanding wrong?

Depends how you set it. You can do that if you want (and it's a good way
to make snares pop out). You can do the reverse too, with an exaggerated
decay and the attack almost completely chopped off. You can also set some
compressors so that they lock at level when the input is silent, so
you specifically avoid pumping up the beginning of each word or note.

There are a hell of a lot of really useful things you can do with compression.
There are, as noted, some evil things you can do too. Tools are like that.
--scott



I see some of the new (digital) compressors have a new control HOLD

So for timing adj you have:

ATTACK... HOLD.... DECAY..

Setting the HOLD to 10 ms or more should prevent any distortion due to gain modulation down to 100 Hz.

M

[Don Pearce[_3_]]

On Tue, 12 Jan 2016 06:15:45 -0800 (PST), wrote:



Or is my understanding wrong?

Depends how you set it. You can do that if you want (and it's a good way
to make snares pop out). You can do the reverse too, with an exaggerated
decay and the attack almost completely chopped off. You can also set some
compressors so that they lock at level when the input is silent, so
you specifically avoid pumping up the beginning of each word or note.

There are a hell of a lot of really useful things you can do with compression.
There are, as noted, some evil things you can do too. Tools are like that.
--scott



I see some of the new (digital) compressors have a new control HOLD

So for timing adj you have:

ATTACK... HOLD.... DECAY..

Setting the HOLD to 10 ms or more should prevent any distortion due to gain modulation down to 100 Hz.

M


Any change of gain at any point causes distortion for the duration of
the change. The hold functions just delays its onset.

d

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[JackA]

On Tuesday, January 12, 2016 at 9:09:34 AM UTC-5, wrote:
Scott Dorsey, geoff, et al:
" thekkkma @ omnibus-brevis.edu wrote in message
...
- show quoted text -
And dummmmmmmm****s are tools. There's one dumb**** here who lays the
whip to his hobby horse, but all it does is decompose, because he rode
it to death long ago. Norman Bates and his mother; Dumb**** Theckma
and his hobby horse. The hobby horse was rode hard and put away dead.

FCKWAFA. AASBDFTOC."


See what I mean? This guy's obviously an
industry insider. But nice way for a "professional"
to act, eh? Probably does stuff to his clients'
projects without consulting them. Class act!
Too bad - you've been exposed, just by the
way you react on here.

I have to admit, I enjoy his use of, "short bus". Use it myself now. See, usenet is of value!!

Jack

[JackA]

On Tuesday, January 12, 2016 at 9:31:29 AM UTC-5, Don Pearce wrote:
On Tue, 12 Jan 2016 06:15:45 -0800 (PST), wrote:



Or is my understanding wrong?

Depends how you set it. You can do that if you want (and it's a good way
to make snares pop out). You can do the reverse too, with an exaggerated
decay and the attack almost completely chopped off. You can also set some
compressors so that they lock at level when the input is silent, so
you specifically avoid pumping up the beginning of each word or note.

There are a hell of a lot of really useful things you can do with compression.
There are, as noted, some evil things you can do too. Tools are like that.
--scott



I see some of the new (digital) compressors have a new control HOLD

So for timing adj you have:

ATTACK... HOLD.... DECAY..

Setting the HOLD to 10 ms or more should prevent any distortion due to gain modulation down to 100 Hz.

M


Any change of gain at any point causes distortion for the duration of
the change.


But how little distortion can humans detect? And, let's say it's an electric guitar, how can you tells it's distorted when distortion is already added, say by Jimi Hendrix?

Personally, I see the value of compression for those who mix audio tracks.

Jack

The hold functions just delays its onset.

d

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Angus Kerr]

There are many uses for compression - effect, and convenience. Distortion is pretty much the same.

I used to do tracking with some compression. One artist that I recorded - played acoustic guitar, and he used to hit the body and smash the strings and do all sorts of stuff that no matter what I did with the gain, would max out the converters and get digital clipping. So I ran his track through a compressor with a high threshold, fast attack and fast release and high ratio, so that the majority of his performance was uncompressed - passed straight through essentially, yet the loud bits didn't clip my converters. Which is really a peak limiter.

For effects, you can use a compressor to pump, wheeze, distort, whatever you want.

For mixing, say a dense rock track, you would need to compress the vocals fairly heavily so that they 'sit' in the mix. You will find that without compression, you need to ride the fader or program fader automation to get the vocal to sit right. I think the art to vocals is to compress them in a way that they do not sound compressed, so they sound airy and natural. Compressing tends to change the character of the sound of the instrument though. However, uncompressed tracks in a dense mix of distorted guitars tend to pop up and down in the mix because the underlying instrumentation is already compressed. Singing particularly is a very dynamic, so compression is justified. It does produce a thickness to the sound.

However, in a airy jazzy track with a good singer who knows how to control their dynamics, you might find that light to no compression may do the trick.

Distortion, is another issue.

What I think they mean by 'distortion' is that the compressed version of the mixed track is substantially, audibly different from the final mix, in not a good way. The final released version of the product sounds very different to the mix that the producer signed off on. Distorted electric guitar on the other hand, is an intentional part of the track and arrangement. Nothing sounds better to me than a guitar amplifier where the tubes are just starting to get tickled into crunching slightly. But this distortion is part of the mix.

Lars's drums imho sound like cardboard being hit with a lead filled sock in the example in the article. To my ears, the overall sound definitely diverged and had way less punch than the GH version. But, hey, if you've been beating those drums in a metal band for the last thirty years, maybe you just can't hear that...

-A.

[[email protected]]

On Tuesday, January 12, 2016 at 9:31:29 AM UTC-5, Don Pearce wrote:
On Tue, 12 Jan 2016 06:15:45 -0800 (PST), wrote:



Or is my understanding wrong?

Depends how you set it. You can do that if you want (and it's a good way
to make snares pop out). You can do the reverse too, with an exaggerated
decay and the attack almost completely chopped off. You can also set some
compressors so that they lock at level when the input is silent, so
you specifically avoid pumping up the beginning of each word or note.

There are a hell of a lot of really useful things you can do with compression.
There are, as noted, some evil things you can do too. Tools are like that.
--scott



I see some of the new (digital) compressors have a new control HOLD

So for timing adj you have:

ATTACK... HOLD.... DECAY..

Setting the HOLD to 10 ms or more should prevent any distortion due to gain modulation down to 100 Hz.

M


Any change of gain at any point causes distortion for the duration of
the change. The hold functions just delays its onset.

d


A compressor creates distortion when the gain is changing __at the same frequency__ as the sine it is controlling. This obviously creates 2nd order.
ie if the signal is 100Hz and the gain is changing in step with the 100Hz, that is non-linear harmonic distortion that creates harmonics.

If the HOLD feature is implemented correctly, during the duration of the tone, the gain isn't changing. HOLD is like a retriggerable one shot....so no distortion.

After the tone ends, and after the HOLD time expires, during the decay time, the gain is increasing but at a steady rate not at a cyclic rate so again there should be no distotion.

I'm assuming the HOLD is set for a durarion longer than the lowest frequency.


I'm using the term distortion here in the pure sense i.e. harmonic distortion.
Squashing of the dymaic range may be good or bad but technically it isn't harmonic distortion.

M

[Don Pearce[_3_]]

On Wed, 13 Jan 2016 11:29:11 -0800 (PST), wrote:

On Tuesday, January 12, 2016 at 9:31:29 AM UTC-5, Don Pearce wrote:
On Tue, 12 Jan 2016 06:15:45 -0800 (PST), wrote:



Or is my understanding wrong?

Depends how you set it. You can do that if you want (and it's a good way
to make snares pop out). You can do the reverse too, with an exaggerated
decay and the attack almost completely chopped off. You can also set some
compressors so that they lock at level when the input is silent, so
you specifically avoid pumping up the beginning of each word or note.

There are a hell of a lot of really useful things you can do with compression.
There are, as noted, some evil things you can do too. Tools are like that.
--scott



I see some of the new (digital) compressors have a new control HOLD

So for timing adj you have:

ATTACK... HOLD.... DECAY..

Setting the HOLD to 10 ms or more should prevent any distortion due to gain modulation down to 100 Hz.

M


Any change of gain at any point causes distortion for the duration of
the change. The hold functions just delays its onset.

d


A compressor creates distortion when the gain is changing __at the same frequency__ as the sine it is controlling. This obviously creates 2nd order.
ie if the signal is 100Hz and the gain is changing in step with the 100Hz, that is non-linear harmonic distortion that creates harmonics.

If the HOLD feature is implemented correctly, during the duration of the tone, the gain isn't changing. HOLD is like a retriggerable one shot....so no distortion.

After the tone ends, and after the HOLD time expires, during the decay time, the gain is increasing but at a steady rate not at a cyclic rate so again there should be no distotion.

I'm assuming the HOLD is set for a durarion longer than the lowest frequency.


I'm using the term distortion here in the pure sense i.e. harmonic distortion.
Squashing of the dymaic range may be good or bad but technically it isn't harmonic distortion.

M

Sorry, but no. If the gain is changing, then the instantaneous voltage
one cycle later is not what it should be for linear operation. It
doesn't matter what the rate is. The result of this is that for the
duration of gain change, there is distortion. And I am also talking
about real distortion that produces harmonics and intermods.

d

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[[email protected]]

Sorry, but no. If the gain is changing, then the instantaneous voltage
one cycle later is not what it should be for linear operation. It
doesn't matter what the rate is. The result of this is that for the
duration of gain change, there is distortion. And I am also talking
about real distortion that produces harmonics and intermods.

d

Don..

OK I guess we will have to agree to disagree.

If you take a pure 100 Hz sine wave and put it through an ideal VCA that is changing gain at a steady rate of say 1 dB per second (with no 100 Hz component in the control), then no 200 Hz component will be generated and I would call that no distortion.

Mark

[Don Pearce[_3_]]

On Wed, 13 Jan 2016 13:32:36 -0800 (PST), wrote:


Sorry, but no. If the gain is changing, then the instantaneous voltage
one cycle later is not what it should be for linear operation. It
doesn't matter what the rate is. The result of this is that for the
duration of gain change, there is distortion. And I am also talking
about real distortion that produces harmonics and intermods.

d

Don..

OK I guess we will have to agree to disagree.

If you take a pure 100 Hz sine wave and put it through an ideal VCA that is changing gain at a steady rate of say 1 dB per second (with no 100 Hz component in the control), then no 200 Hz component will be generated and I would call that no distortion.

Mark


You're right, there will be no 200Hz. The harmonic output is odd order
- so 300, 500 etc.

I've generated a pure 100Hz tone in Audition. Here are two pictures -
the first is tone with its spectrum as generated, and the second with
a steady, regular fade. You will see the difference:

http://www.soundthoughts.co.uk/look/no_fade.png
http://www.soundthoughts.co.uk/look/fade.png

d

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Phil Allison[_4_]]

wrote:


I've generated a pure 100Hz tone in Audition. Here are two pictures -
the first is tone with its spectrum as generated, and the second with
a steady, regular fade. You will see the difference:

http://www.soundthoughts.co.uk/look/no_fade.png
http://www.soundthoughts.co.uk/look/fade.png


Wow interesting..
I hadn't considered the 3rd and odd orders.
The pics were a great way to demonstrate the point!

So I leaned something new today...

A STEADY rates of gain change, casues no EVEN order,
but it DOES cause odd order distortion.


** But as the pics show, a slow fade produces inaudible (ie -108dB ) amounts of odd harmonics. It sounds completely clean - like a slow fade.

Far more concerning are the gain changes that happen during each cycle at low frequencies( ie 50Hz ) if the release time is not long enough. Distortion of the wave shape is highly audible, up to 30%.

Put simply, compressors see low frequency waves as a varying amplitude audio signals and try to compress them, which alters the wave shape.


.... Phil

[Scott Dorsey]

wrote:

Sorry, but no. If the gain is changing, then the instantaneous voltage
one cycle later is not what it should be for linear operation. It
doesn't matter what the rate is. The result of this is that for the
duration of gain change, there is distortion. And I am also talking
about real distortion that produces harmonics and intermods.

Don..

OK I guess we will have to agree to disagree.

If you take a pure 100 Hz sine wave and put it through an ideal VCA that is changing gain at a steady rate of say 1 dB per second (with no 100 Hz component in the control), then no 200 Hz component will be generated and I would call that no distortion.

The effect Don is describing is real, and it's most evident when you have
a lot of sharp transients and a lot of bass in the same signal, and
compression results in the transients modulating the bass. You have to
set the compressor fast enough to respond to the transient but when you
do that, you get the distortion on the bottom end. Ironically the more
accurate the curve on the compressor the worse the distortion seems.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Don Pearce[_3_]]

On Wed, 13 Jan 2016 18:52:43 -0800 (PST), Phil Allison
wrote:

wrote:


I've generated a pure 100Hz tone in Audition. Here are two pictures -
the first is tone with its spectrum as generated, and the second with
a steady, regular fade. You will see the difference:

http://www.soundthoughts.co.uk/look/no_fade.png
http://www.soundthoughts.co.uk/look/fade.png


Wow interesting..
I hadn't considered the 3rd and odd orders.
The pics were a great way to demonstrate the point!

So I leaned something new today...

A STEADY rates of gain change, casues no EVEN order,
but it DOES cause odd order distortion.


** But as the pics show, a slow fade produces inaudible (ie -108dB ) amounts of odd harmonics. It sounds completely clean - like a slow fade.

Far more concerning are the gain changes that happen during each cycle at low frequencies( ie 50Hz ) if the release time is not long enough. Distortion of the wave shape is highly audible, up to 30%.

Put simply, compressors see low frequency waves as a varying amplitude audio signals and try to compress them, which alters the wave shape.


... Phil

Yes, you really need to understand your sources and match the attack
and release to the lowest frequencies. A rule of thumb might be
minimum ten times 1/f at each end. A lot of sources won't allow that,
but it is amazing how much low cut you can get away with without any
loss of fidelity.

d

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Scott Dorsey]

Don Pearce wrote:
Put simply, compressors see low frequency waves as a varying amplitude audio signals and try to compress them, which alters the wave shape.

Yes, you really need to understand your sources and match the attack
and release to the lowest frequencies. A rule of thumb might be
minimum ten times 1/f at each end. A lot of sources won't allow that,
but it is amazing how much low cut you can get away with without any
loss of fidelity.

Well, that is the point of two-band compression.... you can have one speed
on the low end which is slow enough not to have substantial distortion, but
on the other hand you can have a faster compressor on the top end and not have
it affected by low frequencies. Reduced distortion on the bottom end, no
"pumping with the beat" on the top end. Very useful for things like
automatic gain controls and slow gain-riding systems which have wideband
program material going through them.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."