The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
compression: a tin can or automatic volume control? Dynamics Plugins
Old 19th September 2011
  #1
Gear Nut
 

compression: a tin can or automatic volume control?

Why do practically all compression tutorials describe a compressor as an "automatic volume control" and say things like, "...however, the compressor can turn the volume down quicker and with more precision than the engineer ever could." It gives the impression that using a compressor is the same exact thing as volume automation.

Isn't calling a compressor an "automatic volume control" misleading? Isn't a tin can a better analogy? You squash the top down and it becomes smaller. If you bring the can back up to its original height (with make-up gain), the bottom will now be higher up too.

If a compressor was just an automatic volume control, then when the make-up gain was used, the signal would sound exactly the same as it did originally. There wouldn't be much of a point.

Does anyone else find the common general explanation of compression to be misleading or am I way off?
Old 19th September 2011
  #2
Registered User
There are endless debates and books written about compression. It's all true - compressors are flexible tools that do a lot of good (or harm) depending on how you use them.

The "automatic level control" is a better description than the crushing a tin can description.

Your statement: "If a compressor was just an automatic volume control, then when the make-up gain was used, the signal would sound exactly the same as it did originally. " is logically flawed.

UNLESS you applied Make Up Gain with perfectly matching but opposite 'virtual fader movements' as the automatic volume control, the level would NOT be restored to the original uncompressed signal.

Make Up Gain is a fixed static gain boost.

So compression is literally automatic gain control. However, that term is usually used for slow attack, slow release, "lazy" compression that more resembles what a human hand on a fader can do.

Compression is generally faster than a human can do. Limiting is basically very fast, high-ratio compression.
Old 19th September 2011
  #3
Gear Nut
 

Quote:
Originally Posted by Kiwi View Post
There are endless debates and books written about compression. It's all true - compressors are flexible tools that do a lot of good (or harm) depending on how you use them.

The "automatic level control" is a better description than the crushing a tin can description.

Your statement: "If a compressor was just an automatic volume control, then when the make-up gain was used, the signal would sound exactly the same as it did originally. " is logically flawed.

UNLESS you applied Make Up Gain with perfectly matching but opposite 'virtual fader movements' as the automatic volume control, the level would NOT be restored to the original uncompressed signal.

Make Up Gain is a fixed static gain boost.

So compression is literally automatic gain control. However, that term is usually used for slow attack, slow release, "lazy" compression that more resembles what a human hand on a fader can do.

Compression is generally faster than a human can do. Limiting is basically very fast, high-ratio compression.
I understand what you are saying, but my point was that the "automatic volume control" analogy doesn't take into consideration the reduced dynamic range of the new signal and it can make the concept of compression confusing when it is introduced in this way. Even with a slow attack and release, the dynamic range of the signal is still being reduced. That's why I was using the tin can analogy. the reduction in dynamic range seems just as important as the automatic gain control concept.
Old 19th September 2011
  #4
Registered User
I don't follow ... "reduced dynamic range" = crushed can doesn't it?

The main thing to understand about compressors is that they compress. It's not hard.

Technically, you have some sort of gain control (which corresponds to a fader) and a sidechain signal that controls this (which corresponds to human hand on the fader).

It's Automatic.
It's controlling Volume or Level (by controlling gain)

In what way is the concept of "automatic volume control" misleading?

If they were described as distortion or eq effects, that would be more misleading. Although they certainly can create distortion or eq effects.

Your crushed can analogy implies some serious damage - a high quality compressor doesn't have to destructively crush sound.
Old 19th September 2011
  #5
Lives for gear
 

Quote:
Originally Posted by capproach View Post
Why do practically all compression tutorials describe a compressor as an "automatic volume control" and say things like, "...however, the compressor can turn the volume down quicker and with more precision than the engineer ever could." It gives the impression that using a compressor is the same exact thing as volume automation.

...If a compressor was just an automatic volume control, then when the make-up gain was used, the signal would sound exactly the same as it did originally. ...
I don't see your logic. It seems instead you have latched on to the idea that volume automation is not a process of dynamic range reduction.
Sure it's by different means, and the resulting style is very different, but when I automate gain (or the fader- very often do both) the moves are just as apt to include raising sections as well as reducing them.

Now you have essentially floating 'make up, and/or reduction via 'riding.. pre or post a compressor that may be inserted as well for that matter.
The extremely cool thing about automation lines is we can go in and tailor by ear down to the word/phrase level if need be attacks', envelopes, decays in ways 'automatic (static setting) gain control can't.
Old 19th September 2011
  #6
Lives for gear
 
Aisle 6's Avatar
The idea of a compressor being an auto volume automated is correct in as far as it is done with a nonlinear process. Therefore, the outgoing signal is a distortion of the original...this is why we love them.
Old 19th September 2011
  #7
Quote:
Originally Posted by capproach View Post
I understand what you are saying, but my point was that the "automatic volume control" analogy doesn't take into consideration the reduced dynamic range of the new signal and it can make the concept of compression confusing when it is introduced in this way. Even with a slow attack and release, the dynamic range of the signal is still being reduced. That's why I was using the tin can analogy. the reduction in dynamic range seems just as important as the automatic gain control concept.
Compression is essentially a super quick volume control.

Or to put it another way, a human riding a fader is a compressor with an incredibly slow attack and release time.

We don't "hear" the sound of relatively fast attack/release compression in the same way as we hear volume rides, so you're kind of correct - it's a bad way to explain to a novice what compression does, unless you add the caveat "but it doesn't really sound the same".

However, the idea of compression reducing dynamic range is kind of a myth, unless you're using look-ahead limiting or are compressing really slow attack signals. A compressor has a finite attack speed, which means a fast peak will always get through (unless you're using look-ahead) and then the compressor will clamp down.
Old 19th September 2011
  #8
Gear Nut
 

I guess I'm not explaining my initial question correctly or perhaps I'm a bit more ignorant than I thought was in regard to how compression works.

-----

Let me try it a different way.

1. I thought that a compressor ONLY applies gain reduction to the portion of the signal that is above the threshold, NOT the entire signal.

2. Turning down the volume fader reduces the gain of the "ENTIRE" signal, NOT just a portion of the signal.

That difference is why I thought the common analogy of : "the automatic engineer riding the volume fader" was misleading. Because with a compressor you are turning down just the part of the signal that exceeds the threshold, which is not the same as using the volume fader.

Please let me know if I'm thinking about this wrong. I'd really like to understand the difference correctly. Apologies for any ignorance on my part. Thanks.
Old 19th September 2011
  #9
The treshold determines when the amplitude of the signal (that of course means the "whole signal") is reduced, not "which part" of it. As it was said before, the only difference between compression and riding the fader is in the time domain.
Old 19th September 2011
  #10
Lives for gear
 

The other difference is that compression = dynamic attenuation + static amplification (make up gain) while "riding the fader" = dynamic attenuation and dynamic amplification

(assuming we're talking about downwards compression, which is the most commonly used)

But the distinction remains, compression (either downwards or upwards) does not dynamically attenuate and amplify the input level "at the same time", if that makes sense?
Old 19th September 2011
  #11
Yeah, that's what I meant when I wrote time domain.
Old 19th September 2011
  #12
Gear Guru
Quote:
Originally Posted by capproach View Post
I guess I'm not explaining my initial question correctly or perhaps I'm a bit more ignorant than I thought was in regard to how compression works.

-----

Let me try it a different way.

1. I thought that a compressor ONLY applies gain reduction to the portion of the signal that is above the threshold, NOT the entire signal.

2. Turning down the volume fader reduces the gain of the "ENTIRE" signal, NOT just a portion of the signal.

That difference is why I thought the common analogy of : "the automatic engineer riding the volume fader" was misleading. Because with a compressor you are turning down just the part of the signal that exceeds the threshold, which is not the same as using the volume fader.

Please let me know if I'm thinking about this wrong. I'd really like to understand the difference correctly. Apologies for any ignorance on my part. Thanks.
I think I see your problem. It's not like a de esser that only works on the HF portion of the signal.
The threshold is where the fader starts coming down. But once the signal crosses threshold the whole signal is affected. It is not just the same thing as a volume fader. It IS a volume fader.
Old 19th September 2011
  #13
Gear Nut
 

Quote:
Originally Posted by PRobb View Post
I think I see your problem. It's not like a de esser that only works on the HF portion of the signal.
The threshold is where the fader starts coming down. But once the signal crosses threshold the whole signal is affected. It is not just the same thing as a volume fader. It IS a volume fader.
Thanks, PRobb. Obviously my understanding is way off. My apologies.

I guess what has me stuck is that riding the volume fader doesn't "change the actual timbre of the sound." It just just changes the relative volume of the sound.

But, if you place a compressor on a snare drum and set a very fast attack, the transient will be greatly reduced and the actual sound of the snare will change dramatically.

How could that be accomplished with a guy riding the volume fader alone?

If anyone could help me have an "aha moment" I'd be incredibly appreciative.

The issue I'm having difficulty comprehending is that if the compressor is just turning down the volume, then how can it cause such great changes in timbre? Thank you kindly to anyone who has the patience for my ignorance. I know how to "use" a compressor, but obviously my understanding of how it is working to achieve the sound is SERIOUSLY flawed.
Old 19th September 2011
  #14
Gear Guru
Speed.
Someone said a human riding a fader is a compressor with an extremely slow attack and release time.
The compressor can work fast enough to alter the shape of the peaks.
Also, a lot of compressors are not linear. The LA 2 tends to reduce highs as it gets into heaver gain reduction. That's part of why people love them on vocals because so many singers get shrill when they get loud.
Old 19th September 2011
  #15
Gear Nut
 

Quote:
Originally Posted by PRobb View Post
Speed.
Someone said a human riding a fader is a compressor with an extremely slow attack and release time.
The compressor can work fast enough to alter the shape of the peaks.
Also, a lot of compressors are not linear. The LA 2 tends to reduce highs as it gets into heaver gain reduction. That's part of why people love them on vocals because so many singers get shrill when they get loud.
Bingo. Thanks, PRobb. That is completely making sense now. Thanks to everyone for weighing in. Forgive me for creating such a stupid thread. I'm probably the only person in the world who was thinking of this in such a backward way.

Thanks again!
Old 19th September 2011
  #16
Lives for gear
 
travisbrown's Avatar
Quote:
Originally Posted by capproach View Post
I guess I'm not explaining my initial question correctly or perhaps I'm a bit more ignorant than I thought was in regard to how compression works.

-----

Let me try it a different way.

1. I thought that a compressor ONLY applies gain reduction to the portion of the signal that is above the threshold, NOT the entire signal.

2. Turning down the volume fader reduces the gain of the "ENTIRE" signal, NOT just a portion of the signal.

That difference is why I thought the common analogy of : "the automatic engineer riding the volume fader" was misleading. Because with a compressor you are turning down just the part of the signal that exceeds the threshold, which is not the same as using the volume fader.

Please let me know if I'm thinking about this wrong. I'd really like to understand the difference correctly. Apologies for any ignorance on my part. Thanks.

Unless you are using a multi band compressor, there is no "portion" of the signal. You (or it) reduces the volume of the input signal when it exceeds a threshold.

Something I do (more often live than in the studio) is ride an EQ of a track to control the dynamics. I might pull between 1k and 2k on a piano track during vocal phrases, bumping it back up in between to let the piano pop through without raising the whole broadband level of the piano. This effectively controls a portion of the signal. Depending on how you look at is, this could be expansion rather than compression. you could set the same thing up by keying a multi band compressor to the vocals.

Consider this: an engineer like Bruce Swedien eschews compressors in favour of riding the fader. He's obviously using 1. top level musicians with excellent dynamic control where he doesn't have to use compressors to excessively "fix" performances, and 2. he's not using compressors as a colour effect as is popular today. He's not trying to get a compressed sound by squashing the whole track and bringing in the makeup gain; he's mimicking the original intent of a compressor which was to automate the volume.

We don't think of compressors as strictly such since we so often now use them as sound shaping tools - we don't always want them to be transparent and we want to hear them. That's where the tin-can analogy applies.
Old 19th September 2011
  #17
Gear Nut
 

Quote:
Originally Posted by travisbrown View Post
Unless you are using a multi band compressor, there is no "portion" of the signal. you (or it) reduces the volume of the input signal when it exceeds a threshold.

Something I do (more often live than in the studio) is ride an EQ of a track to control the dynamics. I might pull between 1k and 2k on a piano track during vocal phrases, bumping it back up in between to let the piano pop through without raising the whole broadband level of the piano. This effectively controls a portion of the signal. Depending on how you look at is, this could be expansion rather than compression. you could set the same thing up by keying a multi band compressor to the vocals.

Consider this: an engineer like Bruce Swedien eschews compressors in favour of riding the fader. He's obviously using 1. top level musicians with excellent dynamic control where he doesn't have to use compressors to excessively "fix" performances, and 2. he's not using compressors as a colour effect as is popular today. He's not trying to get a compressed sound by squashing the whole track and bringing in the makeup gain; he's mimicking the original intent of a compressor which was to automate the volume.

We don't think of compressors as strictly such since we so often now use them as sound shaping tools - we don't always want them to be transparent and we want to hear them. That's where the tin-can analogy applies.
well put and very helpful. Thank you for taking the time to reply.
Old 19th September 2011
  #18
Gear Nut
 

one more quick question regarding this topic if anyone can stand more of my questions.

Would it be wrong to think of the attack and release time on a compressor as kind of analogous at all to the attack and release on the ADSR envelope on an analog synth?

I'm asking because it seems that it is similar since we are using the "time domain" in both to shape the envelope of a sound.
Old 19th September 2011
  #19
Quote:
Originally Posted by capproach View Post
one more quick question regarding this topic if anyone can stand more of my questions.

Would it be wrong to think of the attack and release time on a compressor as kind of analogous at all to the attack and release on the ADSR envelope on an analog synth?

I'm asking because it seems that it is similar since we are using the "time domain" in both to shape the envelope of a sound.
Not at all - it's the same thing. given fast enough reactions, you could manually emulate an ADSR envelope with a volume control.
Old 19th September 2011
  #20
Lives for gear
Quote:
Originally Posted by timlloyd View Post
The other difference is that compression = dynamic attenuation + static amplification (make up gain) while "riding the fader" = dynamic attenuation and dynamic amplification

(assuming we're talking about downwards compression, which is the most commonly used)

But the distinction remains, compression (either downwards or upwards) does not dynamically attenuate and amplify the input level "at the same time", if that makes sense?
This makes sense, but some of the posts in this thread are confusing me... perhaps I am misunderstanding something.

Wouldn't a 3:1 ratio mean for every 3dB of input above the threshold you get 1dB of output above the threshold? That would not be attenuating the whole signal, and though it is not amplifying the signal below the threshold it would be raising the noise floor relative to the highest peaks above the threshold. I thought that was why it was considered a dynamics processor, and also assumed that is why some people like Bruce would opt to use fader rides instead of a compressor. Not because they are the same exact thing, but because you can accomplish the "gain control" without affecting the dynamic range.
Old 19th September 2011
  #21
Lives for gear
 

Quote:
Originally Posted by capproach View Post
... I guess what has me stuck is that riding the volume fader doesn't "change the actual timbre of the sound." It just just changes the relative volume of the sound.

But, if you place a compressor on a snare drum and set a very fast attack, the transient will be greatly reduced and the actual sound of the snare will change dramatically.

How could that be accomplished with a guy riding the volume fader alone?
..
A small point indeed to add though; I mentioned volume edits at the word' (or partial word) level. In general when we alter the envelope (would this be ‘micro level?) it is 'waveform distortion of sorts. But I can (and do heh do this with automation as well. The motivation in that case is it might be is less noticeably heard than the artifacts of the compressor.
Old 19th September 2011
  #22
Lives for gear
 
travisbrown's Avatar
Quote:
Originally Posted by capproach View Post
one more quick question regarding this topic if anyone can stand more of my questions.

Would it be wrong to think of the attack and release time on a compressor as kind of analogous at all to the attack and release on the ADSR envelope on an analog synth?

I'm asking because it seems that it is similar since we are using the "time domain" in both to shape the envelope of a sound.
Yes, no, sort of.

Attack = Sometimes explained as how long the signal is allowed to go over before the compressor kicks in, but this is technically wrong. The attack time is the amount of time it takes from the time the compressor kicks in to the time it reaches its intended gain reduction.

Release = the other end of the compression action: how long it takes for the signal to return to norm after dropping below the threshold.

On the point of the erroneous description of attack: some compressors are considered slow because of the time it takes them to kick in. I think this is part of the problem because we also refer to this as a slow attack time, but that's different than the attack setting on the compressor. Don't think of compressor setting as slow or fast; think of them as long and short. It's easier to visualize the shape of the sound this way.
Old 19th September 2011
  #23
Lives for gear
 
travisbrown's Avatar
Quote:
Originally Posted by timlloyd View Post
The other difference is that compression = dynamic attenuation + static amplification (make up gain) while "riding the fader" = dynamic attenuation and dynamic amplification

So my thinking on this is a little different from yours. I don't see you amplifying any signal when riding a fader below 0. It's all just degrees of attenuation. You aren't gaining without a subsequent gain stage.

I'd see this as

• compression = dynamic attenuation + static amplification (make up gain)
• "riding the fader" = dynamic attenuation and dynamic attenuation

And dynamic attenuation/amplification presumes aren't returning always to the same fader level as a baseline when the signal drops below threshold. To return to e.g. unity, it has the same net effect as your make up gain - a constant. Well, same effect except lower noise.
Old 19th September 2011
  #24
Gear Head
 

Quote:
Originally Posted by capproach View Post
Bingo. Thanks, PRobb. That is completely making sense now. Thanks to everyone for weighing in. Forgive me for creating such a stupid thread. I'm probably the only person in the world who was thinking of this in such a backward way.

Thanks again!
Just in case you're worried that you're the only one, I've been in roughly the same boat as you on this for a long time--understood basically how to use compressors, but confused on some of the details. Certainly, some confusing language has been used to describe how compression works. I do often hear the word "part" or "portion" used in reference to the signal, (i.e. "the part of the signal that exceeds the threshold...") when in fact that really means a moment in time within the signal. The compressor has no concept of portions of a signal (except of course multiband compression, which divides the signal into frequency bands), only how high the amplitude of the signal is at any given time.

Thanks everyone for the informative posts.
Old 19th September 2011
  #25
Gear Head
 

Quote:
Originally Posted by unitymusic View Post
This makes sense, but some of the posts in this thread are confusing me... perhaps I am misunderstanding something.

Wouldn't a 3:1 ratio mean for every 3dB of input above the threshold you get 1dB of output above the threshold? That would not be attenuating the whole signal, and though it is not amplifying the signal below the threshold it would be raising the noise floor relative to the highest peaks above the threshold. I thought that was why it was considered a dynamics processor, and also assumed that is why some people like Bruce would opt to use fader rides instead of a compressor. Not because they are the same exact thing, but because you can accomplish the "gain control" without affecting the dynamic range.
This is another thing that used to confuse me, too. To say "for every 3dB of input above the threshold you get 1dB of output above the threshold" is kind of true, but also kind of backwards. This math is actually used to determine by how much the signal is reduced. So really, using a 3:1 ratio, if the signal exceeds the threshold by 6dB, the compressor kicks in and subtracts 4dB, leaving the signal 2dB over the threshold. So it isn't that only the part of the signal that exceeds the threshold gets reduced--the whole signal is reduced, it just uses the threshold/ratio to determine by how much.
Old 19th September 2011
  #26
Lives for gear
Quote:
Originally Posted by mightypants View Post
This is another thing that used to confuse me, too. To say "for every 3dB of input above the threshold you get 1dB of output above the threshold" is kind of true, but also kind of backwards. This math is actually used to determine by how much the signal is reduced. So really, using a 3:1 ratio, if the signal exceeds the threshold by 6dB, the compressor kicks in and subtracts 4dB, leaving the signal 2dB over the threshold. So it isn't that only the part of the signal that exceeds the threshold gets reduced--the whole signal is reduced, it just uses the threshold/ratio to determine by how much.
I understand the distinction you are making. But to get the peak of the signal back to where it was you would have to add to the noise floor, decreasing the dynamic range of the signal to noise ratio, which is why I still think it is misleading to say compression is the same thing as a fader ride. Unless I'm just being stupid and missing something obvious. Are you saying the noise floor would not be raised?
Old 19th September 2011
  #27
Lives for gear
 

Quote:
Originally Posted by travisbrown View Post
So my thinking on this is a little different from yours. I don't see you amplifying any signal when riding a fader below 0. It's all just degrees of attenuation. You aren't gaining without a subsequent gain stage.
Yes you're right if we're talking about an actual channel fader. I should have been more specific, but the distinction I made wasn't really worth making.
Old 19th September 2011
  #28
Quote:
Originally Posted by capproach View Post
Why do practically all compression tutorials describe a compressor as an "automatic volume control" and say things like, "...however, the compressor can turn the volume down quicker and with more precision than the engineer ever could." It gives the impression that using a compressor is the same exact thing as volume automation.

Isn't calling a compressor an "automatic volume control" misleading? Isn't a tin can a better analogy? You squash the top down and it becomes smaller. If you bring the can back up to its original height (with make-up gain), the bottom will now be higher up too.

If a compressor was just an automatic volume control, then when the make-up gain was used, the signal would sound exactly the same as it did originally. There wouldn't be much of a point.

Does anyone else find the common general explanation of compression to be misleading or am I way off?
Key word, automatic.

To answer your practical question: no. Since the purpose of an 'automatic volume control' would presumably to keep the overall volume steady across varying input levels, the resulting modified signal would not "sound exactly the same as it did originally."

If one designed enough control parameters into a compressor, he could probably simulate automated gain riding. The degree of finesse one can exert with a compressor is directly related to the choices of control parameters. That's why it cracks me up to see all these latter day RE's going gaga over supposedly classic compressor boxes with just a few knobs/switches, seemingly selecting compressors on a per use basis -- while, if they were on their game, it seems like they might use a compressor with more control options and use their brains and ears to finesse that device's operation to fit the requirements of the moment.
Old 19th September 2011
  #29
Lives for gear
Attenuation is simply that. When a compressor attenuates a signal it reduces the entire signal at that moment by X amount. There isn't any realistic way to separate the portions of a signal above a threshold and treat them special. Over those X milliseconds when the compressor is attenuating the signal, all kinds of frequency waveforms representing the sounds of an infinite number of instruments can be passing through the attenuation. They all get treated the same at that moment. Just as if you had turned down an input trim or pulled down a fader.

This kind of goes back to my issue of people confusing input trims or attenuators with gain. The gain of an amplifier circuit is fixed. The attenuator only changes the amount of input voltage required to achieve some specified level of output.

If you set up a compressor so that it is always reducing or attenuating the signal, you can then add make up amplification to make the output similar in level to what the input was with the comp defeated. The net effect is that since you have turned everything up, but automatically turned down the loud parts, the quieter parts will seem louder in comparison. Just as if you had turned the whole thing up louder and somehow ignored the loud parts. By turning it up, you raised the level of the quiet parts. There is no dynamic processing bringing up the quiet parts in relation to anything.
Old 19th September 2011
  #30
Quote:
Originally Posted by unitymusic View Post
I understand the distinction you are making. But to get the peak of the signal back to where it was you would have to add to the noise floor, decreasing the dynamic range of the signal to noise ratio, which is why I still think it is misleading to say compression is the same thing as a fader ride. Unless I'm just being stupid and missing something obvious. Are you saying the noise floor would not be raised?
Whether it's a human riding gain or a device, the basic action is the same: either the signal gain goes up, goes down, or stays the same. What may differ are the factors provoking those changes in gain: what level triggers what action; how responses to signal gain change vary over time, etc. But the basic function is the same: gain change.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump