compression question - release

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
Post Reply
samsome
Posts: 364
Joined: 18 Dec 2015

03 Dec 2019

PLease confirm if this statement is true or not "the compression still works and compresses even when the signal falls below the threshold"

if not...i don't see why release would be doing anything to the signal...

if yes please further explain so i can understand

thank you

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Dec 2019

Compression only works above the threshold, assuming downward compression (the most common).

What confuses folks is the idea that attack or release only happens when the audio actually crosses the threshold. Instead, attack and release affect any CHANGE above the threshold. So if the signal goes above the threshold and holds steady for a few seconds, and then jumps up further, attack is involved. Same if the signal holds above the threshold and them drops quickly (even if it doesn't go BELOW the threshold), release will still be active.

Also worth nothing that attack and release are time constants, which is in contrast to how an envelope works in a synth (where attack and release are fixed in time). That is to say, attack and release in dynamics devices are RATES, not TIMES. For example, the attack in the SSL channel compressor is listed as "ms per 20 dB." So if the attack is listed as 3ms, and the amount of compression is currently 10 dB, then you have an attack TIME of 1.5ms. Make sense?
:)
Selig Audio, LLC

samsome
Posts: 364
Joined: 18 Dec 2015

03 Dec 2019

then how do you explain this image.

i think we can see that compression is active below the threshold on this one and releases

Image

samsome
Posts: 364
Joined: 18 Dec 2015

03 Dec 2019

if you were to draw a threshold line on that pic where would the threshold be

wouldn't be here as per red line?

Image

User avatar
TritoneAddiction
Competition Winner
Posts: 4219
Joined: 29 Aug 2015
Location: Sweden

03 Dec 2019

samsome wrote:
03 Dec 2019
then how do you explain this image.

i think we can see that compression is active below the threshold on this one and releases

Image
Not sure what you mean. To me it seems like this picture confirms everything Selig just explained.

User avatar
TritoneAddiction
Competition Winner
Posts: 4219
Joined: 29 Aug 2015
Location: Sweden

03 Dec 2019

samsome wrote:
03 Dec 2019
if you were to draw a threshold line on that pic where would the threshold be

wouldn't be here as per red line?

Image
The treshold could be set higher than that red line. We can't really see your settings. But even if the threshold were set at that red line, it still doesn't mean that everything automatically will be compressed down to that level. The amount of compression is set with the ratio control, which could be on a fairly light setting.

User avatar
guitfnky
Posts: 4408
Joined: 19 Jan 2015

03 Dec 2019

is that image representative of actual audio? that’s kind of important.
I write good music for good people

https://slowrobot.bandcamp.com/

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

03 Dec 2019

When the signal falls below the threshold, its envelope may still be above it, so the "compression" goes on. That's if the compressor you are talking about is envelope-based. Compressors are very diverse, they can implement a lot of different schemes.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Dec 2019

@samsome:
This may purley be semantics, but you asked "is it still compressing the signal" after the signal falls below the threshold, and my answer would be "no, it's releasing from previous compression". Of course, the compressor will still be in the release stage even after the signal falls below the threshold. In that sense it's still active, if that's what you're asking.

Sorry if I confused the issue with my additional points - if you had started with the graphic image in your first post I may have answered the question a little differently!
Selig Audio, LLC

User avatar
Benedict
Competition Winner
Posts: 2747
Joined: 16 Jan 2015
Location: Gold Coast, Australia
Contact:

03 Dec 2019

Image

If Threshold were at the Red Line as seems reasonable (one wonders why the image has no guide/legend) then it seems to me that:

1. Audio is playing below Threshold unmolested by the Compression
2. A sudden loud signal appears
3. The Attack phase gets underway and
4. Clamps to its Ratio - assuming 2:1
5. Loud signal suddenly stops - falls below Threshold
6. The Compressor is still in for every 2db in, let 1db out mode
7. This will keep going until the Release phase finishes (in a nice linear fashion it appears)

As a result, the sudden "offness" of the signal catches the slow Compressor by surprise (all machines can only ever work on what has happened - which is why automatic cars can't change down before a hill gets steep) so it can actually "bite" below the Threshold in such a case.

This is a somewhat rare case as few instruments are that ON/OFF. Also, you may decide to shorten the release of the Compressinaor.

More importantly, there is the psychoacoustics of the situation where a sudden LOUD sound will cause the human brain to do this very thing you see here as it "flinches" and takes a moment to recover and open its eyes (ears) again. During that time, the world is actually muted to some extent.

The workings of Compression seem nonsensical a lot of the time, but this is more to do with our thinking than that of compressors. Compression is pure logic, only it behaves in a machine style of thinking which is not the same style of fuzzy thinking we humans prefer (I want it to be therefore it is – or should be).

This last is where Compression truly gets interesting and becomes your friend in Storytelling as you can use compressors (and their various relatives) to not just "flatten" dynamics but shape them to better express the story of one sound, a section, or the whole track.

:-)
Benedict Roff-Marsh
Completely burned and gone

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 22 guests