Posts: 518
Threads: 11
Joined: Dec 2013
Reputation:
31
02-17-2021, 01:09 AM
(This post was last modified: 02-17-2021, 01:17 AM by sunrat.)
I just found this post at Gearslutz in which JonVAudio the FirComp dev is posting prolifically. He explains what the program dependency slider does to the release:
Quote:Hello, yes I agree the program dependency slider is not immediately obvious, that's because it was added later (early versions had it fixed at 100%). I think new versions of the GUI probably need to be taller
With PD at 0%, the release control is fairly standard, except for the fact that the release "curve" starts off much, much faster, and then slows down a lot as it reaches its target. The release is a "release to target" type, and so even with 0% PD it is kind of "program dependent" and should sound fairly natural, but only cares about the current GR and the target GR, and ignores history (as bmanic said).
With PD at 100%, the release is the same shape and type, but can be slowed down by up to 3x as much, depending on the recent history of gain reduction as compared to the current gain reduction. This is continuously varying, ie it's adjusting the release time constants every sample.
So you are getting what you dial in, but that value is used as a starting point. The design goal for the release stage was transparency
The reason it sounds smooth at extreme settings, is because of the novel approach used to reduce discontinuities while tracking the envelope of the input signal (and then the lookahead optionally adds more gain smoothing). It took me a while to implement this part of the chain efficiently, and not get it to use too much CPU use at high sample rates
https://www.gearslutz.com/board/showpost...tcount=176
PS: Gearslutz is really annoying. Haven't been there much lately, or ever.
Mixbus 32C, Debian Bookworm/KDE, EVE SC205 + ADAM Sub 8 monitors, Soundcraft Compact 4, M-Audio 2496, i5 6500, 16GB RAM, WD Blue SSD 1TB, 48" LG OLED, other stuff.
Work as house engineer at a popular venue in Melbourne AU. On a quest for the holy grail, the perfect amount of cowbell.
Posts: 518
Threads: 11
Joined: Dec 2013
Reputation:
31
02-17-2021, 08:53 AM
(This post was last modified: 02-17-2021, 08:54 AM by sunrat.)
I was just reading LSP manual and noticed the Compressor has separate attack threshold and release threshold controls
https://lsp-plug.in/?page=manuals§io...sor_stereo
Mixbus 32C, Debian Bookworm/KDE, EVE SC205 + ADAM Sub 8 monitors, Soundcraft Compact 4, M-Audio 2496, i5 6500, 16GB RAM, WD Blue SSD 1TB, 48" LG OLED, other stuff.
Work as house engineer at a popular venue in Melbourne AU. On a quest for the holy grail, the perfect amount of cowbell.
Posts: 681
Threads: 44
Joined: Sep 2013
Reputation:
36
(02-17-2021, 08:53 AM)sunrat Wrote: I was just reading LSP manual and noticed the Compressor has separate attack threshold and release threshold controls
https://lsp-plug.in/?page=manuals§io...sor_stereo
Cool! :-) I might reopen the project and test it. It's not what I'm looking for, but the default slopes and controls might make it sound perfect both when the drums are going on and when the gongs are taking over again when the drums are fading out.
Posts: 120
Threads: 29
Joined: Oct 2013
Reputation:
2
02-18-2021, 01:39 PM
(This post was last modified: 02-18-2021, 01:43 PM by Flagfoot.)
Jostein I'm sorry for not understanding you earlier
May I take another shot to redeem myself? If this doesn't help you, maybe it will help someone else.
This one is a workaround I use:
1. I want to duck some bells a bit when the hand drum hits.
2. Using the hand drums as a key, I get the attack times I want on the bells, but no matter what, the release is just too long.
3. Since I don't have hysteresis control on any of my compressors, I copy the hand drums to a new key-only track.
4. I put my favorite transient control plugin on the key-only track and dial back the sustain, but leave the attack as-is.
5. That way, my key allows the compressor to release sooner. At least if everything is working well.
Thanks for your patience!
Posts: 681
Threads: 44
Joined: Sep 2013
Reputation:
36
(02-18-2021, 01:39 PM)Flagfoot Wrote: Jostein I'm sorry for not understanding you earlier
May I take another shot to redeem myself? If this doesn't help you, maybe it will help someone else.
This one is a workaround I use:
1. I want to duck some bells a bit when the hand drum hits.
2. Using the hand drums as a key, I get the attack times I want on the bells, but no matter what, the release is just too long.
3. Since I don't have hysteresis control on any of my compressors, I copy the hand drums to a new key-only track.
4. I put my favorite transient control plugin on the key-only track and dial back the sustain, but leave the attack as-is.
5. That way, my key allows the compressor to release sooner. At least if everything is working well.
Thanks for your patience!
Good Stuff! :-)
You have nothing to be sorry about. I was a part of that problem, I did not express myself clearly enough and for every answer, I came up with here and there, I got even worse when attempting to use a new angle when I wrote something again! :-)
I have already finished the project (and explained what I did somewhere) and delivered a CD master (DDP-file) and MP3s to the client. I was also thinking about copying an altered ghost track for controlling the gongs, but what I did was sufficient enough for me to turn up the gongs earlier after the drums faded. This is was I was considering if I would use a copy and why:
The ACE Compressor sounded perfect during the drumming, that's why I still used it (and it's not given that an extra Release Phase Duration Control would do the trick):
- I would have made a copy of the drum track that was not sent to the master, it would only control the Gongs dynamic.
- I would let the drums ring out a little bit earlier (cutting and moving the end around 50-70 ms). That way the gongs would come just a little bit earlier while it still sounded perfect when the drums were doing their stuff.
I'm a very patient person - That's an engineer's and sidekick's best asset! :-)
Posts: 120
Threads: 29
Joined: Oct 2013
Reputation:
2
Now we're speaking the same language.
If I remember correctly, GML used to make a hardware compressor with hysteresis control for problems like this.
Posts: 233
Threads: 13
Joined: Jan 2016
Reputation:
15
I will not comment on compressors. I think that horse has already been beaten to death However, In the hopes of less confusion, misinformation and misconception, please view this article. https://music.tutsplus.com/tutorials/the...-audio-953
I am posting it because it gives a simple easy to understand explanation of what audio compression is and is designed to do. I clarifies terminology. When reading this article understand that it is basically talking about analog compressors, so keep that in mind. It touches on plugin compressors briefly. I think that if we all start with the basics of compression and understand the starting point than we can have much better discussions (about something other than compression ).As always, just my opinion.
Peace to all
Alex
Win10 64 i5 3330 Quad Core, AVL/MXE i5, MB 3-9, MB32C 3-9, Tascam US 20x20(2), Tascam 388, Alesis HD24, Alesis ML 9600(2), A&H GL2400, Soundcraft Studio Spirit 24, Roland Integra7, Roland S-50, M-Audio Hammer 88, ART/ MPA Gold/ TPSII/Pro Channel(2)/Pro VLA(3), lots of tubes
|