All times are UTC+01:00




Post new topic  Reply to topic  [ 9 posts ] 
Author Message
PostPosted: Wed Jan 04, 2017 1:27 pm 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
Is it possible to get the same gain reduction with the SB compressor for sine waves with the same amplitude but different frequencies?

With attack/release shape = 1 it seems that setting peak mode = release_time/attack_time gives an approximation for that or maybe it's just a coincidence.


Top
   
PostPosted: Fri Jan 06, 2017 6:36 pm 
User avatar

Joined: Tue Mar 17, 2009 2:56 pm
Posts: 4151
It should work like that already. it's WB. However check sidechain PEQ is it ON with some settings!

_________________
control point
control point2


Top
   
PostPosted: Sat Jan 07, 2017 12:03 pm 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
Quote:
It should work like that already.
That's only true for attack_time = release_time because auto shapes will be the same in that case, too. However with attack_time != release_time attack and release behave differently enough to settle far away from the theoretical target.
Setting shapes = 1 and peak mode = release_time/attack_time or shapes = 500 and peak mode = 1 seems about to work.
I've only tested analog RMS mode with auto shape on so far.

Here are plots for easier comparison. They show the level computed from input and output (samples vs percentage). The used signals are sine waves with amplitude 1 and frequency as stated below, so they all should've produced the same output. The SB compressor uses threshold = -20 dB, ratio = 2, attack time = 20 ms, release time = 60 ms, knee = 0 dB, mode = RMS and auto fix shape = on. The rightmost plot is from a model compressor.
Image


Last edited by \_/ on Sun Jan 08, 2017 9:27 am, edited 1 time in total.

Top
   
PostPosted: Sun Jan 08, 2017 1:43 am 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 11185
With an attack time that's close to the frequency of a tone (100 Hz -> 10 ms, vs 20 ms), it kinda makes sense that the level it reaches can be a bit different (for example, the release could get an effect on the quieter parts of the waveform). The 1000 and 10000 tones are pretty close, not sure why they are sligtly different. So I would expect that you get better results if you increase both the attack and release times (by the same amount, relatively, so for example 200 and 600 instead of 20 and 60). In the theoretical model the lowest tone is also on top, the difference is just smaller.

I'm also not sure what happens with the 10 kHz tone at Shape=500 - that seems odd.


Top
   
PostPosted: Sun Jan 08, 2017 9:23 am 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
As requested the same plots but with (attack, release) = (200 ms, 600 ms). They look pretty much the same.

Why does the leftmost plot settle at a different target? The input is the same everywhere.

Image


Top
   
PostPosted: Wed Feb 22, 2017 9:17 am 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
I re-did the first measurement now with detection type = peak (all other settings same as here). Not sure if this is looking good. :?
Image


Top
   
PostPosted: Sun Jun 11, 2017 10:04 pm 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
Tested new real RMS mode as of ST 8.40, settings as above. No suprises.Image


Top
   
PostPosted: Tue Jun 13, 2017 11:45 am 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 11185
Looks better than before!


Top
   
PostPosted: Fri Jul 28, 2017 9:12 am 

Joined: Sun Feb 03, 2013 2:39 pm
Posts: 333
.


Top
   
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 9 posts ] 

All times are UTC+01:00


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Limited