Logic Pro 8 A compressor experiment on a sunday afternoon

[Janne]

After reading a lot about how much better Logic 8's compressor is compared to previous versions I set out to try to mimic my to-go compresser, the Sonalksis SV315mk2.

I put the same drumloop on 2 tracks and put a gain plugin on one of them and phase inverted it so the sound totally canceled out.

I then took one of the presets on the Sonalksis (smooth) and adjusted the threshold to taste, then I inserted Logics Compressor on the other channel and started first by mimic the settings numerically.
I've not seen any info on what type of compressor the SV315 is suposed to emulate, but it was quickly apparent that it seems to be based on a optical compressor.

Now the sound started to cancel out heavily and after some adjustments to the parameter by ear I got it to almost totally cancel out. Something is going on differently in the lower registers, I could never get the low end attack of the Bass drum hits to cancel out completely. Also a small "peak" of the snare attack could be heard.
I should also mention that I had to set the Compressors Output Distortion to soft to get as close to the SV315 as possible.

Now comparing the sound of these settings on there own, they were very, very similar. I don't think I could blindfolded tell them apart although I think could notice that the SV315 "smashed" the snare sound differently, were as the snare through the Compressor sounded more "open".

At this point I realized that the new T-RackS 3 Opto Comp, is (obliviously) also based on a optical compressor, so decided to see how it would compare.
Going about it the same way as above I got it to cancel out to aprox the same leve as the first test. This one took away a little more of the bass drum attack when compared to both of the other contestants, so I got the feeling that it's somewhere "in-between" the Compressor and the SV315.

Two things that confused/irritated me during this experiment were:
1) the gain stages of the Opto Comp were strange...
To make this test work I had to either turn up its Input control by 12db, but then the meters basically got stuck at -20db and didn't move.
So to get it to perform as the other contestants, I had to set its output to +12db?

2) Metering...
The SV315 reported a gain reduction around -3db with peaks around -5db.
The Opto Comp were similar but off -3db, ie. -6db with peaks around -8db
And the stranges thing were that the Compressors meter didn't even move!
It's not that it isn't working, if I decreased the threshold I started to get the meter to react but then it of course didn't cancel out any longer...

*** A correction, I just noticed after posting, that in the Gain plugin I had it set to -3db, a leftover while trying to balance the differences between the SV315 and Compressor. Setting it to 0db put SV315 and the Opto Comp into a similar metering range, but Compressor still didn't register anything on its meter... ***

What is completely clear is that You can't trust the metering...
And Logics Compressor can sound as good as the competition.

Here are some snapshots of the metering and the settings, YMMV!

Janne

 

[Per Boysen]

Very interesting. Thanks for posting!

 

[Janne]

Update: I think I have found the difference between Compressor and the SV315 in the lower registers.

In the SV315 there is 4 filters styles that are used to filer the signal that triggers the compression. If I select the one that looks like this -_- I get the closer to what Logics Compressor does. It seems to use the low AND high registers to trigger.

At first I thought that Compressor didn't have anything like this but actually the Side Chain Filter parameters works even when no "external" side chain is used!
So I got fairly close to emulate the SV315 setting by using a HP filter at around 185Hz with a fairly hi Q value of 3.2, the gain doesn't seem to be used here...

Now I got a nice low end coming through similar to what the SV315 does!

Geee... do I feel nerdy..

I have attached the preset, download, remove the .txt and put it into Your Compressor preset folder if You would like to see how close I came...

Janne

 

[jlproductions]

Love this post. There is a lot going on in the Logic compressor that is not intuitive when you look at it. There are a lot of parameters there to deal with ( which is not a bad thing) and it is not always clear how to go about getting the sound you are looking for. So this kind of experimentation is interesting. In terms of the side chain filter you mean the 'extra' data side chain below is set to activity:ON right? Or did you mean that the filter has some effect when it is set to 'off'?
thanks
jamie

 

[Janne]

Thanks Jamie! I was maybe a bit vague, the filter must of course be set to on to be of any use...

I must say that I'm very impressed with Logics Compressor. It's very easy to mimic IK's Opto Comp and I downloaded the Demo of Stillwells Event Horizon, and I could almost cancel that one out too.
I wish I had some more upmarket compressors to test, this was really fun.

I think the problem with Compressors metering is that it will display the gain reduction but will add the make up gain to it, so it might not show any reduction even when heavy compression is going on just because some make up gain has been added. I would like to alway get a reading on how much reduction is going on regardless on the output of the compressor.

I must say that Opto Comp and Event Horizon were very easy to almost 99% cancel out. But the Sonalksis SV315 does somethings that the other compressors don't...

Two things that made it impossible to get as close as the others were:

The release time. The SV315 have what Sonalksis refer to as a Anti-Breathe mechanism that introduces a "hold" time on transient rich material independently of what the release time is set to.
If You download the Compressor preset You'll notice that I had to crank the release time up full to mimic the SV315 behavior on a (transient rich) drumloop, on other material it will probably behave differently.

The transients. The SV315 does something with the transients that I cant put my finger on, it might have something to do with saturation or something.
When I have Compressor set up so close to canceling out the SV315 I still get very short "peaks" where the bass drum and snare hits comes and the channel meter in Logic signals that the SV315 peaks att 5.1db more than the Compressors channel when Auto-Gain is applied. The SV315 manual warns for this, that transients can slip through when Attack time is set to higher values, and now with the make up gain added!

Logics Compressor doesn't do this, if I turn of Auto-Gain i can get almost full cancelation. This might explain why I often turn on the Auto-Gain on the SV315, it will bring out the transients even when a lot of compression is applied.

I also went ahead and tried to mimic the other presets from the SV315 and I noticed that when the SV315 were set to Type II the Opto emulation were not the closest one, I got fairly close with the VCA type but once again it's the transients that differs...

One thing is clear, Logics Compressor have very much to give if You just spend som time with it, I've sure learnt a lot about compressors during this test and Logics Compressor impressed me and I also got to realize what makes the SV315 differs from other compressors I've tried.

Janne

 

[Eli]

I hope no one minds me steering this very informative thread into a slightly different direction. Is it just me, or do others find the auto gain feature in Logic's compressor all but useless?

It seems to me that it works in a theoretical state that never actually exists. It boosts all the signal passing through the compressor up to the level it is set at. But it doesn't seem to account for the audio that doesn't get processed by the compressor.

So, if for example you have an attack time at any value higher than 0ms, you will get clipping. Because the gain is being applied to the signal that isn't getting caught by the compressor the same way it treats signal that is gettign processed by it.

So, unless the attack time is set to a perfect 0ms, which it almost never is in the real world, the auto gain seems to cause more problems than it solves. I know it can be compensated for at the limiter stage, but still. How are people using the auto gain in the "real world"?

 

[orjankarlsson]

I hope no one minds me steering this very informative thread into a slightly different direction. Is it just me, or do others find the auto gain feature in Logic's compressor all but useless?

It seems to me that it works in a theoretical state that never actually exists. It boosts all the signal passing through the compressor up to the level it is set at. But it doesn't seem to account for the audio that doesn't get processed by the compressor.

So, if for example you have an attack time at any value higher than 0ms, you will get clipping. Because the gain is being applied to the signal that isn't getting caught by the compressor the same way it treats signal that is gettign processed by it.

So, unless the attack time is set to a perfect 0ms, which it almost never is in the real world, the auto gain seems to cause more problems than it solves. I know it can be compensated for at the limiter stage, but still. How are people using the auto gain in the "real world"?

Yeah, I never use the autogain function. And btw: the built in Logic compressor?.....it ROCKS!

 

[jlproductions]

I haven't forgotten this post but just have been busy. However yes, autogain is useless!
However one interesting thing that gave me a newfound respect for the Logic compressor was as I have been doing some surround mixing I really enjoyed Logic's Surround compressors. They are basically the same compressors patched up into 6 or more channels - however the interface is changed and the display is now plasma meters. So one gets a whole new visual feedback which is nice. When you go back to the regular one, the sound is similar without the visuals. And what you are describing above reminds me of my Tube Tech compressors which do a whole bunch of stuff to the sound when there is absolutely nothing on the meters...gotta run, best, jamie

 

[tablatom]

Hi,
great thread, this is my first post here. I googled LOGIC COMPRESSOR AUTO GAIN and got this thread. Just what i am looking for.
I really dont understand when using the autogain when the threshold is lowered the output of the compressor increases.

Someone said in another forum that the autogain compensates for the gain reduction. Well that would be great if the output of the compressor stayed the same.

So i also dont use it.

I also like the compressor very much and i sometimes use its limiter too.
tOM

 

[Peter Ostry]

Someone said in another forum that the autogain compensates for the gain reduction. Well that would be great if the output of the compressor stayed the same.

So i also dont use it.

Autogain is a problem. What do you want to get "autogained", RMS or Peak? This depends on the usage. If you want to leave the transients as they are, then you go for RMS. Or you want to stay within a certain peak level, then you aim for peak. If you find that autogain works better with other compressors, then you are lucky and the developer chose the right method for you.

The first thing I do with a compressor is turning autogain off. Then I make my settings and adjust the final level for my needs.

 

[Eli]

If you find that autogain works better with other compressors, then you are lucky and the developer chose the right method for you.

The first thing I do with a compressor is turning autogain off. Then I make my settings and adjust the final level for my needs.

Same here. The more different compressors I use, the more I see how the autogain feature doesn't really work properly, and works differently, on each. And I have found too that adjusting the make up gain manually, in conjunction with a Limiter, usually works best for me.

 

[LSchefman]

Truly cool post! Thanks!

Having the same plugs myself, the comparison is MOST interesting.

 

[tablatom]

If you find that autogain works better with other compressors, then you are lucky and the developer chose the right method for you.

The first thing I do with a compressor is turning autogain off. Then I make my settings and adjust the final level for my needs.

Same here. The more different compressors I use, the more I see how the autogain feature doesn't really work properly, and works differently, on each. And I have found too that adjusting the make up gain manually, in conjunction with a Limiter, usually works best for me.

Indeed indeed.
I wonder if the chaps and chapesses at Apple will look at this?