Compression Listening Comparisons: LiquidMix,Waves, others

[TranscendingM]

Hello everybody, been a while since I've logged in. Recently with the advent of Liquid Mix, this type of external DSP processing is gaining more steam. I know UAD's been around for a little while.

I've also noticed though, alot of listening tests floating around but the tests entail either comparing these "high end" type of external plugins like LM, with other "high-end" plugins such as UAD and/or native "high-end" plugins such as the waves stuff. I want to compare however, if these "high-end" plugins do indeed stand up to what they are, compared to plugins one may not exactly pay so much for. So I wanted to set up a little test here of a short drum mix that compares LM with a popular waves native compressor RComp and a few others that are perhaps not as costly or even free.

The settings on the compressors have all been matched. The purpose of this is to observe how each compressor reacts with a signal according to a shared specific setting. In regards to the LM compressor, I used one of the high-end hardware emulations, Neve 33609.

Just to recap, we should be looking for and evaluating: gain control, density/forcefulness/"meat", apparent volume, openess, transient quality, cohesiveness, smearing (bad type), "bigness" (low-end translation). And of course, this all against the unprocessed track with the foundation of once again the shared specific setting. The gain control factor I will have to reveal at the end and we'll be able to tell by the amount in dB that I indicate was needed to match output levels.

Note: These are all wav files by the way (16bit 44.1Khz)

Unprocessed
Comp1
Comp2
Comp3
Comp4
Comp5
Comp6

Have fun

[PeterSturm]

I don't like 1 and 4. 1 changes the sound of the snare strange and 4 reacts too slow in the beginning.

With these settings 5 sounds best to me.

[Pistolpete]

The only one I really like is 5 also... 1 changed the snare too much, 2 put too much slap into the snare and made it sound unreal, 3 sounded cheap...lol..and too much slap in the snare, 4 brought the cymbals out more in front, but the snare fell behind and soft, not enough body to the lower end on the snare(the same for most of them), 5 to me was the most natural sounding, and the decay was a little more clear in the snare, 6 wasn't too bad, nice and punchy snare, but still unnatural sounding.

5 is the winner in my book....

[PeterSturm]

On second listening 6 sounds good too.

Can you please share us the settings you used on the compression, so we can may be add some other compressors here?

[n8tron]

also like 5 and 6, they are the most clean but brought nice sustain to the cymbals, and didn't kill the snare like 1.. i think 3 did. 2 and 4 brought a lot more color in my opinion, 2 was really compressing... and 4 had brought that rush of volume on the cymbals half way between their attack and release... which I kinda like in some situations, but not most. 1 and 3 were most dull and forgettable. 4 sounds like an 1176 to me, because of that release on the cymbals.

[TranscendingM]

Originally Posted by PeterSturm

On second listening 6 sounds good too.

Can you please share us the settings you used on the compression, so we can may be add some other compressors here?

Sure. Here you go:

Threshold: - 30
Ratio: 3:1
Attk: 10 ms
Release: 200 ms
Gain: Here's the kind of tricky part. On the reference comp, which I can't reveal just yet, the make up gain was set at +4. This gave me a reference average peak of between -3.0 dbFS to -1.9 dbFS with this particular track. So you can use that range to set the gain on other compressors. Take note of that gain for later on when revealing the compressors. The purpose of this is to evaluate the gain control aspect of each compressor, something that gets overlooked.

This is good stuff guys, keep it coming. I like the descriptions of what each one is doing in your perspective rather than just "I like such and such..." and that's it. It's more helpful this way with the details you're giving.

[n8tron]

I'm really curious what the compressors were, can you either pm me or post?

[PeterSturm]

Treshold -30? You mean -13?

[PeterSturm]

Here are two of my favorite comps. I used your settings (except the treshold, obviosly) and I matched the level of the peak with your comp 5.

Comp 7
Comp 8

By the way, I think that your original file will benefit more from a little bit faster rls.

[TranscendingM]

Originally Posted by PeterSturm

Treshold -30? You mean -13?

Hi Peter, thanks for the clips.

But, no the threshold is -30. I wanted a threshold to affect most of the signal but kept the ratio lower to keep the compression a bit more transparent than usual. So, if you can, post your compressors with the same threshold so we can compare more accurately.


I will post the compressors in good time. I would just like to get a few more perspectives and opinions. I thought more of the gearslutz loved listening tests!

[PeterSturm]

If I post the compressors with -30 treshold, the reduction will be around 16 db in the peaks which wont sound like your clips! To me your clips sound with around 4db of redutcion.

[TranscendingM]

Originally Posted by PeterSturm

If I post the compressors with -30 treshold, the reduction will be around 16 db in the peaks which wont sound like your clips! To me your clips sound with around 4db of redutcion.

Pete, if you take a look at my first post I had mentioned that I used a certain amount of gain make-up for each compressor. This is part of the test: gain control/Make-up per compressor and also the character of each compressor with more gain coming from it. That is why I stated an average peak range that you should try to match with the gain make-up (if needed) on the particular compressor itself. And once again, take note of the gain you apply so that at the end when we can reveal the compressors we also reveal these gain make-up values.

[beachhunt]

Originally Posted by TranscendingM

Pete, if you take a look at my first post I had mentioned that I used a certain amount of gain make-up for each compressor. This is part of the test: gain control/Make-up per compressor and also the character of each compressor with more gain coming from it. That is why I stated an average peak range that you should try to match with the gain make-up (if needed) on the particular compressor itself. And once again, take note of the gain you apply so that at the end when we can reveal the compressors we also reveal these gain make-up values.

Excellent inclusion, in my opinion. When I get home I'd be glad to add a few contenders and post my (humble) thoughts on your files.

Thanks for getting this started

Edit:
2, 3, 5, and 6 seemed to have the most character, with 2 and 5 having the most sustain and 3 being somewhat pumpy. Something about #4 sounds like a URS compressor to me, in a bad way.

If I had to guess what was what, I'd guess #3 is the LM, #5 is the 33609, #4 is a URS 1980, #6 is Waves RComp, and whatever #2 is is not too shabby.

[TranscendingM]

Originally Posted by beachhunt

Excellent inclusion, in my opinion. When I get home I'd be glad to add a few contenders and post my (humble) thoughts on your files.

Thanks for getting this started

Thanks alot, that sounds great! Add as many as you possibly can...

[n8tron]

Originally Posted by PeterSturm

Here are two of my favorite comps. I used your settings (except the treshold, obviosly) and I matched the level of the peak with your comp 5.

Comp 7
Comp 8

By the way, I think that your original file will benefit more from a little bit faster rls.

I liked your number 7, but not the later, the cymbals were weakened, on 8 and enhanced on 7, in my opinion.

great thread btw.

[PeterSturm]

Originally Posted by TranscendingM

Pete, if you take a look at my first post I had mentioned that I used a certain amount of gain make-up for each compressor. This is part of the test: gain control/Make-up per compressor and also the character of each compressor with more gain coming from it. That is why I stated an average peak range that you should try to match with the gain make-up (if needed) on the particular compressor itself. And once again, take note of the gain you apply so that at the end when we can reveal the compressors we also reveal these gain make-up values.

There is some misunderstanding here. I am really sorry if it's my fault.
To me, your files sound like compressed with about 4-5 dbs of reduction (on the peaks). If I turn the treshold to -30 then the reduction will be about 16 dbs which sounds totally different from your compressors. With smashed hats and cymbals and lost clarity.
Here the make up gain doesn't any matter.
It will be best if you check how much is your compressor reducing in the strongest peaks.

Cheers, dude! :)

[PeterSturm]

Originally Posted by n8tron

I liked your number 7, but not the later, the cymbals were weakened, on 8 and enhanced on 7, in my opinion.

great thread btw.

Yeah, actually comp 7 really surprised me with his work in the highs.

Here is a mini riddle from me. Guess which compressor is Comp7! :) I'll help you, telling you that it's a more complex plug-in.

[Absolute]

Originally Posted by PeterSturm

There is some misunderstanding here. I am really sorry if it's my fault.
To me, your files sound like compressed with about 4-5 dbs of reduction (on the peaks). If I turn the treshold to -30 then the reduction will be about 16 dbs which sounds totally different from your compressors. With smashed hats and cymbals and lost clarity.
Here the make up gain doesn't any matter.
It will be best if you check how much is your compressor reducing in the strongest peaks.

Cheers, dude! :)

Thats because you cant share threshold settings without knowing the input volume...and not knowing that will affect everything

Thats why saying how much gain reduction YOUR getting on the loudest peaks is the right information..but the poster is not hearing you

[PeterSturm]

Yeah, my point! :)

[knerd]

I am dying to know!

Originally Posted by TranscendingM

Hello everybody, been a while since I've logged in. Recently with the advent of Liquid Mix, this type of external DSP processing is gaining more steam. I know UAD's been around for a little while.

I've also noticed though, alot of listening tests floating around but the tests entail either comparing these "high end" type of external plugins like LM, with other "high-end" plugins such as UAD and/or native "high-end" plugins such as the waves stuff. I want to compare however, if these "high-end" plugins do indeed stand up to what they are, compared to plugins one may not exactly pay so much for. So I wanted to set up a little test here of a short drum mix that compares LM with a popular waves native compressor RComp and a few others that are perhaps not as costly or even free.

The settings on the compressors have all been matched. The purpose of this is to observe how each compressor reacts with a signal according to a shared specific setting. In regards to the LM compressor, I used one of the high-end hardware emulations, Neve 33609.

Just to recap, we should be looking for and evaluating: gain control, density/forcefulness/"meat", apparent volume, openess, transient quality, cohesiveness, smearing (bad type), "bigness" (low-end translation). And of course, this all against the unprocessed track with the foundation of once again the shared specific setting. The gain control factor I will have to reveal at the end and we'll be able to tell by the amount in dB that I indicate was needed to match output levels.

Note: These are all wav files by the way (16bit 44.1Khz)

Unprocessed
Comp1
Comp2
Comp3
Comp4
Comp5
Comp6

Have fun

[TranscendingM]

Originally Posted by Absolute

Thats because you cant share threshold settings without knowing the input volume...and not knowing that will affect everything

Thats why saying how much gain reduction YOUR getting on the loudest peaks is the right information..but the poster is not hearing you

The input volume is whatever the track is. We have a standard to work through any compressor in this so it's a given factor across everytime. I'm more interested on how the compressor controls a signal, within a range, give or take .db on the peaks. I want the gain make-up to be adjusted by this approximation so to understand how each compressor is controlling the signal on a particular peak range. Having that said, the only factors needed were the control settings given earlier. From there, approximate the gain make up to match the range given earlier. That's the point, to observe the rest of the signal behavior after the peak range has been more or less matched using that particular compressors' gain make up. The revelation of gain make up values will then be considered at the end.

[Gear Tramp]

Originally Posted by TranscendingM

The input volume is whatever the track is. We have a standard to work through any compressor in this so it's a given factor across everytime. I'm more interested on how the compressor controls a signal, within a range, give or take .db on the peaks. I want the gain make-up to be adjusted by this approximation so to understand how each compressor is controlling the signal on a particular peak range. Having that said, the only factors needed were the control settings given earlier. From there, approximate the gain make up to match the range given earlier. That's the point, to observe the rest of the signal behavior after the peak range has been more or less matched using that particular compressors' gain make up. The revelation of gain make up values will then be considered at the end.

The input volume will change the way the compressor reacts & sounds.

Use the same input signal for all comparisons. End of story.

[TranscendingM]

Originally Posted by Gear Tramp

The input volume will change the way the compressor reacts & sounds.

Use the same input signal for all comparisons. End of story.

That's exactly the case here. Same input signal.

[beachhunt]

Well, I applied the settings you posted, and -30dB does seem a bit low for the results in the original post. I got about -8dB reduction, and given that fairly fast release I ended up with some serious pumping... Sounds nice to me, but doesn't sound like any of the other files.

Along those lines, you mentioned one of the compressors you used was the UAD 33609 which doesn't happen to have 200ms as an option for release, nor does its threshold go below -20. Can we get a double check on all of the above please?

Anyway, here's a mystery compressor with threshold around -30dBFS, 3:1 ratio, 10ms attack, 200ms release, makeup gain to match most of the other files. It sounds much more compressed than any of the other 8 so far, so maybe I'm missing something.

BH

[TranscendingM]

Originally Posted by beachhunt

Well, I applied the settings you posted, and -30dB does seem a bit low for the results in the original post. I got about -8dB reduction, and given that fairly fast release I ended up with some serious pumping... Sounds nice to me, but doesn't sound like any of the other files.

Along those lines, you mentioned one of the compressors you used was the UAD 33609 which doesn't happen to have 200ms as an option for release, nor does its threshold go below -20. Can we get a double check on all of the above please?

Anyway, here's a mystery compressor with threshold around -30dBFS, 3:1 ratio, 10ms attack, 200ms release, makeup gain to match most of the other files. It sounds much more compressed than any of the other 8 so far, so maybe I'm missing something.

BH

Well, that is the interesting part now isn't it

I promise these are the settings! lol This is why I wanted to go with gain-make up because now we're getting into the territory of character of ech compressor. Knowing discrepencies in how they each sound would eventually arise. With the make up adjusted to an average peak range, with a threshold such as -30db we get to hear different behaviors.
Regarding the 33609 - well - I can't say anymore or it'll give it away somewhat ahahah. What I can say is, you may be assuming "UAD"
But yes, I got 200ms on it OR .2 Seconds, if that's not another hint lol.

Thanks for the clip by the way. Some interesting pump to the cymbals in that one. Good stuff

[jeronimo]

downloading now...

[beachhunt]

Originally Posted by TranscendingM

Regarding the 33609 - well - I can't say anymore or it'll give it away somewhat ahahah. What I can say is, you may be assuming "UAD"
But yes, I got 200ms on it OR .2 Seconds, if that's not another hint lol.

Ah, I just re-read the original post and I misunderstood that bit. You said you used the Neve 33609 preset in the LM, I somehow turned that into using the 33609 plugin. Carry on