New Master Bus Compressor RE VS MClass Compressor

[zero-13]

I searched here but could not find any mention of it, anyone know if the new RE version of the Master Bus Compressor is intended to replace the MClass compressor?

Has anyone compared them that can share a seasoned producer's perspective and insights?

I'm several decades from having trained ears or even functional ears for that matter, if there is a difference I can not tell.

My imagination makes me hear the new RE MBC as sounding "smoother" / "gentler" / "slightly less harsh" with comparable settings... I don't trust my ears and feel it's just some cognitive bias of the MBC being a new thing?

For all I know the new RE MBC is just the MClass in new clothes?

[Kalm]

It's a typical Digital Compressor versus a modeled Analog G Compressor. Two different types and mechanisms. One won't replace the other but your usage can be dependent on which favors you.

If you want to learn more, simply go to YouTube and search G Compressor and you'll get all the info you need on what it is (and why its in the mixer section as well).

[orthodox]

For all I know the new RE MBC is just the MClass in new clothes?

The MBC is indeed almost equivalent to MClass Compressor in "Soft-Knee" mode, and at the same time to the Mix Channel compressor at some settings. Though, MClass can't do more than ratio "4" in terms of the MBC (Mix Channel one can). And the MBC can do faster attack at its minimum setting.

P.S. I am emulating all three of them and more in my Extreme Dynamics compressor.

[zero-13]

Thank you guys for your respones and detailed information!

Learning about compression and various types of compressors has sort of been like noob cryptonite lol.

The basic function of compression has recentl, finally sunk in more clearly as far as the technical understanding goes, but learning to hear more critically / accuratley / observationally beyond perceived loudness is taking a while.

[Loque]

For all I know the new RE MBC is just the MClass in new clothes?

The MBC is indeed almost equivalent to MClass Compressor in "Soft-Knee" mode, and at the same time to the Mix Channel compressor at some settings. Though, MClass can't do more than ratio "4" in terms of the MBC (Mix Channel one can). And the MBC can do faster attack at its minimum setting.

P.S. I am emulating all three of them and more in my Extreme Dynamics compressor.

For my ears i find the SSL compressors way more musical and smooth, with a better overall sound.

Have you any objective comparisons for your opinion? I am just curios.

[orthodox]

Have you any objective comparisons for your opinion? I am just curios.

Sure I have. I made series of tests and measurements and completely reverse engineered them (compressors). I learned writing DSP stuff by analyzing existing things first.

[selig]

I've never "bonded" with the MClass compressor except as an excellent envelope follower (Audio to CV) in Reason (better than any other options to date except maybe the Sweeper).
I cut my audio teeth on an SSL back in 1984, so I've "bonded" with that mix bus compressor from my engineering birth, so to speak!
My first experience with the SSL compressor back when Record came out was 100% positive, as far as setting it like I'd set the hardware version (or any of the hardware clones such as the Alan Smart version) gave me the same feeling as the original. Not saying it's an exact 1:1 clone, or that it's even important that it is - just that it fulfilled the same function for me as the original, and certainly as good as any of the other software options out there.
I'm currently still un-boxing my house and studio after moving out of NYC last week, but as soon as I'm up and running again I'll try to do a side by side comparison - who knows, maybe I'll surprise myself and find them pretty much the same (and would be happy to admit they are pretty much the same if so).

[chimp_spanner]

I really like the Master Bus Compressor. It's very much on a level with the Glue Compressor from Live to my ear. And most of the time I use it on individual channels; it's great for drums with its variable attack time. I kinda wish the Channel Dynamics plugin had an attack control too. I know it's got a fixed time, plus peak and fast modes. But not quite enough control for me.

[Loque]

I've never "bonded" with the MClass compressor except as an excellent envelope follower (Audio to CV) in Reason (better than any other options to date except maybe the Sweeper).
I cut my audio teeth on an SSL back in 1984, so I've "bonded" with that mix bus compressor from my engineering birth, so to speak!
My first experience with the SSL compressor back when Record came out was 100% positive, as far as setting it like I'd set the hardware version (or any of the hardware clones such as the Alan Smart version) gave me the same feeling as the original. Not saying it's an exact 1:1 clone, or that it's even important that it is - just that it fulfilled the same function for me as the original, and certainly as good as any of the other software options out there.
I'm currently still un-boxing my house and studio after moving out of NYC last week, but as soon as I'm up and running again I'll try to do a side by side comparison - who knows, maybe I'll surprise myself and find them pretty much the same (and would be happy to admit they are pretty much the same if so).

I am looking forward to your test andi am curious if i can trust my ears.

[dvdrtldg]

I've never "bonded" with the MClass compressor except as an excellent envelope follower (Audio to CV) in Reason (better than any other options to date except maybe the Sweeper).

Huh. Well, there you go. Will check this out, thanks

[orthodox]

I've never "bonded" with the MClass compressor except as an excellent envelope follower (Audio to CV) in Reason (better than any other options to date except maybe the Sweeper).

BTW, I do an envelope follower too, both Audio2CV and Audio2Audio, mono/stereo, two way relative to the threshold.

[selig]

I am looking forward to your test andi am curious if i can trust my ears.

OK, got a bare bones setup going in the new place, and was able to do some comparisons.
One difference is in the attack release "curves" (shapes), with the SSL been less exponential on the attack and release with the same settings. At first glance, running tones through the two compressors looks very similar, but this is due to the subtleties
As orthodox said, the general shape of the ratio is similar if you use soft knee on the MClass, but there is a difference here too.
As is often the case, the panel settings do not produce identical results when matched to the same value. With both set the same, attack 1ms, release 100 ms, threshold -18 dBFS, ratio 10:1, input and output @ 0 dB, the final output on the MClass was -13 dBFS and the SSL Master was -16 dBFS (with an input signal sweeping up to 0dBFS). This would indicate the ratios are not precise on one machine or the other, with my money on the SSL due to the math.
The math:
If the ratio is 10:1 and the threshold is -18 dBFS, when the signal is at 0 dBFS you should have a total gain reduction of 9/10 of the difference (between threshold and signal level at input. So when the signal is 0 dBFS it is 18 dB above the threshold, and 1/10 of 18 = 1.8, meaning 1.8 dB increase for every 18 dB above the threshold. 1.8 dB is very close to 2 dB, and indeed the SSL gives me -16 dBFS output when the input goes 18 dB over the threshold.
It would appear the MClass ratio is lower than it's reading because it's not giving as much gain reduction. Indeed I have to set the MClass ratio to around 24:1 to get 16 dB of gain reduction with an input level of 0dBFS. Also, using Soft Knee increase the output (decreases the gain reduction) by another 2 dB even with ratio set to inf!

Lesson 1: never trust the front panel setting when matching two devices.
Lesson 2: even when time/rate settings DO match, the curves/shapes may not, which WILL sound different.
Lesson 3: dynamics are a subtle art, a small change in the response of one compressor vs another can leave a noticeable sonic impression.

Conclusions: as heard, there is a difference between these compressors, as there are with many compressors (even two software compressors that are emulating the same hardware design).

[selig]

I've never "bonded" with the MClass compressor except as an excellent envelope follower (Audio to CV) in Reason (better than any other options to date except maybe the Sweeper).

BTW, I do an envelope follower too, both Audio2CV and Audio2Audio, mono/stereo, two way relative to the threshold.

Do tell?

[Loque]

I am looking forward to your test andi am curious if i can trust my ears.

OK, got a bare bones setup going in the new place, and was able to do some comparisons.
One difference is in the attack release "curves" (shapes), with the SSL been less exponential on the attack and release with the same settings. At first glance, running tones through the two compressors looks very similar, but this is due to the subtleties
As orthodox said, the general shape of the ratio is similar if you use soft knee on the MClass, but there is a difference here too.
As is often the case, the panel settings do not produce identical results when matched to the same value. With both set the same, attack 1ms, release 100 ms, threshold -18 dBFS, ratio 10:1, input and output @ 0 dB, the final output on the MClass was -13 dBFS and the SSL Master was -16 dBFS (with an input signal sweeping up to 0dBFS). This would indicate the ratios are not precise on one machine or the other, with my money on the SSL due to the math.
The math:
If the ratio is 10:1 and the threshold is -18 dBFS, when the signal is at 0 dBFS you should have a total gain reduction of 9/10 of the difference (between threshold and signal level at input. So when the signal is 0 dBFS it is 18 dB above the threshold, and 1/10 of 18 = 1.8, meaning 1.8 dB increase for every 18 dB above the threshold. 1.8 dB is very close to 2 dB, and indeed the SSL gives me -16 dBFS output when the input goes 18 dB over the threshold.
It would appear the MClass ratio is lower than it's reading because it's not giving as much gain reduction. Indeed I have to set the MClass ratio to around 24:1 to get 16 dB of gain reduction with an input level of 0dBFS. Also, using Soft Knee increase the output (decreases the gain reduction) by another 2 dB even with ratio set to inf!

Lesson 1: never trust the front panel setting when matching two devices.
Lesson 2: even when time/rate settings DO match, the curves/shapes may not, which WILL sound different.
Lesson 3: dynamics are a subtle art, a small change in the response of one compressor vs another can leave a noticeable sonic impression.

Conclusions: as heard, there is a difference between these compressors, as there are with many compressors (even two software compressors that are emulating the same hardware design).

Thanks for your test and feedback. I am always curious, what the devs add as magic under the hood, like adding EQ to the input signal feeding the compressor threshold, its like using modified sidechain signal. Or other magic, with soe kind of MB compression, which is not visible...in such cases the frequency should respond differently on the matched settings.

This would also mean, you can work with just a few versitile compressors and do 90% of the jobs. Still got to learn something

Thanks, Giles

[selig]

Thanks for your test and feedback. I am always curious, what the devs add as magic under the hood, like adding EQ to the input signal feeding the compressor threshold, its like using modified sidechain signal. Or other magic, with soe kind of MB compression, which is not visible...in such cases the frequency should respond differently on the matched settings.

This would also mean, you can work with just a few versitile compressors and do 90% of the jobs. Still got to learn something

Thanks, Giles

Yes - maybe…
I'm still impressed with how a different compressor can do different things even with the same settings. I was lucky to work in Nashville in my "formative" years, where there was a great deal of vintage and new gear at studios and rental companies to "explore". And much of it was in top shape, from amazing Fairchild, Neve, Altec, Helios, EMI, Decca, Pye, URIE, and Trident compressors to the classics like the LA-2a, 1176, and DBX 160, to "newer" compressors from Tube Tech, Chandler, Valley People, Focusrite, GML, Manley, and Emperical Labs (and many other esoteric compressors I'm probably forgetting about).

It all depends on what you are expecting/wanting, and what your source is giving you. In some cases a compressor that didn't work on everything else will work like magic on a particular source/voice/instrument. Or maybe you work on a different genre and find yourself needing a more (or far less) subtle compressor than what has worked for you in the past.

Certainly, there are cases where a few different compressors will give nearly identical results on a particular source, so it's not like there are obvious differences in all cases. But there are also cases where only one compressor will give you a certain sound! The main thing IMO is to define what types of things you need a compressor to do. For me I want smooth leveling (LA-2a), aggressive coloring (1176), and versatility (Distressor), and creative applications (Dyna-Mite). And of course I'm always looking for something that does things no other compressor can do (Selig Leveler!).

[littlejam]

hello,

brooklyn is not the same without selig
everybody is walking around with their head down
nobody is smiling
it's a real bummer

congrats on the move to your new studio, sir
well played

cheers,

j

(also just wanted to tag this thread for future reference about compression)

[zero-13]

Wow, unreal amount of hard work and technical detail, thank you!

Almost all of it is over my head lol, but my take away is that with the MClass Compressor and the new MB Compressor I should be covered as far as compressors, just some slight sonic "flavor" differences between them, please correct me if I'm misunderstanding?

[selig]

Wow, unreal amount of hard work and technical detail, thank you!

Almost all of it is over my head lol, but my take away is that with the MClass Compressor and the new MB Compressor I should be covered as far as compressors, just some slight sonic "flavor" differences between them, please correct me if I'm misunderstanding?

You are the only one who will know if you're covered or not. But I would at the least familiarize myself with the most common classic compressors out there - they are classic for a good reason!
I would include the LA-2a, 1176, and DBX in that league, and also any variable Mu types like Fairchild or Manley (in addition to the SSL master bus compressor, also a "classic" IMO).

The MClass doesn't really fill any particular void for me, either as an obvious "always try it on this instrument" type compressor OR as a versatile desert island compressor such as a Distressor. There's nothing I could say that I feel is bad about it, it's just that for every application where I use compression, it's not the best tool for the job. After you get familiar with the classics you may still (or may not) find the MClass useful - but you'll only know once you compare! At the least, I suggest getting familiar with an LA-2a (optical) type compressor for transparent "leveling/smoothing", and an 1176 type for aggressive "obvious" compression, and the SSL for bus/mix compression (since you can't go wrong with any of those).

[orthodox]

BTW, I do an envelope follower too, both Audio2CV and Audio2Audio, mono/stereo, two way relative to the threshold.

Do tell?

I mean my device (https://www.reasonstudios.com/shop/rack ... -dynamics/) does. It can output raw gain modulation signal (like the one driving a VCA in VCA model) or using external one as input.

[selig]

Do tell?

I mean my device (https://www.reasonstudios.com/shop/rack ... -dynamics/) does. It can output raw gain modulation signal (like the one driving a VCA in VCA model) or using external one as input.

Doh! My bad, was searching for a dedicated device!
Which is maybe not a bad idea???

[Boombastix]

But I would at the least familiarize myself with the most common classic compressors out there - they are classic for a good reason!
I would include the LA-2a, 1176, and DBX in that league, and also any variable Mu types like Fairchild or Manley (in addition to the SSL master bus compressor, also a "classic" IMO).

I agree, I have the IK Multimedia version of all three (LA/FET/V-Mu) and they are super. LA-2a on vocals is often used because it works, if I need more I put their Manley Vari-Mu after. Their Vari-Mu is really nice as it has Mid/Side (they all have) and in M/S mode you can bring out the sides have have them "dance" around a little more.

M/S is a feature often overlooked when discussing compressors. Suggest you have one with M/S, or at least try one.

[zero-13]

Wow, unreal amount of hard work and technical detail, thank you!

Almost all of it is over my head lol, but my take away is that with the MClass Compressor and the new MB Compressor I should be covered as far as compressors, just some slight sonic "flavor" differences between them, please correct me if I'm misunderstanding?

You are the only one who will know if you're covered or not. But I would at the least familiarize myself with the most common classic compressors out there - they are classic for a good reason!
I would include the LA-2a, 1176, and DBX in that league, and also any variable Mu types like Fairchild or Manley (in addition to the SSL master bus compressor, also a "classic" IMO).

The MClass doesn't really fill any particular void for me, either as an obvious "always try it on this instrument" type compressor OR as a versatile desert island compressor such as a Distressor. There's nothing I could say that I feel is bad about it, it's just that for every application where I use compression, it's not the best tool for the job. After you get familiar with the classics you may still (or may not) find the MClass useful - but you'll only know once you compare! At the least, I suggest getting familiar with an LA-2a (optical) type compressor for transparent "leveling/smoothing", and an 1176 type for aggressive "obvious" compression, and the SSL for bus/mix compression (since you can't go wrong with any of those).

I appreciate your knowledge even though it's at full apogee from my current perigee.

So looking through all my REs, it seems I have had for a while but not utilized or even tried out for that matter the Cakewalk RE-2A (Don't remember at all if I bought this or it was given free at some point?) and also have Selig Leveler (is this similar to LA-2a?), I figure I'll spend some ear training time with these along with the new MBC and try to get some minor discernment going.

Not to derail this discussion, sort of relevant, but I've just recently started using a Selig Gain as the first insert effect and bringing down any and all devices down to -18db before going into anything else i.e. compressor, EQ etc. and then another Selig Gain at the end to get everything to around -12 db... Is this in any way shape or form okay or "correct"?

My understanding from countless varied tutorials and piecemeal snippets of info, is that some effects / devices work optimally at around -18 db so my thinking is just feed anything and everything the same -18 db?

Any quick guidance (by quick I mean even a "nope, try again" is more than enough ) you guys can share in this regard would help immensely to either quickly confirm I'm understanding things, or get myself back to the lessons before passing go.

[orthodox]

... bringing down any and all devices down to -18db before going into anything else i.e. compressor, EQ etc.

I don't think it is necessary. Not many devices behave differently depending on input level. In particular, 0 dB does not mean anything, there is no clipping nor overflow.
You can safely assume most of devices are not input level dependent. Devices doing saturation, some hardware emulations can behave differently. Master Bus compressor, most EQs are not dependent.

[zero-13]

Cool, Thank you!

So sounds like I'm doing an extra time consuming step that is rarely if ever needed?

Is there any sort of blanket scenario or general kind of rule where I should be thinking that maybe make sure incoming signal --> effect needs to be toned down, or can pretty much just stop worrying about this -18 db thing?

[zero-13]

I'm slow, I think you pretty much answered this.