Serial compression - do you use it and how

[Popey]

I have recently started using compression in serial and have been finding this a lot better in most cases than one compressor doing all the work.

Obviously there are different types of comp (1176, opto etc) and wondered if other people work this way and have found any chains that work well on particular instruments.

Something I am going to start investigating myself but as this is quite common I wondered if other rt users were willing to share their experience.

[selig]

I've been using serial compression forever, most have (but don't think about it that way). For example, if you compress a vocal when you record, then compress again during the mix (common practice), you're doing serial compression. Or if you use the Master Compressor on the mix AND a brick wall limiter on the master, you're doing serial compression.

There are two approaches I'm aware of. One is to use heavy compression first (fast time constants, high ratios) followed by lighter compression second. I tend to prefer the other way around, which I always use on mixes – a slower/lighter compressor first followed by a more aggressive compressor to catch any remaining peaks. Both ways "work", so spend a little time learning the difference if you can.

A classic vocal chain can involve a 1176 and LA2a in either order. YEARS ago I did a project where we had two LA-2a compressors, and hearing that Lou Reed did a similar thing we used one in "limiter" mode and one not - can't remember for sure (it was almost 40 years ago) but I THINK the limiter was the second compressor. This was pretty lucky for us to have TWO LA-2a compressors available for this trick, felt 'extravagant' at the time. Now it's common to run multiples of everything everywhere because "you can". I count myself lucky to have learned how to make "less" work (out of necessity) because I know you often don't need multiples of anything! But when you do, it sure is handy to just grab another and see what happens!

[crimsonwarlock]

Same as selig, using compressors in serial, mainly during recording. My vocal chain is actually three compressors in serial. Same as selig I go for a slow comp into a faster comp. Currently that is MooTube comp into RE-160. However, before my vocals hit the audio interface, it goes through a DBX 286s, which is a hardware vocal processor including compression (as well as an expander). This makes it three. Selig"s leveler is also in that chain, so you could argue that makes it four




Now Giles will come back and say leveler.is not a compressor

[selig]

Now Giles will come back and say leveler.is not a compressor

it's a compressor, just not a downwards compressor!

[crimsonwarlock]

Now Giles will come back and say leveler.is not a compressor

it's a compressor, just not a downwards compressor!

Nevertheless, it's an awesome thing. Bought it specifically for my vocal chain, together with your de-esser, even though the DBX286s also has a de-esser. I like using multiples of stuff, doing very little each.

[Loque]

I have no specific "general" setup. I try to adjust the levels as needed, like
* Too high peaks --> Fast attack, and release, quite high ratio, low threshold for a compressor. Alternative i use a clipper or limiter. Depends...
* Sometimes out of control --> I try to level it or use a compressor with slower values and lower ratio and threshold, often followed by brick wall to catch some nasty peaks
* "glueing" is done by lower attack, release, ratio and quite high threshold. Just a little touch from the compressor
* If i want all the background noises or even loudness, i have fast attack and release, low threshold and quite high ratios around 1:10-1:20
* Someimes i just use a limiter or limiting compressor with high threshold to catch or lower some peaks
* ...

Probably the list can go on endless. Try to get an idea what you want to fix/change, than try to setup the devices. There is no rule not to use 10 compressors, different clippers and 2 limiters in serial...well, maybe except if the result sounds crap

[moalla]

Since today i started as well to experiment with a Molot Ge with opto preset, to a Kotelnikov GE with bus preset. This TDR compressors in my master chain working fantastic, gives me this hardware vibe in a way I didn´t excepted. Crimson´s Mojoo-Re 160 vocal chain is also a nice thing, what I´m using but for the analogish vibe this tokyo dawn labs molot ge is a real weapon of creamness. Today you can get molot ge for 10Dollar at audiodeluxe for sale
I think molot is the best coloring comp you can get as plugin, or on par with mjuc!

[crimsonwarlock]

I think molot is the best coloring comp you can get as plugin, or on par with mjuc!

Too bad Red Rock pulled all their paid REs from the shop. Just after those were gone I realized I wanted their C1-L1 comp (I think it was named that), which is basically the Molot compressor in RE format.

[PhillipOrdonez]

I don't. I use a single compressor in brick wall mode at the mastering stage. None of that crazy, modern, serial compression nonsense you kids are into these days with your loudness wars and sh!t

[selig]

it's a compressor, just not a downwards compressor!

Nevertheless, it's an awesome thing. Bought it specifically for my vocal chain, together with your de-esser, even though the DBX286s also has a de-esser. I like using multiples of stuff, doing very little each.

Sometimes I need to do "serial de-essing" on the occasional vocal, even on only one or two songs in a project if the vocalist changed mics or sang differently etc.
There are three basic de-essing topologies and I find combining them helps with tricky vocals.

My favorite type is the DBX902 approach, which inspired the Selig DeEsser. This approach uses no threshold and can reduce the entire signal quick enough in response to sibilance there's no need for a crossover in the audio path. You can also put it anywhere in the chain since it is more or less impervious to levels.

The second type is dynamic, either a dynamic EQ or a multi-band approach and probably the most common of the three. This approach lowers only the range of the actual sibilance so there are crossovers involved (which most of the time is no issue for me).

The third is side chain, probably the oldest approach. In this set you either boost the sibilance on the side chain signal or filter everything else out. The latter works well on the SSL/Reason mixer using the "Filter to DYN S/C" function and running the high pass all the way up to 4 kHz. As always with de-essing, fast attack/release and high ratios work best. This approach can even have success taming sibilance on a mix during mastering, if you know what you're doing!

[crimsonwarlock]

Sometimes I need to do "serial de-essing" on the occasional vocal, even on only one or two songs in a project if the vocalist changed mics or sang differently etc.
There are three basic de-essing topologies and I find combining them helps with tricky vocals.

My favorite type is the DBX902 approach, which inspired the Selig DeEsser. This approach uses no threshold and can reduce the entire signal quick enough in response to sibilance there's no need for a crossover in the audio path. You can also put it anywhere in the chain since it is more or less impervious to levels.

The second type is dynamic, either a dynamic EQ or a multi-band approach and probably the most common of the three. This approach lowers only the range of the actual sibilance so there are crossovers involved (which most of the time is no issue for me).

The third is side chain, probably the oldest approach. In this set you either boost the sibilance on the side chain signal or filter everything else out. The latter works well on the SSL/Reason mixer using the "Filter to DYN S/C" function and running the high pass all the way up to 4 kHz. As always with de-essing, fast attack/release and high ratios work best. This approach can even have success taming sibilance on a mix during mastering, if you know what you're doing!

And as always you drop some gold nuggets of knowledge. Thanks again for that

[Popey]

I've been using serial compression forever, most have (but don't think about it that way). For example, if you compress a vocal when you record, then compress again during the mix (common practice), you're doing serial compression. Or if you use the Master Compressor on the mix AND a brick wall limiter on the master, you're doing serial compression.

There are two approaches I'm aware of. One is to use heavy compression first (fast time constants, high ratios) followed by lighter compression second. I tend to prefer the other way around, which I always use on mixes – a slower/lighter compressor first followed by a more aggressive compressor to catch any remaining peaks. Both ways "work", so spend a little time learning the difference if you can.

A classic vocal chain can involve a 1176 and LA2a in either order. YEARS ago I did a project where we had two LA-2a compressors, and hearing that Lou Reed did a similar thing we used one in "limiter" mode and one not - can't remember for sure (it was almost 40 years ago) but I THINK the limiter was the second compressor. This was pretty lucky for us to have TWO LA-2a compressors available for this trick, felt 'extravagant' at the time. Now it's common to run multiples of everything everywhere because "you can". I count myself lucky to have learned how to make "less" work (out of necessity) because I know you often don't need multiples of anything! But when you do, it sure is handy to just grab another and see what happens!

I had not thought of it that way and have used bus compression and individual track compression so guess i have been doing this a lot longer than i thought. I will certainly try using aggressive/lighter compression and vice versa to see what the different effects are. Thank you for the practical tips.

I have been messing about with it today and am also finding that different plugins bring different tones so think i will be getting to know what i like in serial for different flavours too (i think this is because some emulations saturate a bit different but not really sure).

[Popey]

Same as selig, using compressors in serial, mainly during recording. My vocal chain is actually three compressors in serial. Same as selig I go for a slow comp into a faster comp. Currently that is MooTube comp into RE-160. However, before my vocals hit the audio interface, it goes through a DBX 286s, which is a hardware vocal processor including compression (as well as an expander). This makes it three. Selig"s leveler is also in that chain, so you could argue that makes it four




Now Giles will come back and say leveler.is not a compressor

Thanks for the advice. i do not have the exact same compressor but i will be trying the many i do have in this way.

[Popey]

I have no specific "general" setup. I try to adjust the levels as needed, like
* Too high peaks --> Fast attack, and release, quite high ratio, low threshold for a compressor. Alternative i use a clipper or limiter. Depends...
* Sometimes out of control --> I try to level it or use a compressor with slower values and lower ratio and threshold, often followed by brick wall to catch some nasty peaks
* "glueing" is done by lower attack, release, ratio and quite high threshold. Just a little touch from the compressor
* If i want all the background noises or even loudness, i have fast attack and release, low threshold and quite high ratios around 1:10-1:20
* Someimes i just use a limiter or limiting compressor with high threshold to catch or lower some peaks
* ...

Probably the list can go on endless. Try to get an idea what you want to fix/change, than try to setup the devices. There is no rule not to use 10 compressors, different clippers and 2 limiters in serial...well, maybe except if the result sounds crap

Great tips loque thanks. I must admit i have used compression before but used to always lean towards clipping or limiting as i found them easier to not get wrong. I have been using compressors alongside a scope which has helped me understand their behaviour more.

I guess i never tried serial compression on the same track before as i am not overly confident with it and serial could just make it twice as bad

[Popey]

I don't. I use a single compressor in brick wall mode at the mastering stage. None of that crazy, modern, serial compression nonsense you kids are into these days with your loudness wars and sh!t

Hey i take exception to that......

just the kids part I am far too old to be called that

[BRIGGS]

I use the comp-01 to shape envelopes. It's like a bonus ADSR.

[PhillipOrdonez]

I don't. I use a single compressor in brick wall mode at the mastering stage. None of that crazy, modern, serial compression nonsense you kids are into these days with your loudness wars and sh!t

Hey i take exception to that......

just the kids part I am far too old to be called that

I'm obviously not being serious about compression imagine how my music would sound if I only used one compressor in every song

[Popey]

Hey i take exception to that......

just the kids part I am far too old to be called that

I'm obviously not being serious about compression imagine how my music would sound if I only used one compressor in every song

Ha ha you underestimate how naive i can be sometimes

[crimsonwarlock]

Same as selig, using compressors in serial, mainly during recording. My vocal chain is actually three compressors in serial. Same as selig I go for a slow comp into a faster comp. Currently that is MooTube comp into RE-160. However, before my vocals hit the audio interface, it goes through a DBX 286s, which is a hardware vocal processor including compression (as well as an expander). This makes it three. Selig"s leveler is also in that chain, so you could argue that makes it four




Now Giles will come back and say leveler.is not a compressor

Thanks for the advice. i do not have the exact same compressor but i will be trying the many i do have in this way.

To explain it some more in detail: The serial compression trick for vocals is aimed at getting very controlled vocals with the compression being as transparent as possible. Besides putting a slow and a fast comp in serial, the trick is also to have them act very minimal, with just a few dB of compression each. In the analog days it was called "kissing the needle", as the needle of the compression meter was barely moving

You can use slow first or fast first, it just gives a little bit of difference. I like slow first for my vocals, but fast first might work better for specific vocals. Just try which works best for you. You can also use a limiter instead of the fast compressor, as a limiter is actually a very fast compressor Warren Huart likes to put a limiter first before the compressor for aggressive rock vocals. I saw him doing that during a mixing tutorial and it sounded great. Fair to say that he wasn't going for "transparent"

[selig]

Thanks for the advice. i do not have the exact same compressor but i will be trying the many i do have in this way.

To explain it some more in detail: The serial compression trick for vocals is aimed at getting very controlled vocals with the compression being as transparent as possible. Besides putting a slow and a fast comp in serial, the trick is also to have them act very minimal, with just a few dB of compression each. In the analog days it was called "kissing the needle", as the needle of the compression meter was barely moving

You can use slow first or fast first, it just gives a little bit of difference. I like slow first for my vocals, but fast first might work better for specific vocals. Just try which works best for you. You can also use a limiter instead of the fast compressor, as a limiter is actually a very fast compressor Warren Huart likes to put a limiter first before the compressor for aggressive rock vocals. I saw him doing that during a mixing tutorial and it sounded great. Fair to say that he wasn't going for "transparent"

All of these types of discussion should lead with clarity on what you are “going for”, because that can literally change EVERYTHING!!!
This has been a public service announcement, we now return you to our regular programming.

[crimsonwarlock]

All of these types of discussion should lead with clarity on what you are “going for”, because that can literally change EVERYTHING!!!

Hmmmm... I'm pretty sure I wrote this:

The serial compression trick for vocals is aimed at getting very controlled vocals with the compression being as transparent as possible.

Describes pretty much what we're going for here

[PhillipOrdonez]

I'm obviously not being serious about compression imagine how my music would sound if I only used one compressor in every song

Ha ha you underestimate how naive i can be sometimes

Haha Popey!