Over the years I've read how lots of people prefer to stack compressors say on a vocal (or whatever) to help reduce how much one compressor has to work. Instead of having one compressor with 4:1 ratio, they might use 2 with a 2:1 ratio. I understand how this helps in the analog world, but is it relevant in the digital world? A few youtubers I keep up with do it in their pro tools sessions and I'm not sure I can hear the difference in the digital domain after doing some of my own tests. What's your opinions on this matter? Does it actually offer anything relevant in the digital domain where compressors are lightening fast and don't have a lot of character to them (unless it's an emulation compressor)?
in my opinion it does matter! the point is that you dont hear the compressors pumping as much...
its especially audible if you use some extreme settings, with extreme gain reduction...
if you set one compressor to, lets say, 10 db gain reduction its pretty sure you will hear it pump...but if you use two compressors with an overall gain reduction of 10 db ...you most probably wont hear as much pumping!
in my opinion it does matter! the point is that you dont hear the compressors pumping as much...
its especially audible if you use some extreme settings, with extreme gain reduction...
if you set one compressor to, lets say, 10 db gain reduction its pretty sure you will hear it pump...but if you use two compressors with an overall gain reduction of 10 db ...you most probably wont hear as much pumping!
cheers
Maybe I just didn't set my compressors extreme enough during my tests. Maybe they were too subtle for me to hear an audible difference... Thanks
I don't know about stacking digital compressors but it makes sense with analog compressors but I think its cuss of the unique sound that analog compressors deliver. I mean most people use compressors for tone more then they use it to compress. Maybe with digital compressors you use less CPU? I need to experiment.
You would do it for the same reasons in digital as you would in analog. Typically you would have one that performs fast limiting to control peaks, and another one set much slower for shaping the envelope.
Sent from my DROID4 using Tapatalk 2
__________________ - It looks just like a Telefunken U47 - with leather. You'll love it ... - Jazz is not dead - it just smells funny.
- It doesn't make much difference how the paint is put on as long as something has been said. Technique is just a means of arriving at a statement.
use an opto type comp that's slower and smoothes out overall average volume and a limiter that only reacts to loud peaks. If the vocal track is pretty good to begin with you won't hear as much compression as you would with a track that's all over the place volume/dynamic wise.
That is part of the reason I like to record with analog comp and use another comp in mixing.
It's not stacking but you get the same kind of effect.
I do the same with EQ.
The secret to making this work is to use different compressors/limiters with different qualities, and/or using different attack/release settings that work together to give the desired dynamic results.
It can all be very perplexing until that moment when you get the "a-ha moment".
Yep. It's not really about the ratio. It's about the time characteristics of the individual comps, as well as their tonal fingerprints. Adding up a bunch of different attack and release curves can give results that would be impossible to attain with just one compressor, and the same thing works tonally. A lot of plugin comps are designed to have their own specific tone as well in order to be more like analog ones (just put them on a track and listen with no compression happening, then compare it to bypass), and to that end it makes just as much sense to stack them as it does analog compressors.
Analog= more noise but great sound I often find analog comps just sound better with more gain reduction than plugs so less need for this.
Digital= often find a lot of reduction sounds bad so using to comps can help here. I often use a cla76 feeding a cla3a works great.
Just my op....
Really good responses guys. Thanks a ton.. I'm not so sure I'm going to use this technique since a lot of my songs have 10-20 vocal tracks and CPU could be an issue after doubling all comp instances, but I'll give it a try on a simple mix sometime. Come to think about it, I guess I just did this in a sense. Had 3 vocals all singing the chorus that was extremely dynamic... So each vocal had a compressor, then I sent all those vocals to an aux and threw another compressor on that because even with the individual comps it wasnt tamed enough. So in a certain sense I just did this. Interesting... It was perfect too lol.
Just finished a project with 6 primary vocalists and I used this technique. My signal chain on each vocal track included:
Waves C1 Gate
Waves RenDeEsser
UAD LA-2a
UAD LA-3a
Waves Hybrid EQ
Waves Vocal rider
It was very easy to smooth out the individual qualities of each vocalist, and to make the vocals move to the front of the mix. Each comp was set up to compress a maximum of 2-3 DB.
Sent from my Transformer TF101 using Gearslutz App
Vocal rider would help relieve the stress from the compressors. On the other hand, some compressors sound magic when you're really chugging some signal through them hard.
This might have been said, but using 2 compressors is also a lot about using one to attack on initial transients, and another for some longer reduction. Totally relevant in both the analog and digital world, if you can't hear the difference the problem is the ears.
Sent from my PC36100 using Gearslutz App
__________________ The mix is ALLLLLLLLLLMOSSSTTTT 'perfect'.
This might have been said, but using 2 compressors is also a lot about using one to attack on initial transients, and another for some longer reduction. Totally relevant in both the analog and digital world, if you can't hear the difference the problem is the ears.
Sent from my PC36100 using Gearslutz App
I can hear the difference in this situation, I was talking about stacking compressors with the same settings (except ratio) to reduce the "stress" of one compressor.
Just finished a project with 6 primary vocalists and I used this technique. My signal chain on each vocal track included:
Waves C1 Gate
Waves RenDeEsser
UAD LA-2a
UAD LA-3a
Waves Hybrid EQ
Waves Vocal rider
Total amateur commenting here but interested in the debate:
Vocal rider last? -2a and -3a? Those are interesting choices, I normally use VR first, and in trying to minimise compression would usually aim for only one "colour" compressor as these are?
Incidentally can anyone confirm that VR is anything other than a compressor that can write automation? I like the VR (perhaps a lazyness thing, I only ever do minimal manuals after I use it).
use an opto type comp that's slower and smoothes out overall average volume and a limiter that only reacts to loud peaks. If the vocal track is pretty good to begin with you won't hear as much compression as you would with a track that's all over the place volume/dynamic wise.
or use the Molot or any 33609type/clone for this 2 applications ITB
33609 Unit OTB
Total amateur commenting here but interested in the debate:
Vocal rider last? -2a and -3a? Those are interesting choices, I normally use VR first, and in trying to minimise compression would usually aim for only one "colour" compressor as these are?
Incidentally can anyone confirm that VR is anything other than a compressor that can write automation? I like the VR (perhaps a lazyness thing, I only ever do minimal manuals after I use it).
Wow! How did I rate such a criticism? Amateur? Maybe I am.....or maybe I'm not. I was simply trying to add to the conversation on the topic with a signal chain that worked for me on a specific project. My clients are pleased with the way the record sounds, and so am I, which is the yardstick of a successful endeavor. If such sniping is necessary to make you feel significant, well.........I feel really sorry for you.
Wow! How did I rate such a criticism? Amateur? Maybe I am.....or maybe I'm not. I was simply trying to add to the conversation on the topic with a signal chain that worked for me on a specific project. My clients are pleased with the way the record sounds, and so am I, which is the yardstick of a successful endeavor. If such sniping is necessary to make you feel significant, well.........I feel really sorry for you.
Worship-he was referring to himself as an amateur, not you!
A common thing to do is to use an 1176 first to handle the peal, because it has a really fast attack, and then use an LA2A (or another OPTO) to smooth out the rest. You normally don't want the extreme peaks to trigger the compression on the OPTO compressor because it recovers too slow. Here is an example using plugins: Vocal Chain Example
There is much to be said for compressing each track individually according to the characteristics of the instrument, voice, etc. (listen to any early 1990s Dwight Yoakum record...the guitars are squashed while the vocals run amok) Then use something to smooth it out a tad just before mastering...when it will be compressed again.
Keep in mind that if it is played on the radio it will run through an Optimod (or similar) which will compress it AGAIN.
Stevie Wonder used to have a small transmitter...he test-mixed and listened on the radio in a car. He then tailored the master to fit that format...because that is where most of the final consumers will actually listen.
Ask yourself "how will this sound on crummy car speakers?" Let that be your guide.
I can hear the difference in this situation, I was talking about stacking compressors with the same settings (except ratio) to reduce the "stress" of one compressor.
That's an interesting question. If you use two instances of the same plugin compressor the behaviour of both to an input signal is exactly the same. The ratios multiply if you stack them (two 2:1 and you get 4:1; one 2:1 and one 3:1 you get 6:1 etc.).
The envelope detector's behaviour of the second compressor is influenced by the output of the first compressor. The relation can be expressed mathematically depending on how the detector works. I guess often it is based on hilbert transform.
So, the resulting signal envelope of two stacked compressors of the same type and with the same same attack-/release-settings should be different from the resulting envelope using just one compressor?
...So, the resulting signal envelope of two stacked compressors of the same type and with the same same attack-/release-settings should be different from the resulting envelope using just one compressor?...
yes!
edit:
in this pic:
white = source
green = one compressor, 4:1, 4db GR, 10ms/100ms
red = two compressors, 2:1, 2db GR each (4db GR total), 10ms/100ms
in this example, 2x compressors give a "smoother" curve, but at the expense of slightly higher distortion...(may depend on circuit type/design...?)
edit:
in this pic:
white = source
green = one compressor, 4:1, 4db GR, 10ms/100ms
red = two compressors, 2:1, 2db GR each (4db GR total), 10ms/100ms
in this example, 2x compressors give a "smoother" curve, but at the expense of slightly higher distortion...(may depend on circuit type/design...?)
Thanks for the awesome graph, there's an obvious difference from the graph, but is the audible difference worth the effort of setting up two compressors every time?
New Reply 
Facebook
Twitter
LinkedIn
