Parallel compression in mastering

A friend of mine owns & runs a mastering room. He asked me what was the best way to make a parallel compression device. The idea is to use a pair of 1176 or 160 and blend that sound with the original signal. I do that a lot when I mix on my console and then it´s very easy of course.

He wants the unit to have the shortest signal path possible, no coloration (of the original signal) at all and he wants to have a fader on the unit that controls the blend.

Some kind of summing circuit & gain make up amp would be necessary I guess. How would you construct such a unit? Extremely transparent sound on the original signal is required and it does not have be cheap to build.

Thanks in advance,

Gunnar

I think what you are describing is not really parallel compression but a variable bypass or "blend" control.
My first thought would be to modify the compressor itself; adding a blend control and, if possible, summing it into an in-phase summing node at the output driver circuit. This is the shortest, cleanest path. All passive, No additional power supplies ect.

I think he really describes parallel compression - there are many confusing names (like side-chain compression :? ) for this technique.

If you cannot mod the compressor as described by Crusty2 or want the flexibility to use different compressors, I suggest a simple passive (perhaps balanced) mixing network and a (two for the balanced case) JH990 with DC servo for makeup. Check the mixer META if you are not familiar with mixing networks.

Samuel

Thanks guys!

A splitting transformer will work as well, however you are going to have to take some coloration on the original signal as a result. Unless the guy you know really really really really knows exactly what he is trying to do, I would experiment with a bunch of stuff and see wha works- I use parallel compression on a daily basis and couldnt imagine doing that in mastering, before you kill yourself trying to design a totally transparent system, try something simple just to see if the process is at all useful. Id check to see if lundahl makes an amorphous core transformer with dual secondaries that could be used as a splitter, those are fairly if not suprisingly invisible sounding transformers.

dave

High frequency phasing could become an issue here. This can be measured with a dual trace scope. Some small delay might have to be introduced to get the timing right.

I never really liked this "trick". I believe it is credited to being started in the early NYC studios. You're getting a tiny bit of phase lag which dulls/smears transients (but people say what they're doing is adding more "punch" with this technique--I disagree) and a louder summed signal (which may be perceivably more "punchy"). I would rather use a compressor with a slow attack and a fast release to get more "punch".

But if he really wants to do this, why get more complicated than you need to... Why can't he just use a stereo aux send, to send the signal to his compressor of choice and blend the result of the compressed and original signal at the console/daw? Or must he have a "blend" knob?

[quote author="Ethan"]You're getting a tiny bit of phase lag which dulls/smears transients (but people say what they're doing is adding more "punch" with this technique--I disagree) and a louder summed signal (which may be perceivably more "punchy"). I would rather use a compressor with a slow attack and a fast release to get more "punch".[/quote]

from a users perspective, I would argue the relevancy of factoring any hi freq phasing that smears transients in comparision to a limiter run in series who's very purpose is to ultimately smear transients. I honestly dont think Ive done a mix in eight or nine years without parallel limiting, in fact, its fairly rare that I'll use a limiter in series in a mix at all, sometimes on a vocal, all my parallel busses are generally limiter returns. I dont hook up a scope to my mixes, I just listen to them, so while Im not arguing that there may be the possibility of phasing for a text book discussion, in use, if you are worried about transients the entire question of why a limiter is in use in the first place, series or parallel, would come to my mind.

I think prespective is probably very relevant to a discussion about parallel limiting as there are really a ton of different reasons and goals in mind when a limiter gets patched. with parallel limiting you can use ALOT of limiting for efffect but tuck it behind an uneffected track for depth and nothing else. Most of the people that I know are using limiters as distortion generating devices in the first place and you cant really take that approach to limiting unless you can do it in parallel.

anyhow, dont mean to derail this into technique but thought Id throw in my two cents. I suppose based on my comments you guys can probably see why I think thats a wierd thing to screw with in mastering, I would buss my entire mix in parallel and then master the sum of that. I would still check out the amorphous core lundahls, they are kinda like not having a transformer "sound" at all but still sort of have the "feel" of having a transformer there, might work ok so far as transparency goes.

dave

[quote author="Ethan"]
But if he really wants to do this, why get more complicated than you need to... Why can't he just use a stereo aux send, to send the signal to his compressor of choice and blend the result of the compressed and original signal at the console/daw? Or must he have a "blend" knob? [/quote]

That is how I do it in my studio actually but he does not have a console with aux sends in his mastering room.

Thanks again for your input everybody.

Gunnar

Was thinking about it a bit more and there is a simple solution: take a linear pot, connect the compressor out and the original signal to both "ends" of the pot. From the wiper you get the signal to the make-up gain stage.

Of course, you may want to replace the pot with switched resistors.

Samuel

just to interject randomly but using a 1176 for mastering work......yeeuck.

i dont know anyone who would let this touch their 2 buss for this type of job. these comps can sound really nice in tracking situations but definately a no no for mastering.

Well, I have actually tried this (with a pair of 160 VU) in another mastering room and if used gently, it can add some "meat" to a thin mix.

The mastering guy won´t use this on mixes that sound great from the start of course, only when something extraordinary is needed.

The 1176 won´t touch the original 2 buss, it will only add a little of its character "from underneath".

Blend control was a bad description (by me) of what he wants. The original sound should be at the same level all the time and the fader is supposed to raise the volume of the compressed signal.

does this place only work on rock or something.
160vu on a 2buss, it gets worse :grin:

in all seriousness, these things can be the best on drum tracks, snares come off lush through them if dialled in correctly.
mid range crunch and a dark sound though are ever present characteristics. fine if you want chilli pepper guitars but less so if you want high quality mastering work.
is this place a studio that does mastering alongside to supplement income or do they just use these on very few occasions?

if they want some nice warmth underneath, tell them to swap out their 1176 for an LA2a they they would really know warmth.
still cant think how one underneath would add this to a sound unless there was some strange harmonic thing happening still i wasnt there so cant add more.

Blend control was a bad description (by me) of what he wants. The original sound should be at the same level all the time and the fader is supposed to raise the volume of the compressed signal

Click to expand...

Still can be called a blend control. The simple balance control Samuel suggested is an easy starting point.

Live sound guys use this trick a lot as it is very easy to do on many live consoles.
Something else to try would be a H.P. filter in front of it to restore only top end signal.