The Master Comp can be used for parallel compression by adjusting the mix knob. In his book, Bob Katz discusses parallel compression in mastering and recommends that the output of the parallel compressor be between -5dB and -15dB, relative to the non-compressed signal.

Instead of measuring the output of the parellel compressor relative to the non-compressed signal, the Master Comp takes a different approach. The amount contibuted by the parellel compressor is a percentage of the mix between the processed and unprocessed signals- and is controlled by the mix knob.

What percentage range on the mix knob corresponds to -5 to -15dB relative to the unprocesseed signal (assume the unprocessed signal to be 0dB)?

hi,
it is a bit complicated to get this clear with the mix-knob, but here is my aproach.
set the mix-knob to 50%, then adjust the overall-leveldrop you get by this with the output set to +6 db. then you can adjust the level of the compressed signal with the makeup-knob, which affects only the compressed part of the signal. this is the where you can dial in your -5 to -12 db (does not go lower in mastercomp).

something to think of:
in mastercomp you only have a threshold of -30 db, so if you want to go with bob katz recommendation of -50 db, you have to bring the level up 20 db before mastercomp (so the compressor will go deeper in action) and bring it down again by the same amount after mastercomp. with this "trick" you get the desired -50 db threshold.

when you have a daw with good latency-compensation, you can do it on an aux-bus very easy. threshold-thing is the same, mix knob set to 100% and adjust with the auxlevelfader.

have fun,
mathias


edit: some vst-helpers for the scenario above: blue cat gain suite, acon digital effect chainer

Hello,

Bob has a classic console view of parallel compression which is good for some reasons.
During our development we decided to use the Mix control for its easy of use and direct blending between pre and post processed sound as well as for some other reasons.

The solution by Mathias is clever and worth trying, if you are looking for something less precise please try to work with Mix values between 30 and 70% depending on the compression depth, makeup setting and desired sound, you will easily find the right setting by ear.

Best,

Pushing the level 20dB ahead of the MasterComp in order to achieve a -50dB threshold seems to be a bad idea to me, as it results in a greatly reduced bit-resolution! When mixing this can in certain cases be okay (but far from ideal!) but in a mastering situation this for me is a no-go.

So how about making the MasterComp capable for a -50dB threshold?! I would appreciate this lots!

Another thing bothers me when doing parallel-compression with the MasterComp: why is the VU-meter reduction indicator changing, when turning the mix-knob? It simply doesn't make sense to me, as the compressed signal should always be compressed by the same amount, no matter how much of the direct-signal is mixed back in. And as parallel-compression is meant to be a very subtle thing, it would be great, if the VU-meter could tell me constantly, what's happening to the compressed signal.

-50dBFS threshold is not such a bad idea, especially in the case of the MasterComp where a very low ratio can be used and the sound impact is further reduced by a Mix knob.
The threshold deeper then shown on the Threshold knob is possible with extra plug-in with input of 20dB gain and -20dB output attenuation in the MasterComp's output.

Regards,

i don't know which daw you use.
but you may consider, that modern daws do calculate in 32bit-float internally. that means, that you have a few 100 db of gainrange, which you can use, even in mastering. there is no reduced bit-resolution in these calculations, it is just shifting the whole audiorange up and down, which does not change anything in the audio. the recording is a digital "representation" of your analog dynamic range of let's say 96 db. inside the 32 bit float "range", it can be moved up and down a big amount, without degradation.

however, you have to watchout for a few things:
first is (should be self-explanatory) you have to bring the level down, at the output of the daw, where it goes to the "real" world of da-converters, which do have a fixed gainrange.

second, you have to watch for plugins, that do emulate real-hardware with their saturation and effects, that do occur, because of different levels. with such plugins, you will get nasty distortion, they are meant to be used with "normal" audio levels.

third you have to clear it up, if your daw is really doing all calculations on tracks, busses, master in 32 bit, which is sometimes not the case.

if this is all checked and clear, you can use the technic described without any damage to the audio, even in mastering.

to the vu-meter behaviour:
it shows the dry/wet combined amount of reduction, which is what i want to see, because that is what happens to the audio at the output of the plugin. maybe there could be an option, to switch the metering point, so it will suit your needs better.
if you do parallel-compression with an aux-bus, you will have this metering.

greetings,
mathias

you were faster than me mateusz! :-)

mathias

Well, I almost totally agree on the 32-bit floating point point.

I never said a -50dB threshold is a bad idea - the opposite is true! I was wishing this could be implemented directly into the MasterComp. The thing is, that I learned to be very distrusting concerning most plugins. And I find myself spending too much time making null-tests and so on, just to find out, I discovered a new problem to have and solve. So I would still prefer a -50 dB Threshold in a plugin rather than needing two plugins to achieve the effect of it.

For me, I still don't understand the VU-behaviour: the compressed signal is usually totally squashed using a very light ratio. The dry signal mixed back in is unprocessed. So I can't see, how there can be a dry/wet combined reduction, as I rather have the dry signal enriched by the wet.

Sorry, maybe it's just my lack of understanding, but that's how I understand it.

when the dry/wet knob is set to 100%, you here the effect of the compressor to its full extent and see it on the gain-reduction meter.

when the mix knob is turned counter-clockwise, the direct signal is mixed in proportionally, while the compressed signal is reduced! proportionally.
in the 50/50 position (12 o'clock) this means compressed signal -6 db and direct signal -6 db, mixed together.
if both signals were unprocessed (dry), this would mean the same signal level as before. but one signal is compressed (reduced in level), so the combined outcome is lower, but not as low, as the compressed signal alone. this is shown by the gainreduction-meter.

as you turn it towards 100% dry, the gain-reduction reduces (compressed signal is mixed out proportionally). this is what you hear.
you can see it as to level-knobs (for compressed and dry signal) combined inversely and keeping balance througout.
be sure to have the auto-makeup-gain turned off!!

got it?

mathias

that's true. but when you know your plugins and do not change to new ones, you can trust your experience and findings, and use it in your work.
i work with psp-plugs since years now and i know them very well.
they are really something i can rely on and use them all the time.