EQ compress, or compress EQ? digital vs analogue

[tk2k]

I'm in a debate with a friend of mine, and we're looking for some input.

We started talking about compressors on individual channels, as well as EQs, and then moved on to system wide compressors and system wide EQs.

Long story short...
I said a compressor should be the first thing applied to an audio signal, because on an analogue system the insert port comes before the lo-cut, built in eq, etc. This is supported by the block diagrams of my console (mackie 24*8)

He says no, an EQ should come before a compressor, it's just not feasible to do that on ever channel in the analogue world. The block diagram of his PM5D supports this, as the EQ is applied before the compressor. He argues since the console is digital, they can do the processing correctly.

I extended my logic to the mains, as did he. He would say the compressor should be the last thing the signal hits before the amps, while I would say the compressor should be the first thing the signal hits out of the board.

I'm guessing we're both partially wrong. Anyone wanna set this straight?

[pmolsonmus]

First and foremost, there are no hard and fast rules. But, generally a bit of roll-off EQ may be enough to get the signal clear enough to cut through without compression on an individual channel. So, that would be my first step. If I couldn't manage with that then I'd look to some compression, but because you are changing both amplitude (the compressor) and phase (with the EQ) it is a complicated issue.

I would advise to use as little signal processing as necessary. There is no hard and fast setting (or preset) that works for everything. The mic, vocalist, key of the piece, orchestration, room, audience, are all factors in the eventual audio path. Get it to sound good.

A little help regarding compression may be of some interest. (warning - although the explanation is excellent - the use of profanity is common) Compression is extremely complicated signal processing and many who use it are operating in the dark. Here is a great link from the website I mod on recording.org

http://recording.org/pro-recording-f...mpression.html

Good luck

Phil

[tk2k]

Thanks for the link, it's some good stuff. It's nice to see something subjective about compression other than straight up definitions.

I'm pretty well versed in compression as well as EQ, it's come down mostly to which kind of modulation has the least impact on the quality of sound, the idea being to use that device first. Mostly we were surprised to find it done differently in the digital world than the analogue, and figured Yamaha might know a secrete there.

First and foremost, there are no hard and fast rules. But, generally a bit of roll-off EQ may be enough to get the signal clear enough to cut through without compression on an individual channel. So, that would be my first step. If I couldn't manage with that then I'd look to some compression, but because you are changing both amplitude (the compressor) and phase (with the EQ) it is a complicated issue.

I would advise to use as little signal processing as necessary. There is no hard and fast setting (or preset) that works for everything. The mic, vocalist, key of the piece, orchestration, room, audience, are all factors in the eventual audio path. Get it to sound good.

A little help regarding compression may be of some interest. (warning - although the explanation is excellent - the use of profanity is common) Compression is extremely complicated signal processing and many who use it are operating in the dark. Here is a great link from the website I mod on recording.org

http://recording.org/pro-recording-f...mpression.html

Good luck

Phil

[FMEng]

Scenario one: a compressor on a mic channel. There, I prefer to apply EQ before the compressor because I don't want audio that's going to get filtered out affecting the compressor action. For example, the mic may be picking up a bunch of extraneous bass which I will clean up with the high pass filter. I wouldn't want that bass modulating the compressor. That said, it isn't a sin to have the EQ applied after the compressor.

Scenario two: A compressor, or more accurately a limiter, in the amp chain for driver protection. The limiter must be just ahead of the amp in order to be most effective. If the EQ or crossover were last before the amp, you could adjust more gain to some or all of the audio spectrum which would over power the speakers, and the compressor could not react to it.

[BNBSound]

This may help you to get your head around some of what's going on in the signal chain. Years ago when I would go out to a gig in a tough room I'd get there early and tweak the graphs on my monitor sends to within an inch of their lives. Figuring that creating a feedback free stage would be great for the band, all I would get was complaints about not being able to hear anything. Turns out I had chopped so much signal out of the wedges that the amps driving them had tons of headroom left. So I used the output gains on the graphs to make up the difference and then we had the best of both worlds.

So the guy who said to EQ first had a pretty good idea. If in the act of sonically trimming your signal you happen to remove some of the level, you'll need less compression to tame it down (probably). Although if your desk doesn't have the routing to accomplish that you're out of luck. Something I used to do on festival stages with a 4 bus console was to skip dynamics on the channels and just insert them on the groups. This keeps compression out of the monitor mixes and is surprisingly useful. I'd usually do Kick/Snare on one, bass on one, guitars on one and vox on the last.

[Footer]

Most pro level analog consoles have a switch on each input to decided where the insert chain goes for this exact reason. The way Mackie does it should always be taken with a grain of salt. Mackie's solos also pile on... something most people hate.

[museav]

Scenario one: a compressor on a mic channel. There, I prefer to apply EQ before the compressor because I don't want audio that's going to get filtered out affecting the compressor action. For example, the mic may be picking up a bunch of extraneous bass which I will clean up with the high pass filter. I wouldn't want that bass modulating the compressor. That said, it isn't a sin to have the EQ applied after the compressor.

This is exactly why there is no right or wrong answer as the inverse to the situation noted is that any boost applied via EQ may trigger compression earlier. Whether that is desired or not depends on the result you are trying to obtain. You can get even more complex when you start getting into sidechain applications where you use one signal to trigger the processing or effect for a different signal.

Scenario two: A compressor, or more accurately a limiter, in the amp chain for driver protection. The limiter must be just ahead of the amp in order to be most effective. If the EQ or crossover were last before the amp, you could adjust more gain to some or all of the audio spectrum which would over power the speakers, and the compressor could not react to it.

Agreed, a system limiter must be the last thing in the chain before the amplifier in order to perform its function properly.

[museav]

This may help you to get your head around some of what's going on in the signal chain. Years ago when I would go out to a gig in a tough room I'd get there early and tweak the graphs on my monitor sends to within an inch of their lives. Figuring that creating a feedback free stage would be great for the band, all I would get was complaints about not being able to hear anything. Turns out I had chopped so much signal out of the wedges that the amps driving them had tons of headroom left. So I used the output gains on the graphs to make up the difference and then we had the best of both worlds.

If you had to cut bands for feedback then wouldn't raising the output gain put you right back into feedback?

[Versha]

If you had to cut bands for feedback then wouldn't raising the output gain put you right back into feedback?

not necessarily. it just requires some care to avoid producing feedback when boosting the output gain. when using lower end and a lot of mobile systems this is really the only option you have left since often times there is no signal processing.

[TimmyP1955]

Most mics have a lot of low mid boost - as much as 12dB centered around 200Hz (proximity effect). Not only does this make for a very unnatural sound, it's most of what the compressor's detector works off of, which is why the loud, high, screechy notes sail right through. If the compressor is after the EQ - where the mic's proximity effect is compensated for - the compressor does a better job. Best of all is to have sidechain EQ on the compressor - one band to make it less sensitive to the proximity effect, the other to make it more sensitive to the loud screechy stuff. Best of all is dynamic EQ or multi-band compression.