Well.... where to start?-
And yeah, no way to keep it short and sweet. Sorry!
Part of my delay getting back to this is because after reading the attached IKM forum thread, I wanted to do some experimenting with their stuff specifically. Basically I wanted to see what the poster(s) there were getting on about before I got into this reply.
What I found was that some of those compressors, as described, don't do much if you feed them the suggested nominal input levels.
In one example, at a ratio of 4:1 or less, I was struggling to get more than 1 to 1.5 dB of gain reduction out of it at most. At somewhat higher ratios, I was getting in the 3dB GR range with the compression cranked up.
Pretty much just as described.
Normal? Correct? Maybe.
Honestly, seems a little
off to me, but not by as much as the people in that thread might have thought.
This is a broad and complex subject IMO, and perhaps I am already getting ahead of myself.
My typical disclaimer: I make no claim to be an authority on anything. I have been studying and researching this, have learned some interesting things and formed some opinions. Many of those opinions may change as I learn more. The only thing I can say with certainty is that for every answer I think I have found, it has created at least ten new questions.......
So the general topic is "gain staging"
It would be easier if I could simply get right to the point and answer the core question- "does it matter?" Or "how much does it matter?" Another maybe?
Some of this is technical. Some philosophy. Some shear opinion.
I have a narrow little bit of experience before digital. I had some hands on experience with one
hardware compressor. We considered ourselves very lucky to have that. We used what we had. Now there are hundreds of free or inexpensive plug ins to choose from. And the way I see them used in tutorials often bears little resemblence to how we used it back then.
Does how we did things then have any relevence to how to work in digital? Another maybe.
Heres one thing to consider. Many a decent mix has been made largely ignoring even the rudimentary rules of "gain staging"
And if your method works for you, and gets good results, is it wrong?
We have all heard, and hopefully practiced not clipping the A/D converter when recording. Lower levels are good. But what level is ideal, and how critical is it?
Same thing goes for mix down. Avoid clipping the D/A converter. Watch your peaks.
But what happens in between? Hmm......
All this reminds me of a conversation I had with an experienced producer/mixing engineer recently. The topic was basically "why is mixing well so difficult?"
We agreed that it is in part because there are so many different ways to do it, no real roadmap because there is not really any one "right" or necessarily "wrong" way.
But also because the difference between a "good" mix and a "great" mix often consists of many little subtle things that even an experienced ear may not readily discern. The other side of this coin is that a collection of things you may not be able to really hear well can also add up to harm your results.
Where does your signal level fall into this? Is it possible chaingng the working levels can add up to some degree of improvement? Maybe?
I am told that with many plug ins, it technically really doesnt matter. Time based processes for example. Typically wouldnt be effected much by level. On the other hand, many emulation type plug ins are designed to operate at levels similar to what was done in hardware. Which brings up the 0vu = -18dB RMS etc discussion.
Some people take this to extremes and attempt to be so very precise..... Thats not really how it worked in analog exactly. I think this is yet another digital thing. We have the ability to meter and analyze to rediculous tolerances. And look, somebody somewhere is going to experiment far out of those tolerances and make something cool as a result. Some plug ins "clip" internally and it makes them sound very bad. Others might do something interesting.
Another thought. We know we don't need to record too loud. I have been told by people far smarter about this, that even high-end expensive A/D converters have a range or even "sweet spot" where they function more accurately, and that level is not only well below clipping, but actually may corrospond to... 0vu
And heres even more fuel for thought. We see people talk about "mixing into compression".... well how do we set up that compressor unless we anticipate a range of the level thats going to be hitting the input.....
Then theres templates. If you mix a lot, templates are an awesome time saver. But if you set up an advanced template with your go-to preffered routing and processes already set up, what are the signal levels expected to be?
So while we are thinking about all that, heres another thing to consider.
Prior to digital, I had never heard the term "gain staging" put quite like that. And some of the early explanations and methods I heard made me think it must be largely a digital thing. Until I finally heard an explanation that began to sound familiar and make more sense.
We didnt call it "gain staging" but heres a partial explanation of how we would "calibrate the studio"
First of all, any tube gear was "warmed up" as tubes change a bit after they have been on for a while. 45 minutes minimum was typical for me. Others seemed to think longer was good.
In the next example, I would mute my monitors. (Monitors were always muted except when you wanted to hear them. Turning things on and off, signal spikes, accidents, all could do expensive damage.)
With the sound off I would route a calibrated sine wave signal at 0vu into my compressor. I would set the compressor at -1, or perhaps -1.5 vu as desired. This was how the level of compression was set. And it was left there for the duration of that particular portion of whatever we were working on.
I think the more current trend of higher levels of GR is another digital era thing. I never saw that before. 3dB GR was considered "extreme" to us.
And FWIW, I think there is a case to be made for not using a great deal in any one step now, but thats another subject.
So, the lingering question. Does "gain staging" improve the sound? I think it often may. Does it make a big difference in sound? Not in my opinion. Is it the great answer to better mixes? No.
For me, its a bigger workflow thing than anything.
Some more things to point out. More opinion I guess.
If the desired level going into your inserts is a nominal 0vu or -18dB RMS..... What range should we typically be recording in?
I have loaded a fairly large number of other peoples stems or projects for mixing, or for the purpose of evaluating and advising them on how they might improve their methods. I'd conservatively estimate at over 200.
As of yet I have yet to see one single example where the levels didnt need to be trimmed down IMO for better results.
But maybe its just me, and I admit- still and always experimenting and trying to learn more!