trimph1
I actually think BOTH ways are useful...for me. I like being able to muck around with compressors and other effects but it certainly does not hurt being able to understand what the beep is going on either...
I feel the same as you, trimph. One of the reasons I've sort of stayed out of this thread is due to "compression" being so subjective, there are just too many variables. I completely respect Jeff's hard work as well as everything he has offered here. Starting points and the like are all great ideas.
However, and this is where MY personal experience/opinion comes in.....every sound and transient will be different based on the instrument, the player, how it's being recorded as well as the particular compressor that is being used.
This is what I believe Dean is getting at...which if it is, I tend to favor that method. It's like a few months ago, we had a pretty killer eq/frequency discussion. You can't tell a person how to eq something without hearing it. We can say "high pass at all times" or "high pass on these particular instruments" but we can't give them a frequency number, can we? It's impossible without hearing the sound. If the sound doesn't have a lot of low end in it, it would be senseless to tell someone "well, I always high pass at 150 Hz."
If a guitar tone doesn't have meat in the 600-800 range to start with, we don't just cut it because "it worked on one project I did". The same with compression. I see it as almost impossible to give a starting point with anything other than a rough ratio setting for particular instruments....and even that is flawed thinking until you physically have the sound at your disposal to work with.
How can we give an attack setting when we haven't heard the instrument? How much release is enough? Remove -3 dB of gain or -8 dB of gain? We can't make that call. Then you factor in that EVERY compressor is going to behave/react differently, and whatever "starting point plan" you have, becomes completely useless. This is not meant to say what Jeff ( or anyone else ) has given is "useless". I'm just saying there are way too many things to take into account with this compression animal to where trial and error or "by twisting knobs" is most likely going to be what happens in *MOST* cases.
In other cases you may decide to use the formula that Bobby Owinski has provided for making drum compression work in time with the music. That's a prime example of when a formula or starting point is going to get you what you are looking for with slight trial and error. Or we know that a longer attack and shorter release with the right ratio and threshold is going to simulate what a transient plug would do on a snare drum to increase crack and resonance. Those types of settings are going to work with very little tweaking every time for just about every snare other than the subjectivity of how much crack and ring do you want on the snare.
For example, I can run the PC 4k buss comp on any snare drum using a threshold of -37.7, ratio of 4:1, attack 9.4 ms and release 0.1 and it would simulate what a transient plug would do on a snare drum to where it brings out the crack and the amount of ring. It will work every time for me and to adjust for another instrument, I tweak the threshold, the attack and release for that instrument. But those numbers alre always going to be good starting points for me to make a snare crack a bit more and sustain a bit more.
But for leveling, tightening, conditioning etc, there's no way a starting point (in my opinion) is going to help. Especially if the compressor you use vs. the one the person writing or teaching, behaves differently...and unless you are using the same one, trust me...they will behave much differently. I have some comps that I have to really jump on to hear a difference. If I listened to the stuff I read and applied it to comps like that, I'd not get the results I am looking for.
It's good to read up on the controls....it's a blessing to have people like Jeff share their knowledge for the price of an internet connection....but at the end of the day, it is in my honest and humble opinion that each compressor as well as the instrument it processes, must be treated case specific because of how many different variables there are.
It's like anything else really....each individual is going to experiment and create something that works for them. This comes from trial and error as well as how the comp performs with the instruments that are fed into it. Another example slightly off the beaten path but not entirely...
I've been told for years that a truss rod in a guitar should be set a certain way. 1 1/4 turns or something or whatever it is. I have NEVER taken notice to how many turns my truss rod gets. I turn it until the guitar I am using plays the way I want it. If it's 2 1/2 turns...so be it. When you tighten it, the neck straightens...when you loosen it, the neck smiles. How much tension you need depends on the guitar, how you play and how you want it to sound. A little string buzz can be good and add a little more of a percussive attack to a guitar. This is good as long as it's not fretting out and making a "dead" sound. Some guys (like me) like that, others like less heat on their truss so the guitar breathes a little more.
A compressor is a lot like that to me. It's something you need to feel as well as know what to listen for on YOUR material you are working with. It's also SUPER important to have examples of what a compressor does both good and bad on various instruments. I do a lesson plan with my students covering this exact thing using several compressors. So everything is really important and like trimph, though I will agree on the "both method" my heart will always belong to knob twisting, trial and error.
-Danny