trimph1
And how does one achieve this so that it works in all set ups?
Or rather, should it work in all set ups?
I'm a conflixxed puppy here....
Most certainly you can. First off, it starts with how and what you record. The print is important as well as how dynamic the instrument is to begin with. If you have a loud distorted guitar, well, it was made to be that way. However, with the right set-up...say Eddie Van Halen, rolling off the volume knob allows that particular sound to change dynamically. Ed's volume knob is literally a volume knob with a bit of gain wrapped up into one.
Other guys that use insane gain lose the volume knob part of their sound. The knob becomes strictly a gain knob and controls the gain in the sound more-so than the actual volume. You'll notice with any extreme, high gain sound, that when you do turn down the volume knob on your guitar, that you actually don't lose much volume at all. It's really close to being all full out yet the gain diminishes.
Now if we take piano as our test subject, and the player is tickling the ivories like a pro, you're going to definitely get some quiet passages and some that are hammered. If the piece was tracked with a band, 9 out of 10 times the band is going to adjust their playing and dymanics to this piano because they know the part if coming up and that's what pro bands do. If it's recorded at a later time after the band has recorded and they don't take these dynamic changes into account, you're going to have to automate.
Automation with a bit of compression fixes all this stuff really. Sometimes the automation needs to be done to the other instruments though instead of the instrument you may THINK needs to come up a bit. For example, let's take this piano test in the other direction I mentioned.
A band tracks bass, guitar and drums. Their piano player can't make it that night. He comes later and lays his part down. It's very dynamic...the band forgot about this and sort of played all out. Do you raise the piano in the spots where it drops? Or do you drop the band? I drop the band so that they accommodate this piano IF indeed it sounds like the right thing to do. Other times, bringing up the piano is the right touch on those super soft passages.
Now...the key to loud is to have the mix right first. The biggest problem with mixes today is people keep relying too much on their compressors and limiters to do all the work. This is why you get the sound of @ss, over compression, excessive limiting, choked off instruments, pumping, loss of snare drum crack, even-ness withing the music...all this stems from relying too much in comps and limiters.
I shared some of my methods months ago on how I manually level my mixes. This is crucial to maintaining your dynamics as well as being able to go loud without artifacts. By the time I kick my limiter in...it doesn't matter if I use my PSP, the Waves or the X1 Concrete Limiter, nothing is ever going to suffer dynamically. People always think because something is loud that it loses dynamics instantly as soon as they hear it come on and scare them.
This is not always the case. Something may look and sound like it's going to be loud, but if you analyze a wave form that is done right, you'll notice that it's NOT always full bore at -0.3dB or wherever you set it. There's a difference between loud and done correctly and loud and squashed to heck.
Most mixes should average out to be from about -15 to -12 RMS. That said, something loud may peak at -7 RMS. That doesn't make that mix bad or too crushed. You see, when things are limited right AFTER you've leveled the mix correctly, the over-all volume just changes...not the dynamics. When I drop to a clean part in my mixes...the volume drops no matter how much limiting I'm using. But, you have to understand that the extra volume is an "over-all" thing...it's not losing as much dynamics as you think.
If we did a mix that peaked at -10 RMS and averaged out to -18 RMS, and then redid the same mix to where we peaked at -5 RMS and averaged -13 RMS...the dynamics are exactly the same...the over-all mix is just louder. How you automate and level you instruments and then your mix decides what you can get away with when you get to the limiting stage. If you skip these important steps and just jump to the limiter, you're going to over-use it and it's going to over-compensate and give you artifacts.
The reason being, it reacts to the peaks in your material. When it grabs a peak, it's going to try and make everything else stay along the lines of that peak. If you don't have those types of peaks BEFORE you limit, this allows for less limiting and a smoother end result. Hope this clears some of it up for you trimph. :)
-Danny