Mastering Extremely Dynamic Tracks

[GregV]

Hi Everyone! Long time no see... Sorry, the new gig doesn't allow me to "play" online the way I used to.

Anyway, I'm about to start mastering an EP and 3 of the tunes start and end at a "whisper quiet" level and peak heavily in the middle. (Think of the bands Mogwai or Explosions In The Sky)

I'm just wondering how much of that dynamic range needs to be sacrificed by necessity.

Any opinions?
Thanks in advance

[Bookerv12]

What will be the target playback?
Earbuds?
Cars?
Home stereos?
All of them?

Typically now, dynamic ranges are pretty squashed to allow for the above scenarios.
Most of the stuff you listen to that you like the dynamics of, is more squashed than you would think.
Limiting the range is actually a good thing depending on the final playback environment.
A lot of people now would freak out when presented with the dynamic range of "Dark Side Of The Moon".

I use the term "squash" benevolently as I can't thing of a more discriptive word.

[AlexLakis]

Automation is your friend.

[GregV]

Thanks Guys! I kind of figured I'd have to do some squishing.
One listen on a crappy HP notebook and it was over! Ha Ha

[Bob Olhsson]

It's a matter of gain riding and not squishing.

[JFreak]

maybe they want that old trick of 1st track low levels for awhile when the customer turns up the volume and then suddenly the track becomes VERY LOUD

[simon.a.billington]

There are a few things you can try and maybe a little bit of everything might be what you need to deliver your mix.

Gain Automation: This is the most transparent of processing, providing you're not boosting too much noise in the process. Ideally you'd want to do this before your plugin chain, but thats okay because ProTools plugins are post fader. Alternatively if you want to leave your fader free for general mix adjustments you can try automating Clip Gain on the track itself.

While this is great for relatively detailed adjustments, you would totally burn up all your time if you wanted to get into the minutiae of things.

3rd Party Gain Leveling Plugin: Vocal Rider, Bass Rider, Peak Rider, Drum Leveler, these are only the plugins I know of. They all work in their own way, and are specialised in certain tasks, although they can often be set to work in other roles.

What they do is pretty much automate the gain, but they can do it far more quickly and many times more accurately than what we ever could without burning too much time on it. Products like Vocal Rider also works well on things like guitar solos and can write out the automation so you can make any adjustments you feel is necessary. So don't feel its just designed for vocals.

Low Ratio Compression: As opposed to hitting something harder with a compressor for overall tonal shaping, with more dynamic sounds it can be more desirable to set a ratio of less than 2:1, often less than 1.5:1 and a low threshold in order to tame the peaks. You'd want an attack quick enough to catch those transients and but the release can mostly be set to taste, the longer the release the more levelling is done on the signal. Usually I'm very light handed with my compression, but in this instance you can even get around 8-10dB or reduction without much side-effect, this is because the ratio is set very low.

Standard Compression: Not much point in explaining this one, it's fairly common and straight forward. Compress any sound as you normally would, usually leaving the main transients untouched and with a ratio set between 2-4:1, usually.

Parallel Compression: Also a very common technique. Either use a compressor with a Mix or Wet/Dry knob, or duplicate a track and apply a compressor to that. Slam the track really hard with the compressor, really crush it, pull it out of the mix entirely, then slowly add it back in until you feel there is enough body in the track.

Soft Clipping: As opposed to hard clipping this can tailor your transients much more gently. It's commonly associated with the effects of tape, but its not exclusive to it. There are plugins that do this, IK has one, Alpha Compressor has it as an option, you could use a tape emulation like a J37, for example, or you could use a compressor, but you really need one with a soft knee and a short enough attack time. Just set the threshold low enough to catch most the peaks, set the attack to fast and release also relatively fast. The knee needs to be set to soft and the ratio set between 1-2:1, depending on taste. The higher you go the more it will develop into hard clipping, although, this may also be a desirable sound, depending what you are after.

Hope there is enough here to help. As I said earlier, you may bellowing at using more than one of these techniques, if not all, to get the balanced sound you need.

[Bob Olhsson]

Ideally everything would be done with gain automation because you can choose where the gain will be increased and do tricks like reducing it slightly before climaxes to retain impact.

[simon.a.billington]

It's true, but can simply take forever to do properly, particularly of you have a high enough track count.

If time is money, it's not always the best solution. That's why I'd usually do some quick overall levelling with it and focus on some troublesome areas. Then just load it a preset, quickly adjust a few settings and move on. Revisit it later for final tweaks, that's always done best with fresh, more objective ears.

You got to think that for decades we got away without much in the way of level automation, at least not to the degree we talk about these days. Then it was also a matter of relying on tape to do a bit of levelling, cleaver application of compressors etc...

It's a different ballpark now we have DAWs and automation, but how much is too much before it begins sounding synthetic or over produced??

I think each person will have their own answer to that question, so it's meant to be rhetorical.

[GregV]

You guys are awesome. Thank you so much for the input. It will all be put to good use!!