Orchestral music for TV use - dynamic range?

[reid]

I'm wondering just how much dynamic range to allow in the music I'm currently writing for a TV series, and would appreciate any advice available.

The score is purely orchestral, and so has the potential for daft swings of dynamic range (nice quiet bits at -24dBfs quickly followed by big stabs at -0.5dBfs - that kind of thing). I'm guessing that if I submit files to the dub with that kind of range to them, and the mix engineer doesn't have time to level ride them all through the mix, then they'll get stuck through a set-and-forget compressor/limiter, and never be heard from again.

So, is there an optimum dynamic range I should be trying to squeeze everything into? Say, nothing quieter than -12dBfs, nothing louder than -3dBfs; that kind of thing? Actually, 9dB of range is ridiculous, but you see what I'm driving at.... Should I be trying to pre-empt any dynamic processing that might occur at the dub, and actually write in gain changes to the stereo master before sending it off to the dub?

A few composer colleagues happily smash their orchestral stereo masters through an L16 and the like, but frankly it sounds like poo once it comes out the other end, so I'd like to steer clear of going down that path - but I still want an end result that can find its place in amongst the dialogue, FX and atmos.

Thanks!

[londontown]

The massey limiter is your friend (if you're Pro Tools based) or other gentle transparent limiting on your 2 bus that'll preempt a bit of the dub carnage and won't mess with the sonics of your mix.

It'll probably get screwed with a bit either way though, orchestral stuff being so full in terms of frequency range; best that you control the dynamics in a friendly way so it doesn't get screwed with so distastefully... (they might just set and forget it anyhow so there's a modicum of risk that it might get screwed with even more if you 'pre master' it for want of a better term);

but of course the usual concept applies to write around the dialogue and sound and compliment as opposed to reinforcing, both in terms of frequency and in terms of location...as much as possible so that they can find a level that works without ducking your music and so it sort of mixes itself without too much dynamic skullduggery...I've always thought TV dubs can be the most damn cruel to music. Oboe or mid rangey orchestrations on female dialogue, trombone or low mid textures over a male dialogue line...these things can conflict with such ease (obviously a colossal generalisation but the point remains), and the music will lose that battle at the dub every time...not to mention there's never enough time and the dialogue is obviously the most important thing at the end of the day, as much as we might like to think otherwise, if you can't hear what they're saying, the whole damn thing is usually a bust.

some folks do smash the shite out of it to way too much of a degree, and that's equally bad as usually, it then gets squashed for broadcast again, and you end up with NO dynamic range, none whatsoever...just a rectangular waveform when zoomed out...but judicious use I've found of a really good limiter plugin such that you don't hear it but it's reducing it (the Massey is just bloody extraordinary at this, esp. on orchestral stuff, bizarrely) has made a bunch of TV dubs and film temp dubs so much easier it's ridiculous.

Originally Posted by reid

I'm wondering just how much dynamic range to allow in the music I'm currently writing for a TV series, and would appreciate any advice available.

The score is purely orchestral, and so has the potential for daft swings of dynamic range (nice quiet bits at -24dBfs quickly followed by big stabs at -0.5dBfs - that kind of thing). I'm guessing that if I submit files to the dub with that kind of range to them, and the mix engineer doesn't have time to level ride them all through the mix, then they'll get stuck through a set-and-forget compressor/limiter, and never be heard from again.

So, is there an optimum dynamic range I should be trying to squeeze everything into? Say, nothing quieter than -12dBfs, nothing louder than -3dBfs; that kind of thing? Actually, 9dB of range is ridiculous, but you see what I'm driving at.... Should I be trying to pre-empt any dynamic processing that might occur at the dub, and actually write in gain changes to the stereo master before sending it off to the dub?

A few composer colleagues happily smash their orchestral stereo masters through an L16 and the like, but frankly it sounds like poo once it comes out the other end, so I'd like to steer clear of going down that path - but I still want an end result that can find its place in amongst the dialogue, FX and atmos.

Thanks!

[danijel]

i second londontown on solo woodwind etc. they have to really go down against dialogue. percussive sounds are very hard to mix too - if they are over a line, they will either cover it, or (if mixed low) they will be mistaken for a diegetic sound.

for the show i'm currently working on, i, unfortunately, get music mixdown (no separates), so i can't duck just the offender, but also the strings/pads - it often ends up sounding like solo clarinet as the rest of it gets inaudible.

if you want to pre-mix, i suggest you slam a limiter (-10dB) over temp dialogue you got, and for every scene roughly adjust the dialogue volume to about -20dB RMS (short term, measured only when they speak). then ride music levels against that while monitoring on an actual TV set or small cheapo speakers with your window open. you'll be amazed at how much you will have to turn up the quiet parts.

and again, +1 for what londontown said - it's composing around dialogue that is the most important part of the mix.

[reid]

londontown, danijel - there's a lot of food for thought in there from you both. Writing around the dialogue goes without saying, but both your points about certain instruments and frequencies have really got me thinking. And the tip about leaving the window open - top stuff

Massey's plugs are out for me unfortunately, I just rely on Logic so heavily for writing that switching to PT for mixing would eat too heavily into schedules. But his AU port of Tape gives me an idea of what I might be missing out on re: the limiter. However, PSP's Xenon limiter is doing a good job of seeing me through most peaks and troughs.

Thanks guys