That 'Commercial Soundtrack' sound

[jonparadise]

This is a questions that may not have an answer, or may have a million answers, I'm not sure.

The general consensus on most forums seems to be that very little plugin processing should be required when mastering an orchestral track, with maybe some compression for synths and drums, and a touch of reverb.

I hope I can produce a half-decent mix, but when I A/B with a commercially produced orchestral track from the movies, it always lacks that 'bass' and 'warmth'.

I've tried a little EQ and applying Ozone's exciter to try and add some warmth to the lower end but never seem to be able to replicate the overall sound.

Is it down to the way the orchestra was recorded to begin with, or is it that the big guys have a vast selection of gear to process the sound?

[tehlord]

It's nothing more than good arranging and mixing like any other kind of music.

Just practise. There are no tricks or secrets.

[jonparadise]

Thanks for the reply tehlord.

Back to more mixing techniques then!

[CableChannel]

Yeah I think it is a lot about high quality sounds and delicate EQing.

I know a guy who produces perfect commercial sound. He uses the best libraries on the market, and as far as mixing is concerned, there are no secrets except his EQing and compression.

[Gamma-UT]

There's a difference between orchestral as in live orchestra recording and cinema orchestral. Soundtracks don't exactly hold back on the limiter settings - they go LOUD.

You might want to compare soundtrack versions of classics versus CDs of classical recordings, eg Ride of the Valkyries in Apocalypse Now versus Das Ring or even soundalikes, such as Horner's soundtrack for Aliens versus Khachaturian's Gayane (obviously not the Sabre Dance bit, although that might have been fun for when the aliens scuttle across the ceiling in their big attack scene).

I can't think of more recent examples off the top of my head but that's largely because in recent years directors have gone for EPIC DRUMS, EPIC GUITARS, EPIC STRINGS, everything feels EPIIIIIIC.

[Casiquire]

The arrangement is all-important!

[Xenobt]

A scooped upper mid eq to keep midrange instruments out of the dialog is one thing I've heard done, and used myself for spots where I had to use a library track. That also explains the airy top end, and warmer bass you're hearing. When you take mids out, you can add a bit more volume without affecting the dialog too.

And it's true, there's a HUGE difference in a real orchestra in a hall, and the soundtrack version. Lots of things I've heard from East/West and the Project SAM stuff are VERY cinematic, so maybe the samples are half the battle as User123 suggests.

Storm Drum is also a big contributor to the epic druminess out there, especially in trailers.

But really, they're just blends of close and distant mics and all the other modern eq/limiter and mix tricks you'd use on any project to make those libraries. There's no reason you couldn't do it too~

KVR/eSoundz: Xenobt

[jonparadise]

Thanks for the info guys.

I think my biggest task is perfecting the EQ techniques. I've also tried the scoop and it does work very well indeed.

[chokehold]

One thing I found out that seems to work pretty well - use more elements at lower velocities rather than less at higher velocities.

I started out with full blast 127 velocities, because hey - it was supposed to sound EPIC so it MUST be loud, right?

But most sampled instruments are velocity layered, which means that they have different sounding samples for different velocities, capturing the dynamic behaviour of each.

Say you strike a piano key slightly, the sound is relatively muffled, because the sound itself is quiet and it has a rather (in comparison to the actual sound) long attack phase with a very short tail. (no offensive jokes, pls )

But if you strike that same piano key really hard, it will have a quick and very intense attack phase, a longer sustain tail; and as it was a loud sound to start with, the string takes longer to fade back into silence which means the note sounds longer, even if you take your finger off the key again.

Most sampled instruments will have captured this sort of behaviour in their samples, which means shorter and softer, less "attack-y" samples for quiet notes, and longer and harsher, more "attack-y" samples for the loud notes.

If you mix together a horrenduous amonunt of lound samples, each with an intense attack phase and dragging a longer, louder tail behind it, you'll end up with a very undynamic mess. Might sound good, but not very intense.

You can investigate this by listening to some easy-to-follow commercial works, just examine a short phrase from a soundtrack and compare it to what you would create with your DAW and sample library to sound roughly like it, all Mastering and Mixing aside. Short clip of 1 or 2 bars is enough, so it doesn't get too complex.

Like that, I came to the conclusion that most "epic" soundtracks with some very loud passages don't actually use loud playing, but the loudness mainly comes from layering.

These are made up values out of context, but might help you understand:

Say the song starts quiet with a cello that plays at a medium 80 vel, accompanied by a soft background ensemble patch at 55-ish vel. Quiet, soothing, calm.

To add some action, raise the vel on the cello slightly, maybe to 90 or even 100, just to draw attention to it - but without triggering the harsh 127-ish samples.
Leave the background ensemble vel untouched, and just bring another element into the equation, maybe a quartet of high-pitched strings or some brass, all at also a 55-ish very soft velocity, maybe place occasional 55-ish soft percussive hits.
This will make the sound more complex and raise the overall intensity, again without really blasting out the 127-ish harsh max-vel samples.

Then, in that short climax that usually comes before a sort-of breakdown, where you might think "this calls for 127 full-blast velocity on all" ... brace yourself. Don't.
Maybe slowly raise the volume of the drum hits from 55 up to 100, because percussive elements tend to live from their attack phase rather than from the sustain, and then just let them play a more complex pattern. Or add another type of drum to raise the complexity, not the velocity.
So even though they instantly make the sound more hectical, they're still quiet enough samples to NOT punch the peak meter into the red and drag that long sustain tail behind it, which would mix with all the other instruments and muddy up your mix.
Might sound impressive on its own, but you don't want 1 timpani taking up all the frequency range for a bass section 5 seconds after that timpani has already been forgotten again.

When it comes to loud parts, always try to think in complexity rather than velocity.

If you feel what you're composing is too soft or too quiet - don't worry about it. You can always bring missing volume and "distinguishability" of elements back in with better EQing, better mixing or slight compression, but you can't make them sound better once the overall sound is drowning in full-blast samples.

Just turn your monitoring fader up, not the velocity of the samples. You should instantly notice how much "richer" and "fuller" everything sounds, but without over-emphasizing the top and low(-mid) end as using full-bast samples would do.

Think about it rationally - someone who plays one VERY loud note a couple of times to capture it as a sample has to do one thing: play a VERY loud note.
But someone who plays a complete track together with an arry of many other musicians has to do so many different things like "feeling" the rhythm, remembering his parts, keeping up with the position - where in the track he is and so on, plus he's playing more than one note, probably quite a few more.

So even though a "live" player could probably occasionally play notes like the 127-full-blast samples during a performance, he is very much more likely NOT to, because it would affect his further playing, like tire out his arm or hands etc.

That's my personal opinion I formed over the last years, it's a conclusion I reached after listening to a lot of different "soundtrack" style pieces of all sorts of quality, from commercial to hobbyist. And, of course, by making my own.

Knowing how these dynamics work doesn't make me a better composer, or my tracks any better, but I think by listening to these two tracks you should be able to notice what thinking in complexity rather than velocity has done for me. Soundwise, it was a big step in the direction of that "commercial soundtrack" sound, don't you think?

Old:
http://soundcloud.com/chokehold/regret-and-forgiveness

New:
http://soundcloud.com/chokehold/wip-track-1-v3

Cheers,
Rob

[annode]

Simple question, simple answer;
- Try using an EQ that supports upper and lower 'shelving'(shelf filter).
I'm sure your Isotope EQ has them.

Is it down to the way the orchestra was recorded to begin with, or is it that the big guys have a vast selection of gear to process the sound?

As for processors, yes, expensive processors are surely used. They have complex algorithms you can't compete with. You can only do the best you can with what you have.