Wall of sound - sifting through it

Question

I sometimes get sound effects turned over to me that have little to no skilled premixing done to them.

And, I am expected to mix it within that day of receiving it (for a 2 minute trailer for example).

Sometimes there are quite a few elements to work with.

I'm curious to how you sort through and essentially "mix" the elements. Do you solo each element and listen to what it is before determining what level it should be at?

That's essentially what I do but it's time consuming... I'm trying to speed up how I do these things.

I'm curious to know if there was premixing done before-hand do you let it roll and just tweak small things and leave the premix in tact? Do you delete all the automation and start from scratch?

Also, how often do you mute and get rid of elements from a mix?

Thanks - Ryan

Answer 1

Try putting everything into stems and then be dramatic with the EQ and volume automation. That way you can leave, just about, everything in so that it comes in on cue nice and loud and then drops down into the mix without cluttering everything up.

• Start of cue = flat and loud
• Mid cue = eq'd to sit in mix and moderate level
• End of cue = heavily eq'd and barely audible

I learned how to do this with bands who refused to cut anything, wanted to hear everything, and still wanted the mix to be 'clean'.

Answer 2

Hey Ryan,

I guess you're mainly talking about sounds that have been layered up together. Sadly I never get anything pre-mixed. I normally just improvise and play through with randomly set faders and see how it sounds. If it sounds cool and works with the picture great, if not I will actually scrub each track to see what's in there. I will more often than not strip out material that is surplus to requirements.

For atmos/ambience I normally solo that group, suspend automation (I use ProTools) and get the balance right. I then apply that automation to the scene.

I like the mixture of improvisation and careful thought.

Answer 3

The editors here do rudimentary volume premixes before they deliver to me which I usually bring in with the tracks. Do a pass to listen and place markers at points where stuff needs tweaking (usually I mark them depending on the work needed eg A for total remix/re-edit, B for volume tweaks, C for too many fx layers etc - I've about 10 'codes'). Then I Go through marker by marker and either remix scenes in touch mode or use Trim automation for tweaks. Generally I don't go digging unless its a total redo on a sound. Once I'm happy I have about 10 stem faders which I do a overall mix with.

On smaller shows where I might only have 20-30 tracks, the marker workflow isn't as necessary but for my current series (over 80tracks per ep) its easier to manage.

Answer 4

When I was first delivering FX stems to one of our mixers, I would go through painstaking effort to premix. I would pan and automate everything to be able to deliver a final, re-recording ready stem. It was only after about the 4th or 5th show I delivered this way that I found out that he was deleting everything the moment he imported the session data. Meanwhile he was wondering what the hell was taking me so long... As a result, pretty much everyone here is delivering elements with zero premixing.

So first and foremost, communication up and down the chain is key. If you can make sure that whomever is providing your elements has a clear understanding of what you're expecting from them, then you're not wasting their time or yours.

Now, if you don't have that luxury, or (as your phrasing suggests) the ones providing your elements are just doing a crap job about premixing, here's what I do. Since I've wasted a bunch of time myself soloing up individual elements of a build to see what's what, I start by throwing up all the faders to get a balance. If something sounds out of whack, rather than soloing each channel, I mute where I assume the problem is. Usually I'm looking at meters and waveforms to make my best guess. I've found it faster to listen in the context of the entire sound, mute the offenders to find them, pull them back, unmute and gradually add them back until I like what I hear. This way I'm mixing more of the sound at once, which usually results in fewer passes of the sound.

Now, because I know the guys who are effects building for me, and I know how they work, I know that elements higher up the track constitute more "important" elements to the design. As a result, an effect placed on the "FX 4" track will typically play louder than something on "FX 12." But that's usually just a starting point. So I'll write the basic balance automation, then dive into trim to detail its mix. After I've gotten the effect mixed, I'll use a VCA to feather it in against the rest of the mix.

As far as muting clips goes, I rarely do it. There are occasions where I don't agree with what they delivered and I may have to go hunting for the sound that I think should be in there instead. But more often than not, rather than shift gears completely, I'll EQ or mix down the bad elements sooner than getting rid of them completely. There usually some reason that it's been added to the design.