SFX mixing

[Dali_dBStudio]

One of the recent tricks I use is to have a FX Master BUS and DIALOG Master BUS. Then you compress FX Bus with sidechain DIALOG input using a special tool called spectral dynamics.

It behaves similar to other sidechain compressors, compressing FX bus based on vocal sidechain input, only it doesn't compress wideband signal, but only frequencies in DIALOGUE tracks.

Simple it creates a freq space inside FX signal for dialog to ride over. That way you can have a blast of effects, but will always cut a freq space when Dialogue comes.

Plugin is called:
MeldaProduction MSpectralDynamics, spectral dynamic processor & free-form linear phase equalizer

Br,
Dali

[ggegan]

Why would you want to arbitrarily apply compression to everything, and then at the same setting? That is a very imprecise and random approach.

I don't compress the vast majority of sound FX, with the exception that I will compress Foley footsteps if needed (usually when they are poorly performed and too peaky). When I do use a compressor on sound FX I always do so at the source track, never from the Aux sub master, because I want the processing to be specific to the individual sounds and their issues. The only processing I put on the Aux sub masters is a limiter set to the highest threshold I can get away with and I try not to hit it too hard.

I find that well-recorded, properly chosen sound FX often don't need anything more than a slight tweaking of the extreme high or low end or maybe some reverb, and many times they need nothing at all other than the right fader moves.

[FullFrequency]

Originally Posted by ggegan

Why would you want to arbitrarily apply compression to everything, and then at the same setting? That is a very imprecise and random approach.

I don't compress the vast majority of sound FX, with the exception that I will compress Foley footsteps if needed (usually when they are poorly performed and too peaky). When I do use a compressor on sound FX I always do so at the source track, never from the Aux sub master, because I want the processing to be specific to the individual sounds and their issues. The only processing I put on the Aux sub masters is a limiter set to the highest threshold I can get away with and I try not to hit it too hard.

I find that well-recorded, properly chosen sound FX often don't need anything more than a slight tweaking of the extreme high or low end or maybe some reverb, and many times they need nothing at all other than the right fader moves.

Thanks for the reply Gary! this is exactly the sort of answer I was after!

I'd say it's not the case of wanting to compress everything the same, it's just more that time and budget haven't in the past allowed getting so precisely into it. I've had specs to adhere to so the aux compressor was allowing me to push louder sfx through it whilst still keeping within spec, but not coming close to touching more natural sfx that didn't need compressing for any reason. For the most part, none of the SFX were getting compressed the way I was using it.

I think also that because I've spent so much time on dialogues, with quite simple SFX lays up till now I've naively thought I could have a similar thought process towards SFX than I do the dials. Now that I'm getting more interesting work in terms of SFX I've realised I can't achieve what I want with the way I'm working, and hence the question.

Obviously SFX tracklays can get pretty intense once you start going up the budget ladder. Would you find yourself using other additional methods than the comp and EQ combo in special situations, and if so what might those situations be?

Thanks

[MixMasterM]

I think the reason there is so much more discussion regarding dialogue mixing, is that dialogue mixing requires background noise clean up, camera or tonal filterring, matching of several different takes, etc. while SFX mixers generally are working with elements that have been mastered or are well recorded and layed out......generally :-). So, a sfx template can be much more straight forward than a dialogue template, and if the tracks are well layed out and well recorded, they shouldn't need more than volume moves and occasional touches of reverb.

In my personal setup, I only have eq's inserted on each of my tracks for the occasional shaping. I also have compressors on my foley tracks, as they will require additional work. Each track has sends to several reverbs (2x monos for production matching for SFX and foley, and 1 or 2 stereo and/or 5.1 verbs for effect. For film work, I generally just have a peak limiter set at -1dBfs on the stems master, but occasionally for TV work I'll set it lower, depending on the broadcast spec.
The other thing I have set up is group vca masters so that I can do group pans, eq's or trim whole sets of elements (ie guns or cars etc). Also depending on the project, I may have additional auxes set up for specific things. Ie LFE sends, futzes, or specific processing.

[danijel]

A neat trick I recently discovered is having an aux for loud SFX (gunshots, bodyfalls, kick'n'punch, explosions) with a limiter on it, routed to the main SFX aux (which has a limiter of its own). That way, I can limit the sh_t out of loud stuff without affecting other SFX, especially continuous ones, like car engines, and without doing it off-line.

Otherwise, I only have a bus limiter on SFX aux, EQ on every individual channel, and generally use Transient Designer (off-line) more than compressors.

[JSt0rm]

Originally Posted by ggegan

Why would you want to arbitrarily apply compression to everything, and then at the same setting? That is a very imprecise and random approach.

I don't compress the vast majority of sound FX, with the exception that I will compress Foley footsteps if needed (usually when they are poorly performed and too peaky). When I do use a compressor on sound FX I always do so at the source track, never from the Aux sub master, because I want the processing to be specific to the individual sounds and their issues. The only processing I put on the Aux sub masters is a limiter set to the highest threshold I can get away with and I try not to hit it too hard.

I find that well-recorded, properly chosen sound FX often don't need anything more than a slight tweaking of the extreme high or low end or maybe some reverb, and many times they need nothing at all other than the right fader moves.

Listen to this man.

[renec]

someone said somewhere that sfx mixing is 90% editorial (wish i could remember the source)

with that said, I'm personally VERY conscious of frequency content, dynamics, dynamic range and distortion content when I'm cutting sounds. My philosophy is that sound design is ingredient based cooking, and as such if you find or record excellent source material and lean towards having those things out front and alone then you'll end up with better stuff than if you layer 100 things on top of one another to try and find the "beef"

often, I'll look for specific textures and types of movement in leiu of specific sounds when I'm building something fantastical or hyper-real. I'll make sure that the low end is big and punchy, then find the right balance of high and mid textures to get the sound just right.

i'll also often do extreme clip-based processing as part of the design process to get things into the place that I want them. This includes things like rendering distortion, crushing transients, bandpass filtering, etc. This all happens well in advance of the sound design being mixed in context - and as such its not as much about making it play well with others as it is about making into something new that I like initially. hyperbandpassing something is as much a design (as in non-mixing) decision as reversing a sound is IMO.

On the larger tracklays, its common to "sandbox" sounds out in a separate session and then print food group stems that end up back in the main sfx sessions. This allows me to go nuts with plugins and processing, and forces me to make certain decisions. If those decisions don't end up working in the context of the mix with moderate effort, then its my responsibility to re-work the initial ingredients in a way that does.

If I have temp music or a score, I'll always design in context. Sometimes score only leaves so much room to play, and in those cases arrangement is better than esoteric mixing techniques.

My .02

[postprosound]

Perhaps I'm being to philosophical here, but whether you're an fx editor or mixer your one and only job, (other than clearing for dialog, of course) is to advance and add to the story. Nothing more, and certainly nothing less.

How you get there is of secondary concern. In fact, it probably doesn't matter that much. If it sounds good, the director is happy, and you've done your part to collaborate with dial and music, then it is good. Enjoy your success.

Obviously this is a gross oversimplification, but in principle, it's true. We have specs to meet, and deadlines to make, but don't make this more complicated than need be. I think anyone who's been on a stage with 250-500 or more fx tracks can attest to this.

I most certainly agree of the importance of good editing. It doesn't have the glamour of the knobs and blinking lights of a dub stage, but it's where fx are won or lost.