Mix Advice - EQ suggestions for this sample?

Matthew Bennett

Matthew Bennett

Well-known member
Hello experts...

I'm currently doing a rough mix (rough meaning my abilities at this time) for my film.
Below is a sample that perhaps some of you might be kind enough to give me advice about.

Here's a brief project history.. I shot a film recently, and used a Sennheiser G6 wireless mic for the entire production. I only had a single mic set so I placed the lav on the lead actor and left it there for the whole shoot. We worked 90% indoors in a very reverberant, creaky wood-floored apartment with a lot of ambient street noise leaking in. This gave me no usable production sound, but did give me a decent scratch track to do ADR with.

After some great ADR technique advice from members of this forum, I recorded the ADR, and feel confident that I've not only captured the spirit of the original performance but worked with the actors to add to their performances as well. I used a Audio Technica AT815b long shotgun mic, close technique, phantom powered, overhead, pointing right at the mouth/chest area, in a well insulated, blanketed area. (recording to tape on my DVX) I know that some say that the AT815b is a little bit hollow, and I would tend to agree even if I haven't heard samples from many other mics. (its all i got) But, I felt like my technique was very good, so at least I have a very EVEN set of new vocals to work with.

In the last months I've been painstakingly engaged in the sync process, which is a whole zen unto itself, but I'm learning as I go along.

Now I come to the mixing and EQ'ing aspect. This is how my mix goes... I take all my very flat foley and vocals and mix them together. (This is what the sample consists of)
Then I take that mixed track (I figure this would be a simulation of what a good production track would sound like, an 'image' of the room if you will, only an image of the room as if there were sound blankets everywhere)... and I bring it into Soundtrack Pro and add a little Reverb and then some EQ (a preset I've found which favours the mids a little). Then I take this EQ'd, louder track back to Final Cut and add room tone, city ambient tone, streetcars, sirens, dog barks, street walla, and music... whatever the scene needs.

The final mixes are sounding a little harsh and cold to me. I want a bit of a warmer tone, a bit more masculine. Perhaps a bit of a rounder, fuller sound.

So I guess what I'm asking is... based on the sample below (this a mixed foley/vocal, straight from the shotgun with no EQ, panning, anything, all recorded in the above conditions)... how would you EQ this for a final, out to DVD to sound good? (Good ie: see above, warmer, fuller)

My final final output for the audio is an AC3, dialogue normalized at -27db, for DVD video.

I've also included the Soundtrack Pro settings I used as a jumping off point...
Thanks for any tips or any advice at all!

Link to sample: (remember this is non-eq'd, straight from the AT815b)

http://www.stickypod.com/videos/uploads/432/Tahiti_Sample_1.aiff


Pics of my settings:
Reverb_1.jpg


EQ_1.jpg



Thanks!!!
 
Last edited:
Hey man... that stuff sounds really really solid. And it's got plenty of high end, so I wouldn't apply that EQ curve that I see there to it. If anything I would do a slight inverse of that... by maybe pulling out touch of 4K and adding a bit of wide-Q 400Hz or something if you want to get a tad more warmth. But it sounds damn good, so don't go too tweak crazy. And you shouldn't put much (if any) reverb on the stuff... if you listen to well-produced stuff, it's almost always slightly artificially dry, choosing clarity of the dialogue over room-realism.

Also a bit of compression never hurts, again, don't go overboard. It shouldn't "sound" compressed, but it smooths things out and helps keep all the dialogue pleasantly intelligible over whatever b/g sound & music you add.

A little bed of room tone/walla under that for whatever location, and I think you've got yourself very pro audio.

-Kevin
 
hemophilia said:
If anything I would do a slight inverse of that... by maybe pulling out touch of 4K and adding a bit of wide-Q 400Hz or something if you want to get a tad more warmth.

-Kevin
Click to expand...

I tried it, thanks for the advice, it sounds a lot smoother now. I guess the theory is that there's already a lot of top end in the vocals, so to 'smooth' it you mute the top frequencies a bit an then do a wide-band boost on the lows, which evens out the overall frequencies in the signal.

Thanks also for the advice about reverb. I've read so many times about adding a touch of reverb to ADR, so I was sort of addicted to the idea even though I never really liked the results. I think I'll scrap it altogether and go for the 'dry clarity' you mentioned.

Thanks Teacher!!!!
 
Happy to help.

The easiest thing to do when EQing... if you're not liking the sound is:

Well first of all, don't load a preset. God only knows what that preset was intended for. Presets in things like EQs really boggle my mind, except for very simple stuff like a rumble filter, but then again you could dial that in so fast why would you need a preset?

So back to the tweaking methodology: Play your sample, load up your parametric EQ. Do a stupidly large boost in one band (like all the way up... turn it down first so you don't blow up your speakers and your ears), with a really narrow Q, and then sweep it around until you pinpoint what you DON'T like about the sound. i.e. if the sound is too harsh to start with, boost the EQ and sweep it around until you find those harsh frequencies and the sound is ripping your head off. If the sound is too muffled, find those boomy frequencies that are totally unitelligible, etc. You want to make it sound as bad as you possibly can.

Then set the Q... by broadening until you're hearing the more pleasant components of the sound start to creep in, until you're hearing a broad enough spectrum that the sound is beginning to head in a tolerable direction, then back it off a bit (make narrower) so it sounds unbearable again.

So that pretty much sets your frequency band exactly where it should be... then you can get the gain set properly, which will be a little cut... depending on the source (in the case of those clips you sent I would only recommend a very small cut b/c it's pretty close). One of the things I like to do when setting something critical like an EQ on overall dialogue like that... is to first reference something else, say a Hollywood movie that you think sounds great that you want to emulate. Then switch over to your sound, and grab slider, then set it without looking at it. Close your eyes if you have to, and LISTEN listen listen. People get too caught up in presets and their time-tested-routines that worked once in the 70s and habit and all that. The ONLY thing you should trust is your ears. Once you slide that thing around a bit with your eyes closed, you'll have totally forgotten where the zero mark is, and you'll have no idea if you're boosting or cutting or not changing it at all, and you'll be forced to make a decision based solely on what sounds best, which is as it should be.

This of course being 'subtractive EQ', in that you're reducing the volume of a band instead of boosting. As a broad rule of thumb, subtractive EQ ends up being more transparent, i.e. it doesn't "sound EQ'ed" as much. So for something that is supposed to sound natural and unprocessed like dialogue, it's not a bad idea to generally shoot for a subtractive EQ. If you do need to boost something, just try to do it with a wider-Q setting, which again will make it sound more natural and less like it was EQ'ed. Err on the side of underprocessing it, rather than tweaking it all to hell.

And above all, trust what sounds good, not what the sliders are telling you.

Who knew I could ramble for so long about setting a band on an equalizer? :)
 
I was going to give the subtractive EQ advice and then I saw you covered it.

One thing I noticed, and it may not be a problem with the full mix, is that it's a little "roomy". It's got a bit of that small room bounce to iot that doesn't sound so great with close micing. Rolling off some of the highs probably helped. If it was a real problem you can use an expander (sometimes) to kill a bit of the room sound. But it sound really good, clear etc. The room tones (the ones you have added) will probably eat up all of the "roomienes" i was refering to.
 
Such great advice, thanks. It's actually the first time I've had a handle on the EQ process in general. Thanks for the knowledge.
 
I agree with everything else that was mentioned. I would roll off a little on the low end, though. It sounds like the guy with the mic on is a little boomy. Besides that, it sounds natural, which is a good thing.

-Dave
 
I agree with hemophilia. You really want to take out what you don't want first. If there is any room left you can boost a little. People tend to boost first and they are really only kind of hiding the problem.
 
These comments have really made the lightbulbs go off for me as far as the how and why of the eq process... Thanks so much guys.
There is a real art to the final mix ... sort of 'planting' the dialogue into the mix so everything seems natural, but without sacrificing clarity. I have been learning a lot these last weeks.
Cheers!!
 
You must log in or register to reply here.
Back
Top