Need a little advice from the audio experts :)

Hi everyone!

I'm having a little trouble with the dialogue in one of my scenes and was hoping some of the audio experts here could point me in the right direction towards making this all fit together.

Below are 4 separate dialogue tracks that all mix together for the complete scene - at the moment there's a distinct difference between them and whilst I've spent some time clumsily EQ'ing them I've yet to come up with anything that sounded good so I figured it's time to defer to more learned minds than my own :)

These are the tracks as recorded (i.e. without my 'experiments') with +5db gain added:

HotD_Sc18_Blake.wav: This was recorded fairly close and since the actor has quite a deep voice it's got a lot of low-end - sounds a little 'dead' and is noticeably different from the wide track.

HotD_Sc18_Dexter.wav: Recorded close again only this actor has a higher-pitched more nasally voice so it doesn't sound quite so out of place when mixed with the wide track.

HotD_Sc18_Waitress.wav: This actresses voice is a little muted and indistinct (even though she was recorded up close) - I'm looking to enhance it to make it more pleasing/fitting.

HotD_Sc18_Wide.wav: Recorded (necessarily) from a distance and seems to have some slight reverb - I'm guessing that it'd be easier to match the other tracks to this one to make it all seamless but I'm not entirely sure of how best to do that.

HotD_Sc18_All.wav: The four tracks above combined into a single mono track. All tracks are 58MB 24/48kHz.

Any and all advice would be gratefully received!

P.S. I'm using Audacity and Sony Vegas for my audio editing - not the best tools but I've got access to some good compressors/limiters and graphic/parametric EQ's - I've read up on and understand the processes to an extent but my main issue here is a lack of experience and a trained ear...

P.P.S. Oops - meant to post this in Post-Production... *doh!*
 
Last edited:
There's no set rules here. A big issue may be your listening environment and your playback system. Being in a quiet, balanced room and listening on clean, neutral speakers when "detailing" sounds is, unfortunately, very important, however. As it is, you are compensating for the deficiencies in your room and playback system while trying to correct several different but interlocking problems; this is why you're having trouble getting close.

And there's no "magic button" that will fix the problem. :lol:

Magic-Button.jpg


I do suggest that you can start with Dialog Editing by John Purcell.

http://www.amazon.com/Dialogue-Edit...Motion+Pictures:+A+Guide+to+the+Invisible+Art
 
Caveat: This is not advice, as I'm an audio noob, and have zero experience, and don't have a proper listening environment.

I can hear more noise and low frequency rumblings when you introduce the wide track, which is unnaturally absent when let's say the waitress is talking. I was wondering, and I'll leave it to the audio gurus to suggest whether this is a good idea or not, if it might be okay to introduce that noise and bg across the entire mix, just so the mix sounds uniform?
 
P.P.S. Oops - meant to post this in Post-Production... *doh!*

In a sense, you may have posted in the correct section! The situation you find yourself in, and indeed the fundamental workflow of all dialogue in a finished film, is a combination of 4 film audio roles (over-optimistically avoiding any mention of an ADR team!): The Boom Op, the PSM, the Dialogue Editor and the Re-recording Mixer. The more skilled and well equipped the production sound team and the better the preparation and filming environment they are presented with, the less time, effort, equipment and skill is needed in post to correct production sound issues. The inverse is also obviously true, the poorer the production sound, the more this has to be compensated for in post. In other words, your current issues could possibly/probably have been eliminated during Pre-Prod/rehearsals with a good PSM noticing the potential audio perspective problems in this scene and taking the necessary steps to alleviate them in the first place. As a general rule, this is the best solution for providing the highest quality final result.

Given that you're presumably well beyond this point, the only solution remaining is on the audio post side of the equation. I'm also assuming that these are the only takes available or that these are the best and all the other takes have the same issues. If this is the case, then the solution to your audio perspective issues is restricted to only what can be accomplished by the Re-recording Mixer (you!). However, there are also many significant dialogue editing issues, in addition to the audio perspective problems, and Alcove's book suggestion is therefore an excellent one, as it covers not only dialogue editing but also wider dialogue considerations. I also agree with Alcove on the vital importance of the audio monitoring environment.

Providing you with accurate/useful advice is next to impossible without the accompanying visuals. The audio perspective issues might be just as bad or potentially even worse than they sound in the audio mix you posted or, they might not be nearly so out of place, depending on their relationship with the visuals (edits/shot widths/angles). This could make the difference between some relatively minor tweaks, a fairly complex combination of specialised audio processors or anywhere in between. The final solution will definitely involve some EQ'ing (as you have identified to a limited extent) and some re-balancing but what other types of processing, I can't say without reference to the visuals.

G
 
Last edited:
Strictly technical review....

You need to use more LPF (low pass filter) in the EQ to eliminate sounds below about 60Hz (rumble). I'm surprised phase problems aren't worse than they are.... I do hear some phasing but it's not terrible.

Biggest problem.... you need some room tone to patch these edits together. Hope you've got some from that location! You also might be able to "unlock" the audio and drag your handles out... Right now it sounds like the audio editing is just rough cuts from the video editor. Pretty choppy.

I also think you'll need to do some leveling with automation, compression, et al.
 
Last edited:
Firstly, I want to thank all of you for taking the time to respond - it's much appreciated!

Alcove Audio: Sadly all I have available to me is a pair of Sennheiser HD201 headphones and a pretty capable workstation - no speakers and no budget to buy any (and I really mean *no* budget) - the audio post was never supposed to be done by me in the first place but that's a story for another day. It's my hope that if I can get the movie finished to a standard that I consider 'presentable' that I might find the means to have the audio done from scratch by a professional - until then I'm forced to work with what I have. And yes, as a VFX guy I'm well aware of the mythical nature of 'magic buttons' - thanks for the book recommendation - it looks like an excellent place to begin :)

trueindie: Thanks for the observation - my PSM did record room tones for all the locations/sets so perhaps adding that sound into the close-ups will help tie it all together.

AudioPostExpert: To be fair to my PSM, he did point out issues with the space and we followed his advice in trying to mitigate them as much as possible but shooting elsewhere wasn't an option (it was the only place within our budget large enough that we could build the sets we needed). And yes, you're correct in assuming these are the best takes and I'll certainly be obtaining a copy of the book Alcove recommended.

One rather large directorial mistake I made in this scene was not picking up on the actors projecting in the wide shots and then talking in more hushed tones in the close-ups - there's supposed to be a somewhat conspiratorial aspect to the conversation but they're actually in a club so the close-ups should've been a bit louder.

In retrospect, it was dumb of me not to post the video - since I'm new to this, I find I can focus on the audio better without the distraction of the image but then, I know the type of environment I'm going for and context, of course is everything - I've uploaded a video of the scene here. Unfortunately this scene is a green-screen shoot so I'm not sure how much use it'll be on its own - to help give you an idea of the kind of ambience I'm going for, please see these stills from the second film in my favorite trilogy ;)

There'll be other patrons (though less than in the images) and some music - likely an instrumental version of something by Bessie Smith - played somewhat subdued on a gramophone or radio rather than live.

soundslikejoe: Thanks Joe - yes, there's no HPF/LPF on these clips - they're just sound rushes straight from the edit with +5db applied. It's not quite time to do a full audio pass on the whole film but since I'm doing *all* the post-production work, as I come across things I know are going to be a problem later, I'm flagging them and doing some research/learning so that when the time comes there aren't any nasty surprises. And yes - I've been reading up on EQ and compressors, limiting and loudness and Dialnorm and EBU R128 and... and... *head explodes* :D

If any of you have some pointers on how I might be able to balance these takes out, it'd be a really useful starting point for me - thank you!
 
If you could even get a decent set of speakers for your editing computer just to give you some perspective it would be $100 well spent. (BTW, I have a pair of medium quality computer speakers for checking the mixes of projects aimed specifically at the small screen market.)

To make it "presentable" the dialog tracks need to be relatively smooth. The problem is that your monitoring situation will not let you hear the worst culprits. You have four interlocked missions - noise reduction, putting the disparate pieces of dialog into perspective with each other tonally, reducing the background sound (as opposed to actual noise reduction), and creating a smooth, hopefully unnoticeable, space for the dialog to live in. Then you get to mix all that in with your music and whatever else you may be putting in. (Foley? Sound Effects?) Do a basic noise reduction of all the clips - you'll have to do a lot of automation. Identify the biggest noise problems, usually rumble, hiss, or the infamous 60 cycle hum. Now lower the volume between the lines; I prefer cutting those sections out and dropping them on another track. Put in your track of room-tone and do noise reduction on it. Now you just have to play with the EQ until you get everything sounding the same. Keep lowering the room-tone and the volume between the lines of dialog. Keep working that noise reduction. Keep tweaking back and forth until you get it to your satisfaction.

A well constructed ambient/background track can cover a multitude of issues, but only do it AFTER you've made the dialog as good as you can get it.

That's about as "in a nutshell" I can make it.
 
Speakers are on my list of things to get but even $100 is more than I can currently afford unfortunately - I'll have to make do until my circumstances change knowing that the audio will need another pass before I can monitor it with any degree of accuracy.

Thanks for your 'in-a-nutshell' explanation - it's certainly given me plenty to think about - I've just downloaded a 60-day trial of Reaper (which looks like the only DAW within my budget that'll take EDLs) and I'm reading through the book you recommended - lots of new concepts to get to grips with but (so far) it all seems to make sense. Thanks again for the help!
 
Back
Top