3 dialogue editing questions

[pechnatunk]

Hi, can someone help me out with the best technique/tools to matching spectral qualities of different dialogue files - ones recorded with different mics and different noise floors etc.
I know most of you might say trust my ears - but it is hard to get it perfect.

If I need to use some noise reduction - is it best to try and match the spectral qualities of the dialogue before using any noise reduction or after.

On a different topic - I've been setting my nudge value to 40 ms to match 25 fps but is it best to actually use the grid so the cursor snaps right to the frame.

Also - are there any audio advantges to using lapel mics - On every film I've done post for I've ended up disregarding the lapel mic because of crackling, muffled sound and what seems to be poor quality in general.

[Uncle Bob]

I'm going to say the obvious - you need to use your ears.

Your methods will depend upon your editing and mixing system. If you have tons of processing power you can have NR and EQ of every channel. If you are using a native system you may have to process and print - and then go back and start over if it doesn't come out right.

I work with a native system, so I start with checkerboarding. I first separate out by the noise. Track one is room tone, then I work down from horrible production sound to really bad, then bad, then passable. Using NR and EQ I try to get all of them sounding the same. Then I edit out the production sound "air" between lines of dialog. Then I EQ to get the individual lines of dialog sounding the same. Once I'm satisfied I reorganize by character and print the results.

It's much easier when I'm freelancing and working on a TDM or HD system; the unlimited processing power means I don't have to print anything and can leave it all up for the rerecording mixer to tweak as s/he desires.

Nudging is a personal preference and may vary from project to project. I like quarter frames myself.

I dislike lavs myself unless they were set up by a real pro; they tend to be clearer and have fewer clothing noises.

I'm sure that others wiser and with greater experience will chime in with more and better ways of working.

[Lipflap]

Originally Posted by pechnatunk

Hi, can someone help me out with the best technique/tools to matching spectral qualities of different dialogue files - ones recorded with different mics and different noise floors etc.
I know most of you might say trust my ears - but it is hard to get it perfect.

If I need to use some noise reduction - is it best to try and match the spectral qualities of the dialogue before using any noise reduction or after.

On a different topic - I've been setting my nudge value to 40 ms to match 25 fps but is it best to actually use the grid so the cursor snaps right to the frame.

Also - are there any audio advantges to using lapel mics - On every film I've done post for I've ended up disregarding the lapel mic because of crackling, muffled sound and what seems to be poor quality in general.

Matching - Since this is the Holy Grail of dialogue mixing, you will undoubtedly find as many valid answers as you will find editors/mixers. But two things will make your life much easier:

• Organize your regions by shot (which conveniently is the same as sound source). Don't start editing until this is done, as it will save you lots of grief throughout the process.
• Remove as many transient noises as you can through editing, rather the through aggressive cooking in the mix

.
NR- Noise reduction and "shaping" go hand in hand. You can't manipulate the one without affecting the other. It's "a bit of this, a bit of that" until you have more or less what you want. And again, ask 100 mixers, you'll get a like number of "best way" options.

Nudging - I like using a grid, but some say I have control issues. Truthfully, po-TAY-to, po-TAA-to.

Oh yea, it really is about trusting your ears.

Good luck,

[charles maynes]

I think you should get John's book, his reply is drawing me towards buying it, and I dont edit that much dialogue.

[Flix]

His book is great, even if you already know how to do it.

[Pedro_vde]

Indeed, trust your ears, and before all read Johns Book !!!

Just to add, I never use grid on dialog editing, just use slip, as your ears don't listen in frames... an edit in dialog has to come where it has to come, and that can't be chopped in frame-like parts.

In general I use the lapel-microphones as a problem solver. When the boom is to late or to far or something like that. In general I tend to use just the boom(s)

It all depends on the Production Sound Mixer and how much effort he did on matching the sound of his lapel-microphones with the boom microphone. It also depends on how much attention is put in the placement to prevent rustle etc.

kind regards from Belgium

Pedro

[Sventeck]

I agree with all above... I personally work in Grid mode as well, I assume you're using PT, if you want to work in grid, hold down command and you temporarily can edit in slip mode, without changing back and forth a lot. Handy with fades etc. I change nudge between 1 frame, and 1 subframe but I nudge a lot.

Thanks for the book John, really gave me a clue when i started!

[pechnatunk]

Thanks so much for your replies - it is very helpful - especially the common sense stuff like the region arrangement advice which is actually different to how I was taught ( I can switch back to character arrangement for mixing. I'll definitely get that book - however not in time for this project.
I think the best tool for NR is actually sound notes and subtle time-stretch/compression - maybe I was being lazy but I didn't search for alternative takes since I didn't have any sound notes - but after stumbling upon a usable take I sifted through the 300+ 2x audio tracks and luckily found some useful stuff.
As good as izotope is as cleaning jackhammers and circular saws from my cocktail bar scene - an alternative take is better.

The matching is coming along fine - just wondering if there is an easier way to automate plugins - at the moment I've been turning automation off - auditioning my eq and making note of the settings then "pointing' in the automation. If I use write, it writes every parameter into automation when maybe only one or two is actually different from the last region.

Also - I think I'm nearing an end to the dialogue editing - read a fair few threads here about compressing dialogue - I've never felt I've needed to before but might give it a try on this one.
Would this process change much where there is a fair bit of foley on the same take - I've found there is nothing wrong with it.
And would you normally expand a little to hopefully get rid of some noise at the same time?
Vic

[Jussi]

Originally Posted by pechnatunk

I can switch back to character arrangement for mixing.

Don't. The whole point of organising by angle/slate/perspective is to make mixing smoother and faster. Also editing tends to be faster and results in less clutter and problems. There is really no need to organise by character.

Would this process change much where there is a fair bit of foley on the same take

You could mix atmos and dialog at the same time to get the dialog sit in the environment and see if there is need for expansion or noise reduction. But this is mixing. If you're doing dialog editing, I would do only that and leave all eq'ing and compressing etc. for the mixing engineer to handle.

[Jfriah]

Nice 'props', Charles. I agree with the theory: even if you've been doing something awhile (or not), it is always good and educational to read / hear someone else's ways of thinking/working.

And am I loopy or did you recently get "made" as a moderator? Congrats.

-Jeff

[Jfriah]

[QUOTE=Sventeck;4486891 if you want to work in grid, hold down command and you temporarily can edit in slip mode,[/QUOTE]

Not really sure what you mean here, Sventeck, unless I'm just missing it?
Holding down Cmd when in I-Beam/Cursor brings up the alternate takes list; and in Grabber/Hand mode, will shift the material within the region, but by whatever your nudge mode currently is.

Neverminnnnnnnd----in Grabber/Hand mode, if you are DRAGGING/moving a region, if you hold Cmd you can, yes, move regions within a grid space. Nice one.


Learning every day till I die,
-Jeff

[charles maynes]

Originally Posted by Jfriah

Nice 'props', Charles. I agree with the theory: even if you've been doing something awhile (or not), it is always good and educational to read / hear someone else's ways of thinking/working.

And am I loopy or did you recently get "made" as a moderator? Congrats.

-Jeff

thanks -

I was made one of the moderators in the Political forum....

[danijel]

Originally Posted by charles maynes

I was made one of the moderators in the Political forum....

Haha, I bet they need a lot of moderation over there

[Sventeck]

Neverminnnnnnnd----in Grabber/Hand mode, if you are DRAGGING/moving a region, if you hold Cmd you can, yes, move regions within a grid space. Nice one.

Sorry I should have been more clear. I always work with the multi-tool.... if you hold command and grabber, you can drag in slip. Command and trim will trim in slip. Also Command and fade portion of the super tool will also fade in slip mode.

Another nice one is control+trim will allow you to trim out a region up to the next one without overlapping it...

[Lipflap]

For normal dialogue edits (within a scene), the grid mode is not of a great deal of use, and it may even lead you to trust visuals over trusting your ears as to what makes a good edit. But I highly recommend using this mode for scene transitions, as it creates more uniformity and control in scene-to-scene transitions (since you know how long the crossfade is) and it helps you to make the same kind of scene segues as the background editor is doing, thus creating more solid scene changes.

[RScott]

I use grid mode in dialog editing for general changes, because it makes it easier to correct if you need to nudge it back to the original timing (say if something gets bumped out of place and you don't notice it until later, grid can help you get it back exactly where it was with less hassle.) Then if needed for fine tuning and things that need to be tweaked beyond a frame (many things do!) I'll work in slip mode.

For matching mics, EQ and dynamics tools can get you into the right ballpark, but sometimes a different mic is a different mic and you will just have to get it as good as you can. Sometimes shaping two sources in the same processor can give them a similar color (eg: reverb, some harmonic plugs, etc.) but I would probably not go to a processor that I didn't want for it's primary effect just to add uniformity.

For noise floor matching, your room tone recordings are a good tool to use. On set, you should have a minute or so of recording in the room from the mics that were used in the shot. If you use this room to fill inbetween lines of a character (while character B is speaking on mic B, add the room tone from mic A to Character A's track.) This prevents a lot of the jarring dialogue edits in a two shot. This raises your noise floor because you will constantly have noise A + noise B, but you won't have to worry about matching as much if this technique is working and you can sneak one in or out if that character/mic is not doing anything for a while.

I do a NR pass before matching dialogue, but maybe if your dialogue matching is bringing out some of the nasty artifacts of your noise reduction process, then try swapping. If you're really keen you can do it all in one big pass, but I like to have individual processes separated out so that it's clear later on what in the world I was doing.

Lapel mics can make decent "oops" mics for when you have something screwy on your main mic. They can sometimes give good results, and are always nice to have as options. They can be good for getting a dialogue recording with less room, especially on a wide shot when the boom can't get in. They can give good coverage of multiple sources. Often times performance situations, poor mic quality and poor mic placement create major issues for lapel mics.

[Airon]

I go for: same sound, same track. Same dialogue sound same track. Chequered when necessary for the mixer(which is me a lot of times). Perspective cuts, different track, but watch the roomtone transition. Does the mix require enormous gain change on the next line ? Then cater to it. If possible get to know the dialogue mixer and their preferences at the first opportunity. That person is your collaborator as much as the production sound mixer, the picture editor (and their hopefully excellent assistants who know where all the files are) and the director.

You'd best attend all the dialogue mixing sessions you can, to get a feel for it. The mixer depends on you, or else someone will need to fiddle with your edits during the mix.

And keep a non-denoised edited copy of the material handy, ready for use during the mix. Deactivated tracks are your friend(Protools).

[K5000]

Ah, a book on dialogue editing. Might be a nice read.

I so far don't have post experience outside school, and unfortunately the plugin package that our Pro Tools HD's have is pretty basic. Just the Digidesign stuff and a basic Waves package. I find room tone often doesn't pose a problem, really... by doing careful work on volume automation it seems rather easy to drop the level of the room tone quite a bit after the actual line without it being obvious.

I haven't touched a lot of NR, more like finding the most problematic frequencies on the bg noise and applying sharp EQ nudges as much as I cain without too much reduction of the voice quality. Any tips on situations where NR might be of exceptional value?

Also, suppose you've got a line that was recorded on a MKH60 or some other interference tube, and the boom op has messed up a tad and the mic turns during a line, producing that funky yet unwanted phasey effect. Any ideas on how to lessen that?

[philper]

Buy the Purcell book. Display it in your edit room so your clients will know you are a well-read person.

Philip Perkins