How do they edit film sound?

C

clank72

New member
I come from a very extensive audio background with multitrack editing and music production and sampling. But using off camera sound and video to make a movie totally makes me lost

So lets say you have a ton of WAV files from a field recorder, most recordings are slated or have claps. How do they normally get all the audio merged with the video. Then lets say you have all the audio merged with the video, then which software do you fine tune the sound cuts in and out of edits? Non-linear video editors (Like Final Cut) have very limited audio editing capabilities and sound tools and are somewhat tedious to use apposed to an App like Cubase or protools.
 
Not an audio pro, I only hire them, but I come from a music background so I understand how important they are.

My guy uses protools which is pretty standard I think. getting into an area of workflow I am a little fuzzy on, but I believe you export the project as an OMF file from the NLE which can then be imported into Protools for the audio edit. The lack of OMF support was one of the reasons FCP X was almost universally panned by pro editors.
 
Last edited:
So the OMF is a data file which then contains the position of the audio clips along the timeline, then pulls in the "linked" sound clips from a nearby folder placing them into the timeline for editing. Those clips can then be moved to different tracks for let's say... adding a noise reduction plugin on a dialog clip etc. And then sound effects are applied to other tracks while viewing the video track?

So in a nut shell, there is only very basic audio cuts until the entire movie is edited from start to finish. Which means the film editor is listening to raw video audio for reference purposes to perform good edits, all fine tuning and enhancements to sound are completed by the sound engineer. Does this sound accurate?
 
Last edited:
clank72 said:
So the OMF is a data file which then contains the position of the audio clips along the timeline, then pulls in the "linked" sound clips from a nearby folder placing them into the timeline for editing. Those clips can then be moved to different tracks for let's say... adding a noise reduction plugin on a dialog clip etc. And then sound effects are applied to other tracks while viewing the video track?

So in a nut shell, there is only very basic audio cuts until the entire movie is edited from start to finish. Which means the film editor is listening to raw video audio for reference purposes to perform good edits, all fine tuning and enhancements to sound are completed by the sound engineer. Does this sound accurate?
Click to expand...

That sounds exactly correct to me.
The sound engineer would also insert all the foley, ADR, music, etc... into the timeline. when the audio mix is complete he outputs the audio to an uncompressed (I think) audio file like a WAV that starts at 00:00:00 on the timeline, so the editor can then replace his audio with the final audio and render out the finished project.
 
clank72 said:
... How do they normally get all the audio merged with the video...
Click to expand...
Maybe not what your asking with this but that's what the slate is for (or the syncing part at least), when they clap the clapoard it provides a visual reference (when the clapboard hits) and an audion reference (the sound of the clapboard hitting) you bring the footage and the audio into your NLE and sync at that point. You can also sync audio to audio, meaning the clean sound from the seperate device to the camera audio or "scratch track". Scratch track is tecghnically when the Sond Mixer runs sound from his set up to the camera for a "scratch track", but it's also (less correctly though widely) used just to mean the crappy sound that the camera natively records. So you can sync up the sound by clapping loudly on or off camera then you sync the audio spikes in the wave forms.

Alternatively to doing this yourself manually in post, programs like Plural Eyes, will analyze the waveforms and sync them for you.

All of the above methods still sometimes take eyeballing and hand shifting due to things like phase and delay.
 
clank72 said:
So the OMF is a data file which then contains the position of the audio clips along the timeline, then pulls in the "linked" sound clips from a nearby folder placing them into the timeline for editing. Those clips can then be moved to different tracks for let's say... adding a noise reduction plugin on a dialog clip etc. And then sound effects are applied to other tracks while viewing the video track?

So in a nut shell, there is only very basic audio cuts until the entire movie is edited from start to finish. Which means the film editor is listening to raw video audio for reference purposes to perform good edits, all fine tuning and enhancements to sound are completed by the sound engineer. Does this sound accurate?
Click to expand...

This is all pretty true on one hand, but on the other, I've only ever had an OMF workflow come into play about twice in a dozen films. Lots can be done right in the NLE and in the indie world it usually is. Reverb, EQ, infinite tracks, automation, even Noise reduction via third party plugins or via software bundled with the NLE like Sountrack for FCP or Sound Booth for Premiere.

Also -- just using scratch track or rough dialogue with no sweetening to edit by can make it hard to judge performance, etc. I'm more extreme than most, but I usually cut below 100hz and bring out consonants as needed and drop off any hiss as I edit. A lot of this is analogous to temp VFX for the editor. It can be hard to get a sense if something works or not without roughing in some elements besides just where you cut. You may also have ambient noise differences, a truck sound in the background of audio from person A's close-up and not person B' close-up, so that has to be artfully faded to not make the cut jarring.

Audio does a lot to sell cuts -- this is obvious with motion graphics -- swishes and such make the camera movement make sense, but also make cuts work sometimes. More subtely, in editing a scene, the feeling of aural continuity can make a cut seem more seemless, so you may do some more tweaking, mixing in just the edit to make sure the cut sells or works. Big studios have sound and visual FX people that work in the editing department doing temp stuff to make sure the cut works / communicate well with the VFX dept and director etc. The little guys will do their best wearing a couple of hats.

Yes in the real pro world, all the mixing will be delegated. When you mix a major picture like Lord of the Rings, one group mixes all the foley, one group mixes all the dialogue, one mixes the scrore, (and even more), then they all meet and try to mix their submixes together in in one master mix, because he submix may have hundreds of tracks. On one of the Lord of the Rings discs there's a great behind the scenes features where they show the submixes being mixed together and another feature where you can toggle between the submixes via your remote. So you can hear just body movements foley, just weapons foley, sweetened dialogue only, set dialogue only, score only, etc.
 
Here's the LOTR Stuff (I pretty sure)
Part 3 shows putting the final mix (all the submixes together)
The last 10 seconds or so of Part 2 shows the Foley Mix isolated with picture.

LOTR2: SOUND DESIGN 2

LOTR2: SOUND DESIGN 2

LOTR2: SOUND DESIGN 3
 
I've made 4 shorts, the biggest with a budget of only $6K, and every one of them had the audio given to a person whose sole focus is audio to do the audio mix. With notes from me (the director) of course. It generally takes less revisions than it does with the video editor where we might go through 10 to 15 cuts as we work our way to the final product. With audio He can usually do a mix, I give one set of notes, and we're done.
 
All the programmes and films I have worked on for the past 20 years have involved OMF importing or on older projects re-conforming the audio from an EDL (edit decision) list.

In the old days it was most common to re-conform the audio as sep sound from DAT or even from the video or quarter inch master tapes and the audio was just used as guide in the edit suite, I mainly used an AMS Neve AudioFile in he 90's and up to seven years ago to do this, extra stereo audio, foley and music tracks were then added and the final dub done on AMS Neve digital consoles.

As technology has moved on OMF has become more prevalent as more audio can be handled in the edit suite. If location sound is done direct to the video tape then this is what is loaded into the edit and generally the editor will just give the dialogue via OMF this way.

Sep sound needs to be syncronised first if it is going to be put into the edit suite, this may take the form of an off-line type format in the case of film work but with digital cinema the pictures and sound can be put together before any post work or left separate if the same timecode is used for picture and sound. it may also be that all the digital camera has is a guide audio track for editing purposes.

Personally as I edit picture and sound these days I do a lot of my audio track laying in FCP and then export an OMF for final dubbing that then goes into pro tools, it may be that I add all my stereo effects and music in FCP too but depending on the project it may also be that the edit only handles the dialogue in a raw form and sep sound is re-conformed straight into pro tools. They may be instances where audio is track layed in the picture edit so that cuts can be done to music or sound effects but this may only be for guide with the full production sound being added in the audio track lay or dub.

Larger projects may involve a lot more with dialogue, effects and music editors all working alongside the picture edit, ADR may also come into the equation as does foley etc and they all get merged together for the sound dub via pro tools or sometimes via an AMS Neve cinefile playback machine.
 
There are whole books and courses of study that cover this, and to get into the nitty gritty takes way more time than can be covered in one post here. Each shoot is going to require its own workflow, and that workflow needs to be developed based on the knowledge of the capabilities of the production cameras, production audio recorders, picture editorial, and sound editorial. I've developed many workflows for all kinds of productions, and no two are exactly the same.
 
Hi Dave,

Thanks, I agree, There are so many scenarios. if anyone knows of any books off hand that would be helpful. Although it sounds pretty self explanatory bridging the gap. Originally thought timecode played a big part. It's hard to believe people are still clapping away on 300 million dollar films.
 
I would guess that most large budget films now are using TC. Though the last huge budget film I worked on (admittedly a while ago) we used the TC to auto assemble the dialog but then assistants went in and phased it all to the guide tracks, so it was still a lot of hand work.

You may want to read the stickies on post, they are probably one of the best sources of good info and... it's free.

I,ve done post on a lot of low budget films and shorts and the last one I did that didn't use OMF (or AAF) was on analog tape - so quite awhile ago. The exception to that is on a big budget film cutting on real film, hardly ever happens these days. They don't have a NLE to output an OMF so you will be assembling by hand from EDLs and pull sheets. Spielberg cuts on film, but they may mirror the work on a NLE? If not they are bringing in a bunch of extra bodies. Up to the late 90s OMF's were not used much because AVID usually had crow sound quality. Not such an AVID issue (though it was a bit) but the folks digitizing were thinking guide track and so they went for speed of transfer. So we always had to re-digitize off the original tapes. Once producers realized OMF could save a bunch of money and time the audio coming from the pict dept got much better and OMF became the standard way to get it to post sound. Of course when your looking for alt takes etc you still need to bring in and sync other takes, so it's not as great as it sounds.
 
From an editor who learned wthis the hard way. This is not generally accepted as "right" but believe it or not, it's actually much faster, IMHO. I've had producers with film experience stand back in awe as they watch me perfectly sync like this, every time.

I cut the shots I want (in FCP), using the reference audio from the camera and put them in the timeline like I would any other video. When the sequence is right, I lock the video track in the timeline so I don't overwrite it.

Now I find the same phrase in the original audio file from the first clip in the timeline like "These oranges are great!" Then I place my "In" at the beginning of the phrase and the out somewhere around "oran--" and drag it into the timeline and partially overwrite the reference audio from the clip. I then play it and see how out of sync it is with the lips.

Keep in mind, I've only overwritten "These oran--".

And then I use the slider tool to sync the phrase to the lips. Delete the remainder of the ref audio and drag the end of the audio clip to the edit. Before I go onto the next clip, I make sure it is sync'd and then quickly unlock the video track and select the video clip (now with good audio underneath it) and the new (good) audio clip and press "CMD+L" ("Link"). Now FCP treats the good audio clip like it was that video clip's own true son.

Once you have done this with all the clips and your timeline is ready for sweetening, go File>Export>Audio to OMF and give that OMF (short for "Oh My F-, that's a big audio file!") file to your sound engineer to work however he likes.

Important: Learned this part the hard way. Make sure your audio guy understands this: his software thinks in Seconds and fractions of seconds like milliseconds. Your software thinks in "frames". If he moves ANYTHING, he moves EVERYTHING. His job is to sweeten and to add to, not to subtract or tighten in any way.

I know it sounds like a lot of work, but it's actually a great deal of fun. Like completely rediscovering your audio as you go.

I will just preemptively say, knowing others will disagree, that I have NEVER seen Pluraleyes work for anyone and I have NEVER seen someone cut to the clapper and not want to tear their hair out. Editing is supposed to be fun! Enjoy it. Love it. Embrace the work of it and if it doesn't kiss you back, give it to someone else (ha ha!)

Peace, brotha,

Joseph
 
Last edited:
Most modern filming does use timecode but the clapperboard is there for more than just sync reference, it identifies each shot (375 take 2.... etc) and more importantly highlights to everyone on set that we are now in take mode and people need to concentrate on what they are doing.

It does help in syncing though as the clap is a good quick reference with the timecode being there to maintain sync throughout the take and act as a numerical reference, in the old days sound and camera ran to a 50 or 60 Hz pulse derived crystal reference and once the sound was copied to sep mag (a sound tape with sprocket holes) film and sound were then in frame sync with each other, film and sound tape would then have rubber numbers printed on them and that was the equivalent to a timecode number reference for each take.
 
"Most modern filming does use timecode but the clapperboard is there for more than just sync reference, it identifies each shot (375 take 2.... etc) and more importantly highlights to everyone on set that we are now in take mode and people need to concentrate on what they are doing."

Absolutely. I would slate every shot even if I didn't need to for audio.
 
JoeRawlins said:
From an editor who learned wthis the hard way. This is not generally accepted as "right" but believe it or not, it's actually much faster, IMHO. I've had producers with film experience stand back in awe as they watch me perfectly sync like this, every time.

I cut the shots I want (in FCP), using the reference audio from the camera and put them in the timeline like I would any other video. When the sequence is right, I lock the video track in the timeline so I don't overwrite it.

Now I find the same phrase in the original audio file from the first clip in the timeline like "These oranges are great!" Then I place my "In" at the beginning of the phrase and the out somewhere around "oran--" and drag it into the timeline and partially overwrite the reference audio from the clip. I then play it and see how out of sync it is with the lips.

Keep in mind, I've only overwritten "These oran--".

And then I use the slider tool to sync the phrase to the lips. Delete the remainder of the ref audio and drag the end of the audio clip to the edit. Before I go onto the next clip, I make sure it is sync'd and then quickly unlock the video track and select the video clip (now with good audio underneath it) and the new (good) audio clip and press "CMD+L" ("Link"). Now FCP treats the good audio clip like it was that video clip's own true son.

Once you have done this with all the clips and your timeline is ready for sweetening, go File>Export>Audio to OMF and give that OMF (short for "Oh My F-, that's a big audio file!") file to your sound engineer to work however he likes.

Important: Learned this part the hard way. Make sure your audio guy understands this: his software thinks in Seconds and fractions of seconds like milliseconds. Your software thinks in "frames". If he moves ANYTHING, he moves EVERYTHING. His job is to sweeten and to add to, not to subtract or tighten in any way.

I know it sounds like a lot of work, but it's actually a great deal of fun. Like completely rediscovering your audio as you go.

I will just preemptively say, knowing others will disagree, that I have NEVER seen Pluraleyes work for anyone and I have NEVER seen someone cut to the clapper and not want to tear their hair out. Editing is supposed to be fun! Enjoy it. Love it. Embrace the work of it and if it doesn't kiss you back, give it to someone else (ha ha!)

Peace, brotha,

Joseph
Click to expand...

I don't disagree that you've never seen pluraleyes work for anyone... I believe you... but I disagree if you're saying pluraleyes doesn't work, because i use it frequently. maybe some people have had bad experiences with it or not used it properly, but it's never failed me... and saved me a lot of time and frustration.
 
You must log in or register to reply here.
Back
Top