Kudos to you dialog editors!

[dcwave]

I am working on a low budget 90 minute indy film, originally just to mix the final stems. The dialog editor who is also doing Foley and FX was running behind so the producers asked if I would be willing to take on the dialog in order to make the deadline. Silly me; I said yes. Most times when I work on film projects I am doing FX replacement and sound design, minor edits to dialog stems and the final mix or composing - enjoyable work.

All I have to say is that you guys who do dialog full time must be miracle workers!
I want to kill myself or the actors or the directors or someone or something.

The biggest challenge - inconsistent background noise/hiss between actors speaking. So much hiss that any amount of NR makes too many artifacts. The other pain has been matching ADR with outdoor scenes, again with inconsistent background noise. Oh and no room tone unless I cut between phrases and past some stuff together.

Any tips on how to deal with the background noise and hiss? And blending ADR into outdoor shots?
Thanks

[gsilbers]

lucky you that u had ADR! i worked on 6 fox TV shows with daily showing that didnt use ADR! it was sound restoration! plus the location was awfull and i guess there where no cuts when airplane or trucks passed near the set! still, they sounded great, or at least ok. not to mentioned beach scenes! "please people, can we have more Ocean sound on that scene... i can still hear the actor!

long fades of course between diferent sound sources (mics). once small trick is to cut a piece of noise floor and duplicate it and reverse it and reapet several times so when played back sounds like one long one. (if they dont give u that in the start/stop of the action). then use those to help to do longer fades. the longer the fade the better as u wont hear those transitions.

on that note. if the scene is a guy talking and the other actor is just responding short and it has another ambience...just use that extra long pieace of noise and added to another track all through out the scene. if not the same noise as the other something similar, and save them to one folder for easy access.

another simple handy thing is pro tools 7.2 will let you stretch out fades instead of -selecting region-increase size then fade it. minor stuff bu wil save u time.


another thing i did while cutting DIA was adding a compresor on the tracks and an L1 on the master and using a hipass filter set on 200hz or so. this way i could easy hear what the engineers (closer to final result) will listen. sometimes the ambience sounded difrent or there where camera hums on difrent freq i didnt hear before.

BTW, we didnt use an NR plugins! as it creates artifacts that took too much attention. to itself.
we used the waves C4. its a miracle worker for those things. it took me time to learn how the engineers used it but its worth it. we have it so its parameters are set on the Icon for those wierd transitions.

[dcwave]

How is C4 used for dialog?

Coming from the music production side of things, a MB compressor is often looked upon as a bad thing.

[Jazzpunk]

Originally Posted by dcwave

How is C4 used for dialog?

Coming from the music production side of things, a MB compressor is often looked upon as a bad thing.

It can be used to selectively compress certain frequency ranges thus helping to squash the offending frequencies in noisy audio clips.

The C4 is also great for bringing VO and dialogue levels up when trying to pass restrictive LM100 specs. An extremely useful tip that I learned right here on these boards.

[brucie]

"once small trick is to cut a piece of noise floor and duplicate it and reverse it and reapet several times so when played back sounds like one long one."

Wow I thought I was the only person to do this and get away with it thumbsup

Phew!!!

[Branko]

Hi,

C4 can be used not only as a multiband compressor, but as a multiband expander as well, and that is exactly why it can be very useful in production sound pre-mixing.
With some careful fine-tuning, one can get a very good emulation of Dolby's old Orange Box (cat 43) and a very efficient de-hisser. Just make the range control positive and gain control negative and choose frequency ranges for 4 bands. It takes some time to get all the parameters right. The fact that the range is limited to +/- 18 dB helps to avoid overprocessing.
Branko

[dcwave]

Very useful tips! Thanks, the trick about reversing the noise floor is pretty good and help salvage a number of spots.

I will for sure being looking into the C4.

How do you guys (or gals) deal with off-axis production dialogue? Just EQ automation, especially if they turn away?

[doug_hti]

Originally Posted by Branko

Hi,

C4 can be used not only as a multiband compressor, but as a multiband expander as well, and that is exactly why it can be very useful in production sound pre-mixing.
With some careful fine-tuning, one can get a very good emulation of Dolby's old Orange Box (cat 43) and a very efficient de-hisser. Just make the range control positive and gain control negative and choose frequency ranges for 4 bands. It takes some time to get all the parameters right. The fact that the range is limited to +/- 18 dB helps to avoid overprocessing.
Branko

Yes, a very effective way dealing with NR. I do this on troubled sources. However constant fairly broadband hisses should be able to be knocked out in a pretty good way with actual NR tools...

all in all, I find that it's almost never about one tool, but a combination all doing a little bit of work.

[gsilbers]

Originally Posted by doug_hti

all in all, I find that it's almost never about one tool, but a combination all doing a little bit of work.

yep.


just to add more even though is not for dialog editors, is the EQIII in pt which u automate for troubled passes with noth EQ. with extreme Q and gain u sweep the freq until u find those hums and camera noises and then cut them. works great if the mic on the actor has hi frq hum but the boom mic doesnt. you do "copy plugin settings, and then paiste plugin settings comands" and "write to next" to cover that section of troubled DIA (all very fast with an Icon!)
but before the dialog editor will have to guess where and if the engineer will be able to do so or not. (if theres no $$$ for ADR)


also for Dialog editrs, which is obvious. is to do cuts for phone scenes appropietly.
make 2 long noise passes from each scene so it can be place whenever the other actor is on the phone but the scene is the other actor listening. and its about a frame before and after of fade between cuts, not start the fade when the scene changes (but a frame before) ...

there is new setting in pro tool 7.3 (i think) which lets u choose en the prefs to NOT use avid fades. which will save tons of time be not having to erase the avid fades and then do yours.
also in 7.3 which i mentioned before is the bility to grab and hold a fade and drag in or out. VERY VERY important to make editing faster.

also when u mix you can talk to the engineer of how to labels the tracks so he can import it into his mix template. some or more picky about this than others.
also you can name a track or the last track is where you put all the "futses" like phone speaker, small speakers, announcers from radio or tV you know stuff it will be EQ like its coming from a phone. and rememebr to add noise underneath it.

also divide the track appropielty, like if the scene has 3 actors you can separate the track by actor but i rather split them up by the mic used. this way the mixer (sometime me can EQ out offensive stuff no matter whos talking. it can vary but make it consistent. this way the mixerengineer can automate by selecting the track of the whole scene, notch filter/c4/comp/etc and copy this automation to the whole scene and start from there instead of doing little parts at a time.

[kailpot]

Great thread..any more tips/thoughts would be appreciated! I am especially interested how much of the eq/Noise reduction gets done by the Dialogue Editor and how much by the mixer in a major production?

[dcwave]

Here's another question:

How do you remove "film stutter"? Its like the sound that a film real would have if it wasn't in the gear and kept playing. I have several spots like that and I have tried noise redu but that just makes it sound like a low thud - if I try ti EQ it with a tight notch it removes too much of the male voice's natural timbre.

No chance for an overdub either.

[gsilbers]

Originally Posted by kailpot

Great thread..any more tips/thoughts would be appreciated! I am especially interested how much of the eq/Noise reduction gets done by the Dialogue Editor and how much by the mixer in a major production?

dont know about "major production". in TV or if u have abudget you will just print out a ADR sheet with the lines to be replaced. if its somethingin production then its foley or sfxs.
usually NR and that sort of stuff is done by the engineers. if u both are the same then why not. or if u know the engineer well then its cool. but when I was overwhelmed by TV shows we hired extra dialog editors and the engineers where very particular con how they (we) wanted stuff. some could handle anything that came in. others liked their tracks arranged certain ways, labeled like this, sorted like that etc. and also not mixed or anything.



noise reduction is used more for old recordings and stuff like that. for dialog/prod the c4 kicks ass. then EQ/notch filter out static noises and/or use x-hum to remoove 60hz and hums.
when i started mixing i used wavearts restoration bundle and waves but 2 things occured; one was that it sounded reverby/thin/artifacty and second its delay compennsation on pt was humongoues and unuasble at times.
i lefted simple and besides long fades and smooth transitions, automate EQ for each scene to filter out noises and difrent mics sounded similar and automate the C4 if the scene was too noisy.
of course there difererent ways to do the same thing.

[gsilbers]

Originally Posted by dcwave

Here's another question:

How do you remove "film stutter"? Its like the sound that a film real would have if it wasn't in the gear and kept playing. I have several spots like that and I have tried noise redu but that just makes it sound like a low thud - if I try ti EQ it with a tight notch it removes too much of the male voice's natural timbre.

No chance for an overdub either.

interesting. never heard that.

so its digital hit or as if it was done with film reel and the audio record it?

[Jazzpunk]

Originally Posted by gsilbers

interesting. never heard that.

so its digital hit or as if it was done with film reel and the audio record it?

Sounds like they skimped on a boom guy and tried to rely on the camera mic. Funny how much those pick up.

[gsilbers]

when he mention no budget "digital cam" poped into my mind. thus, film reel noise seems oout of the equation. but who knows.

ill ask the sound recordists

[dcwave]

Yeah - it was a DV cam. I am not sure what the sound was, the best way to describe it is like a film trying to run with a gear not aligned so its a constant stutter sound. The wav form is funny too in the parts with no speaking - no negative image, only postive.

The best I could do was use NR to get it at least low enough that you could hear the voices more clearly. The producer thinks the mic cable may have been shorting out or not plugged in solidly.

[dcwave]

Another question pops into my mind while I am editing:

How fast do you work? I am not sure I am working fast enough. They have given me five 18 minute cuts and it's taking me about 4-6 hours for each one.

[gsilbers]

Originally Posted by dcwave

Another question pops into my mind while I am editing:

How fast do you work? I am not sure I am working fast enough. They have given me five 18 minute cuts and it's taking me about 4-6 hours for each one.

crap. i forgot now, but something like 8 hours per 30 minutos of film for big production. damn i forgot. nevermind.

the old cats DIa editors wher doing 8 hours for a 45 min tv show.

but it really depends on the quality f recording. sometimes i do an hour show in 3-4 hours, other times is like 12.

if the scene is outside with lots of actors and they forgot to replace the bateries on the wireless mics then its going to take a heck of a long time.

it varies a lot.

my friends records for a top reality show and other TV shows and he wont record if there is a plane, truck or the like passing by. he will get into fights with the directr cause of it. cause he answered to the producers becuase how much are u going to direct for reality show, really? "camera 1 follow the half naked girl", "can we make it more "real" in the interviews! please people its a reality show act natural!" hahaha.
I have astory about one of those "reality shows" with tori spelling that went somethign like that... haah. oh boy our kids are going to laugh at our period.
anyway
most directors dont know this kinda of stuff. same ones they want music prced at almost next to nothing and dont want to pay for audio post. why not just do the audioin avid or FCP it supports audi oplugins. !

[Branko]

...it's taking me about 4-6 hours for each one.

What do you exactly do in 4 - 6 hours? Dialog editing, Dialog cleanup, Sound editing, Mixing...?

[dcwave]

Originally Posted by Branko

What do you exactly do in 4 - 6 hours? Dialog editing, Dialog cleanup, Sound editing, Mixing...?

Clean up, editing, eq matching ADR to the scene (which was bounced into a "dirty" production bed by someone else), trying to grab room tone from heads and tails, trying to eq and match room tone that changes every time the camera changes from one person to the other. Playing back the scenes with different eq automation, correcting the automation. Then there's the outside scenes.

Thankfully I finished this project that spawned this thread. The producers are pleased with the results and understand that there were a few rough spots that nothing more can fix them and they can live with that.

[starcrash13]

I don't know about how it works in televsion, but in my experience in film 4-6 hours to cut 18 minutes of dialogue is not typical at all. That would mean cutting an entire feature film in 2-4 days. And then premixing (ADR matching, EQ, noise reduction) on top of that?

Hmmm.tutt

It's easy for these types of "how fast are you" discussions to turn into a dfegad contest but keep in mind that part of being an experienced editor is being able to budget time and assess how long something will take. Also, since we mostly get paid by the hour, it's not a sprint to the finish to save the client money. It's about doing the job properly in the amount of time that's budgeted. There is always going to be major pressure from the post super or producers to do it faster. That's their job. They'll always tell you that there is no money left for post sound. That's just standard and usually an experienced supervisor knows how to manage that type of pressure. And, even if you break your neck to finish in record time, the client will immediately expect that turnaround every single time from that point on.

I don't specialize in dialogue editing, but I've done a few projects and I work with dialogue editors that are at the top of the heap in the industry. I see them typically spend 2-3 full days on a reel of dialogue and that's just editing, no ADR, no major EQ, and usually no NR. The reason editors end up working on major features for several weeks or months is because of all the picture changes that happen up until the last minute.

[dcwave]

Originally Posted by starcrash13

... keep in mind that part of being an experienced editor is being able to budget time and assess how long something will take. Also, since we mostly get paid by the hour, it's not a sprint to the finish to save the client money. It's about doing the job properly in the amount of time that's budgeted.

Hence the reason for my question. I have no real reference to know if my pace is good, bad, or ugly - I just know the deadline and was trying to move at a pace that did a decent enough job to meet the deadline. This thread, as well as many google searches, has helped me have a better understanding of what "doing the job properly" is. Having had a couple of days to step back and do a postmortem I realized that there were things that were not done properly - both on my part and the Audio Director's part. But knowing what is proper is half the battle so that I don't make the same mistakes in the future.

Something I learned for sure is that more time is going to be required to do a proper job. I felt rushed

[petessound@hotm]

for one reel of film (10-20mins) i will spend anything from 1-4 hours ediing the dialog. it all depends on the recording quality. and aslo if there are any song in the film

its usually around 2 hours tho and try to factor in any missing lines/distorted lines etc that may need to be rerecorded and supplied.

dont know if that helps

[Branko]

for one reel of film ... i will spend anything from 1-4 hours ediing the dialog

How wide is your editing? How many channels do you deliver to the dub stage? I have to admit it seems too fast for what I'm used to...

[gsilbers]

again. its depends on the scene(s). if it was recorded in a quiet room and there are only 2-3 actors and cameras and mics, then itll be faster than an outdoor scene with more variables.

after a few films/projects and if u see mixers mix the film then u get a better idea of how fast or slow and what it takes. also if u see other editors do it.

sorry for the "it all depends" answers. usually the faster the better so learning little tricks and keycomands help a lot if u doing repetetive stuff. when PT7.3 came out the prefernce of "not preserve avid fades" saved me about an hour of edting per show.
also the "stretch fade" instead of extending the region and then fade. little stuff like that helped. also learnign how to do those noise floor regions fast helped. well its kinda tral and error. if u get noisefloor from the begingin or end then cool, you duplicate it/reverse it a couple of times and then save it if theres another similar scene. but if there is no noisefloor at the tail or start then grabbing parts from inbetween dialod takes way more time to make sound desent.
if u get an OMF with music and effecs (unmixed or mixed) you can also tell if u need to really make too much of an effort or not when the scene will be covered in music and effects. but if its only dialog in the scene for a long while, then you have to be very very detailed.
what helps is to listen a high volumes with a limiter for wierd peaks. and no, im not saying **** up your ears. a compressor on the master track with a strong ratio helps, cause sometimes u hear that two cameras mics sound the same but when u go to the mixing stage its very difrent.

[Branko]

Sorry to insist, but how wide is that editing?
Branko

[smurfyou]

Originally Posted by starcrash13

I don't specialize in dialogue editing, but I've done a few projects and I work with dialogue editors that are at the top of the heap in the industry. I see them typically spend 2-3 full days on a reel of dialogue and that's just editing, no ADR, no major EQ, and usually no NR. The reason editors end up working on major features for several weeks or months is because of all the picture changes that happen up until the last minute.

I would agree with this generalization, I usually end up spending 2-3 days per reel on average (if time/budget allows of course). And to say it one more time, it really depends on each individual project. What the quality is like, if room tone is usable (or if it exists!), how bad the editing is, what the OMF layout is like, etc.....I usually make sure I have at least 1 day per reel if it is a serious project.

I agree with the above comment about arranging tracks by microphone opposed to character. By character appeals to me logically, but practically by mic works out better. Its always best to talk to the mixer anyway.

On a side note, I cut dialogue with cans. I know I'm not the only one I know, but does anyone else on here use headphones? It makes the job a lot more fatiguing on the ears but I can be sure I won't miss anything. I'm a big fan of DT-770s.

[Eric L]

When I first started cutting dx, a 20+ year veteran told me the following:

"10 minutes a day, or you don't stay."

this was feature film dx editing, after auto assembly was done (Post conform days)


Now-a-days, with all the different recorders, meta-data, bwfs, Titan, OMFs, picture changes, virtual katy, edls, change notes....

and no assistant editors (only interns), it is a different game.

Assuming you have done everything else and you have audio in the proper place on the timeline, you should be able to cut 10-15 minutes of program per day depending on the sound quality. Anything less is unacceptable.

Of course, there is much more to DX editing than writing fades. There seems to be a trend now, where the "dx edit" being done is nothing more than splitting out the OMF tracks, removing duplicate regions and slapping on fades.

The art of cutting in fill, removing breaths, pops or clicks is being lost. Too often I have heard the dx editor say, "Can't you use de-click or de-crackle on that?"

A great dx editor is hard to find, but worth every penny they earn when you do find one. They just are not being trained anymore. It's a shame. I have tried to train a few people now, and it just isn't sticking. It seems cutting sfx is more exciting.

[petessound@hotm]

sorry for the late entry...

there will usually be around 8 main charecter tracks 4 various tracks a few wallas and possibly vocals on top of that.

i will usually be concentrating most on sync, lip smacks and external noise removal. when i started it used to take me more like 4 hours per reel but now its down to 2ish for two reasons.

1. i have learned protools a whole lot better (i thort i was the man before!)

2. i know what you can let go. it can be very easy to get sucked into things that frankly people will not hear

[starcrash13]

I agree with the above comment about arranging tracks by microphone opposed to character. By character appeals to me logically, but practically by mic works out better. Its always best to talk to the mixer anyway.

On a side note, I cut dialogue with cans. I know I'm not the only one I know, but does anyone else on here use headphones? It makes the job a lot more fatiguing on the ears but I can be sure I won't miss anything. I'm a big fan of DT-770s.

I split tracks by microphone setup, rather than character also. I also use the Beyer-Dynamic DT 770's. They're great! I used to use 7506's, but now I could never go back to them. Night and day difference. I always use cans for cutting DX.