Help with audio workflow

H

HarryR

New member
I'm about to get working on a short film and am wondering how you guys/girls handle your audio. I'll post my muddled workflow below. I'd be grateful for any help to iron it out.

- Import dialogue tracks and denoise if necessary with iZotope
- Add all video clips and corresponding audio files to Premiere timeline. Do the export, pluraleyes, import thingy.
- Checkerboard audio and video for actors
- Export omf, add timecode and render out low-rez QuickTime video
- Open omf in Pro Tools and import video
- Work on the dialogue - cross fades, room tone, eq, compression, etc.
- Record, add, and sync, Foley, ambiance, atmosphere tracks
- Do whatever I need to do to make the mix sound right
- Add a brickwall limiter on master bus, bounce to disk
- Import mix into premiere, sync, export final video

I don't really know at all if this is a good workflow and I have trouble understanding how to ensure sync using timecode. Should I use stems in Pro Tools? Am I missing some crucial steps? I have done a lot of searching and reading and this is the best I could come up with so far. Any help would be awesome!

Cheers,
Harry
 
I would personally not use any kind of denoiser or processing of the dialogue at the video editing, so that the OMF imported into pro tools contains the raw dialogue tracks. The dialogue editing is where I would do any denoising, and only after I've made a copy of the raw clip on a new track (muted or inactive), so you always can go back to the original clip.

I don't know much about the workflow when importing and editing video, but depending on how much time you've got to do the audio post, you can prioritize certain things. Dialogue almost always comes first, but what you do next depends on the kind of project and time allowance. Some prefer to do Foley, then fx or ambience. Sometimes you can save some time by doing certain ambiences before any Foley and fx (or even dialogue editing), because it might be loud enough to cover more subtle defects and fx, so you only have to fix what is clearly audible. Some scenes might be dominated by music, so you don't have to spend much time doing ambience or soft fx and Foley on them.

Lastly, what medium will this film be shown on? Certain mediums have standards regarding sound levels (both maximum and average), and that will have to be taken in consideration when mixing and exporting it.
 
Last edited:
Well, you can't "assure sync with TC". Your in a file based workflow and really TC is more of a locational thing. TC ruled when you were syncing tape to whatever. It's useful as a reference as in I want an explosion at xx:xx:xx:xx, since that should be the same place for everybody but other wise it's not a lot of use in low budget filmmaking.

If you do use it for post remember that it should start on the hour of the reel number. Since you are probably only doing a single "reel" your First Frame Of Action (FFOA) should be at 01:00:00:00. Your head pop is at 00:59:58:00.

IF you have TC on your recorder and camera synchronized (or at least manually set to the same TOD) you can use TC to roughly sync takes. But to use it to really sync you would need all your equipment locked to trilevel sync and if your posting here I doubt that that is an option.

I'm guessing your a one man band since some of the above is the picture editors job, but. DON'T use any Noise Reduction till your in sound post. It is the first "destructive proses" you want to do but you won't really know, in most cases, how much you need to do till it's in context and you absolutely want to be doing it on the best sound setup you have, not in e picture edit suite.

For sound post you only need the final cut picture. You don't nec. need the picture editor to checkerboard the audio, and you may not want to have them do it because you may well want to group them in another way. You do want "handles" on the OMF. Not sure what PP's default is but you want between 4 and 10 seconds if you can get it.

"add TC" I assume means add a TC burn to the video. Some really like that some don't care. When we were using tape I definitely wanted it and would add one if picture didn't. It is a bit useful if the director is going to be making notes using a video player, since the numbers should match, or if you are sending to a composer with places you want music. But from a pure sound post position it's not a big plus and can be a problem depending on where and how the numbers get placed since they will block some of the action.

The video doesn't need to be particularly low res. If you set the Codec to render a key frame for every frame (default is something like one every twenty four) you can use HD resolution h.264 video with ProTools with out problems.

You left out FX or are confusing Foley with FX. Foley is a very specific, and you are probably not doing much or any real Foley. If you record someone making sounds in sync with live playback of the video it's Foley, if you are cutting in sounds that were acquired any other way it's FX.

Also someplace in there you are probably going to be adding some music.

The brickwall is good, but keep in mind what you are delivering. If it's a "film" your max is someplace around -6 - -3 dBfs. It's good to have but you don't want to be hitting the limiter very much. You can also use one on the dialog sub. A lot of mixers will use a limiter rather than a compressor. Similar result but the limiter won't raise the noise floor on it's own, so you have a bit more control. A compressor pushes the top and bottom toward the center and the limiter only pushes the top. Both are used but the most common for dialog is a multi band limiter which also can effectively do a bit of noise reduction if set correctly.

Do you need stems? Maybe. You need to go from the end backward to really know. Are you preparing for a foreign language dub? If not probably you don't need stems. I generally do a kind of "virtual stems" in that I buss Dialog, Music, AFX, BFX, Foley, ADR, etc. to separate submixes.

The reasons are that it's handy to turn off everything else when working on a single "food group" as in dialog. Also various groups often have slightly different things you want to do to them. For instance you might have a brickwall limiter set to -3 on the hard FX, like explosions, but that is too high for dialog so the dialog submix might have a brickwall at -6. Dialog might benefit from an over all EQ adjustment where you don't want that on FX. Music may have a number of things that you don't want on the over all mix.

I as a rule don't create "stem" files, but I can if needed.

I also generally set up a PrintMaster track and record the mix to that rather than "bounce to disk".

The reason for that is three fold. First you can listen to picture while you "bounce" and stp if there is a problem. Second is you can go back into the PrintMaster and destructively record fixes. On a short not that big a deal but on a feature the difference in punching in a couple of short fixes over rerecording the entire file is very significant.
Third is a combo. When you catch that problem (in first) you can punch in (as in second) and not have to start from the beginning.

In a big mix there are just a lot of things to miss, especially if you are on your own or tired. Like a track you muted that is off the screen. So the reality of having to do fixes is that you will almost always have to fix something either because of your error or a directorial change of mind. If every change is a hour and a half recording you are going to get tense fast. If that missed door close 01:32:00:03 is only a five minute fix you will be MUCH more relaxed.
 
I also generally set up a PrintMaster track and record the mix to that rather than "bounce to disk".
Click to expand...
I've heard several people do this, and it seems logical. So do you just name the track the same thing as you would the bounce, consolidate it after all "fixes" are done, and copy the consolidated file from the projects "audio" folder?
 
I don't do much dialogue clean up (noise removal and the like) during editorial. Every time I've done that I've had re-recording mixers get mad at me. They like to have control over that, and that's fine with me because they usually have better tools for dialogue clean up on the mix stage than I do on my editing rig anyway. I've had re-recording mixers tell me to do different things with the dialogue. Some want each character on their own track. Some want it cut according to the type of shot. I prefer to cut it per the shot (close, mid, wide)..makes for easier mixing later on.
 
Jema said:
I've heard several people do this, and it seems logical. So do you just name the track the same thing as you would the bounce, consolidate it after all "fixes" are done, and copy the consolidated file from the projects "audio" folder?
Click to expand...

I don't consolidate because I destructively punch in. I generally just name the track with the project (or Printmaster) and then rename the region to what I want the file name to be. Then I select the region in the region list and "export region as file". That way I can generate a file type that picture wants. I work at 48/24 WAV but FCP for instance likes AIFF better. I can also generate an MP3 for listening if needed. Not great quality but it can be emailed or downloaded quickly so OK sometimes when you need to send a temp.
 
davefisk said:
I don't do much dialogue clean up (noise removal and the like) during editorial. Every time I've done that I've had re-recording mixers get mad at me. They like to have control over that, and that's fine with me because they usually have better tools for dialogue clean up on the mix stage than I do on my editing rig anyway. I've had re-recording mixers tell me to do different things with the dialogue. Some want each character on their own track. Some want it cut according to the type of shot. I prefer to cut it per the shot (close, mid, wide)..makes for easier mixing later on.
Click to expand...


All true. I was assuming that he was doing it all himself. If you are sending it to a mix stage then you should check with them on how much if any processing they want you to do on dialog.

And as I think someone else mentioned. Always duplicate and disable the original OMF import tracks so you can go back if you trash something. Very big time saver. The other option is to reimport the OMF to get the bit you trashed and toss everything else.
 
I don't consolidate because I destructively punch in. I generally just name the track with the project (or Printmaster) and then rename the region to what I want the file name to be. Then I select the region in the region list and "export region as file". That way I can generate a file type that picture wants. I work at 48/24 WAV but FCP for instance likes AIFF better. I can also generate an MP3 for listening if needed. Not great quality but it can be emailed or downloaded quickly so OK sometimes when you need to send a temp.
Click to expand...
Alright, I'll try this out. Thank you for the explanation.
 
Hi Noiz and Jema,

Thanks for chiming in with some good info. Let me clear a few things up first off. I have a tiny crew that I work with: a DP and a boom op. We are hoping to make a film or two to submit to festivals. I do all the post work myself. As regards standards I calibrate my monitors to 85db and mix to taste but generally I try to have average levels at -12dbfs and peaks around -6dbfs. Is this a sound approach?

I have read about plugin delay and that noise reduction plugins are the real culprits so do my (light) de-noising (if necessary) first in iZotope. I know this is destructive so I have the raw files backed up. I will do as suggested though, from now on, and wait till the dialogue is seated in the mix before doing any de-noising.

I was referring to Foley. I don't have much cause for hard FX in the films I make. I have a small Foley stage and I actually do a lot of it. I take what I can from production sound but a lot I add because I am extremely specific about the way things should sound. I just set up clips on a monitor or ipad for reference. And yes, I forgot music.

Seems I can dispense with the TC burn in.

For some reason I thought I needed a blackmagic card or the like to get HD playback in Pro Tools? Can you explain in a little more detail how I can do this without such a card? I have a lot to learn about Pro Tools.

I really like the idea of recording the mix to a printmaster track. Being able to punch in and make adjustments rather than having to bounce again and again sounds very good indeed - so thanks!!

Good tip with duplicating the OMF - something I haven't done before. It's fair to say I'm a novice at all this.

Brickwall on the master is purely a precautionary measure. Realistically I don't want to hit it, ever. I hadn't thought about putting a limiter on the DX bus but will try it out.

It seems I am already using "virtual stems" so that is good. I have a read a few different approaches regarding stems so I just wanted to get some clarification.

I know I have more questions but I'm not sure what they are. Sound and video are so different. Sound is the scary, deep-end of filmmaking where it is harder to define the semi-abstract rules that help to make good audio. My video workflow is a cinch compared to my audio workflow. It just feels like so many things can go wrong and I always feel colossally unorganized, and that I am doing everything backwards. I wish I could learn hands on with a professional. Books and practice are what I have so they must do.

I'm sure I will run into some problems soon that I will need advice on. I feel extremely grateful to be able to lean on you good people here for that advice.

Sincere thanks,
Harry
 
Last edited:
I don't use NR as a track plugin, I use it as an Audio Suite plugin, so really there is not a delay issue.

Good on you for doing Foley, that should help a lot with the over all quality. I would recommend you tag and catalog the sounds though and add them to your library for future use.

I also think you would benefit from thinking more about FX. Your not going to be able to really build backgrounds with Foley, at least not efficiently. And unless you have a really big room with a ton of props there are a lot of things you are not going to be able to do well with Foley. But how big a deal the above is depends a lot on the projects so?

Picture has a LOT to do with story and sound has a LOT to do with the "world" the story takes place in. Picture editing is largely subtractive where sound is largely additive. Sound is much closer to doing visual FX where you are adding lots of layers to create a single effect.

Where are you located, there are probably folks in your area that you could pick up tips from.
 
I forgot about the Audio Suite option. I have been doing noise reduction backwards it seems. I just watched a few clips of John Purcell's lecture included a long one on noise reduction that was very interesting. I'm making my way through his book too - what a wealth of information!

Good idea about the little Foley library.

I sometimes think more about sound than picture so I'm not skipping over anything. Probably just a misuse of terminology by me.

I'm located in North Queensland, Australia, where there is a dearth of serious filmmakers. I haven't met one yet.

Thanks again for your help,
Harry
 
You must log in or register to reply here.
Back
Top