Dual System Audio and Sync Issues

[Andrew Peterson Padilla]

First off, I'd like to thank everyone for making this board such a great resource. I wouldn't be anywhere near close to capable if not for all the stuff I've read on here.

That being said, I need help. I recently fell into the position of directing a documentary for a fairly well known band that will eventually be purchased and aired by a major music network. It will consist of both concert footage interspursed with the documentary footage covering this band on tour (before tour, during, etc.. as well as a few things to make it unique). The budget is $100,000 and the guitarist and I have decided it best to purchase the equipment given future projects we'll have together (videos, projects for other bands, etc...). The biggest variable in this budget is the concert footage - we're going to outsource it and have a 6-8 camera production during one (at most two) of the concerts during the tour. Unfortunately, we're still too early in the planning stages to have recieved quotes for such an undertaking...so I can't be sure how much of the $100,000 actually ends up spent on my portion.

We're both hoping that this will be a film that is capivating in its own right, not just to fans of the band. With that in mind, we want it to be high-quality.

For cameras I've settled on an HVX with a Brevis 35mm stored using either an upcoming P2 Gear w/ 4 16gig P2 cards dumping to externals or 2 firestores doing the same. I'll have a multirigpro and panasonic field monitor to go with it. We'll have an aditional (for some legs of the tour, two additional) HVX's and operators for b-roll.

My main issue now comes with audio. And I'm going crazy over it. The band consists of 6 people, and we also will be profiling the road crew. Now, hooking wireless lavs to each band member is not really something I want to do, both because of logistics and sound quality. So I'm aiming for a boom with good quality hyper and shotgun mics. My problem is I can't figure out whether to go dual system...and how.

Now, I want to go dual system for many reasons: camera still gets a rough track, arguably higher quality, and most importantly no tether between me and the boom operator (I want to be mobile). Problem is, I can not for the life of me find resources explaining a decent workflow. HVX timecode out is firewire only, so that knocks out my plan of using lockit boxes. That leaves me with the plan of purchasing a SD 702T and setting it and the camera to time-of-day TC for a rough approximation. Then I can use a dumb or smartslate to sync in post. The thing is, I keep reading here and on dvxuser.com people stating that a smart slate isnt neccesary - that you can just slate once on your first take then again every few hours to adjust for drift.

How? Now obviously the first take is easy matching slate...but after that, how are you syncing clips without a slate to match? Is it just noting the TC difference between the audio and video tracks and simply using that knowledge for all the others? For example, syncing the first take with slate and noticing that audio lags video by 15 frames...so in Final Cut (which I'll be using), just matching up subsequent audio and video by TC then bumping the audio forward 15 frames? I understand the system with smart-slating each take...but I really would like to avoid that if possible.

We will have our own van and my team will be 3...possibly 4 people including myself (Me, Boom Op, and B-Roll/Assistant). I plan on doing dailies to save myself time in the editing room later on. We'll have atleast 2 MacBookPros between us so it should be feasible...I just can't figure out the best workflow for syncing.

If anyone can help me figure this out, I would be forever grateful. And if any of this plan sounds broken, or you think you have a better suggestion - please, let me know! I'm just now turning in the first draft of the budget, so there is still plenty of time before purchases are made.

Also, though not audio related - my plan for archiving is to dump to 500gig HDs with 5 set aside as redundant. Once 4 drives fill (with identical backups) I will ship them home (or take them, if that leg of filming is over) where they will be offloaded onto a redundant RAID array. Once I know that has occured, the 5 backup drives will get cleared and used again for the next series of drives with footage. Does this seem logical? Tape backup seems prohibitively expensive and not worth the money.

For reference I find the films I Am Trying to Break Your Heart (the subject of which is Wilco) and Drive Well and Sleep Carefully (the subject of which is Death Cab for Cutie) very well done. They are definitely a bar I'm setting myself to.

[Dean Sensui]

I haven't done this but have thought about it a lot.

I would use time-of-day TC to provide references to video and audio media. For actual sync I would suggest slating all the shots.

Stick an AT 4051 on your camera. You'd be amazed at what that mic can pick up and how good it sounds at distances up to about 6 feet. I was surprised as heck when I shot a string quartet for B-roll and discovered that it sounded amazingly good. I also got two-shots of fishermen on boats chatting it up and it was almost as good as having a decent boom mic on them.

All that audio was incidental as these people are generally set up with wireless lavs. But at times the on-camera mic was selected for the final cut since the lavs might have had clothing rub or just didn't sound as evenly balanced in terms of tone.

As for storage, you might want to transfer immediately to a level-1 RAID. Should one drive fail for whatever reason, you'll have a backup. Otherwise all your efforts have the potential of being destroyed in a single blip at the production location.

Check out Firmtek for options that will work with your MacBook Pro. I have their Express 34 RAID card and it works nicely.

If using a MacBook isn't practical, then consider investing in some P2 Stores. They take a bit of getting used to, but once you settle in on a disciplined transfer process, they're very reliable and extremely portable.

Also, consider using 8-gig or 16-gig cards instead of the 4-gig cards. You'll find that for documentary work you're going to appreciate the increased working time, without having to transfer data from the camera.

[Andrew Peterson Padilla]

Thanks for the input! I'll definitely look into the RAID option but I think I may have not explained my setup clearly. 5 of the external drives would exist solely to serve as backups of the current captured footage. So for every drive of footage there is an identical one serving as backup. Once four of the backups fill, I will ship (or take myself) the four drives of footage home to put on a RAID server. Once I know the transfer went well, I will erase the backups and begin again backing up newly shot footage. The fifth drive serves as an overlap during the process. But if the RAID setup you're describing is decently portable (will work in a van with an AC car-lighter converter) then that is certainly a better idea.

Also, I guess I wasn't clear, but Im getting (4) 16 gig cards. I agree that 4 gig cards wouldn't cut it! The P2 store is an option...but I kind of like the P2 Gear form what I've read. The ability to offload directly to external hard drives (and power them through bus!) seems to offer the most flexibility.

And I definitely will check out the AT 4051! That sounds excellent...if the on-camera audio can be turned into something usable then that is just icing on the cake. I know our B-Roll camera has a nice shotgun on it, so it'll be great if the main cam can have the same.

As far as slating...I suppose I may have to slate every take. It's just so inconvenient and really a bit imposing for the subjects..it really puts them in front of the lens, you know? But oh well, you do what you have to. I'm still hoping to discover what people are talking about when they say you only have to slate once and then again every few hours for drift.

[Dean Sensui]

Andrew...

I must have had a case of dyslexia and saw "16 4-gig cards".

Ooops!

I hadn't worked with a P2 Gear so I didn't want to endorse that. The P2 Store's biggest issue is the 60-gig drive. Not much capacity considering what one might shoot in a documentary setting. But it's very compact, handles run-and-gun nicely and also features weather-resistant seals. Rain, for example, shouldn't be a problem.

Slating can be as simple as a hand clap or tapping on a hard surface. Even without a slate, it's often fairly easy to sync up with dialogue alone and I've done that a lot to correct for the occasional slipped sync.

Good luck with the project!

[Steve House]

Why limit your options by selecting the HVX? A fine camera to be sure but looking a step or two up the ladder to cameras that have full broadcast capabilities and your issues with lack of dedicated timecode and genlock I/O goes away. I'd suggest you at least take a look at the Sony XDCAM HD offerings before making your final decision. And the Sony optical drive storage is far more cost effective than the (IMHO incredibly expensive) P2 system or external firewire hard drives - at about $30 for a 23GB disk holding at least 45 minutes of HD video at the highest data rate (compared to $900 for a single 16GB P2 card) it becomes cost effective to archive your camera original, no need to dump it off the cards to disk for permanent storage.

I have to disagree partially with Dean's suggestion of an AT4051 as an on-camera mic. It's a good mic, don't get me wrong, but it's a cardioid pattern and that can present problems due both to lack of reach and susceptibility to pickup of camera noise if mounted on the camera. For proper pickup of speech a cardioid needs to be within a very short distance to the speaker, something around 18" maximum. If you're recording music, then there's no problem with much farther distances, and if you're recording ambience, audience reaction on b-roll, etc, and can control the camera and handling noise then recording from the camera position is fine, but I'll go so far as to say it is virtually impossible to record high quality dialog from the camera position with any type of mic on the market unless you're working within a very few feet of the subject - you simply can't violate the laws of physics.

[Dean Sensui]

The HPX-500 also provides TC via BNC connector.

[Steve House]

Originally Posted by Dean Sensui

The HPX-500 also provides TC via BNC connector.

...and genlock also. It too would be a good candidate to choose for this sort of project I think.

I still like the workflow and the economy of the XDCAM's rewritable optical storage compared to Firestore drives or P2 cards transferred to hard drives for long-term storage. By the time you transfer and archive to hard disk for permanent storage all the footage for a major documentary you're going to have dozens of hours of video to contend with and that adds up to a lot of hard drives to buy, manage, index, transport, and store, not to mention the time spent copying the files. For that matter, backup storage as well, especially if you're using inexpensive hard drives for archiving - the footage is too precious to trust to one fragile hard drive. Permanent storage on $30 disc cartridges approximately the same size as a conventional CD in a slimline case that are recorded directly in the camera without any transfer being required seems a lot more manageable and cost effective.

[Jim Boda]

Originally Posted by Andrew Peterson Padilla

...How? Now obviously the first take is easy matching slate...but after that, how are you syncing clips without a slate to match? ...

Dual system syncage...

Send a reference "TONE" before each take to your dual recording system. You can even develop a tone code that will help you identify the shot if you are that organized.

Send a unique tone and line them up visually by using the waveform in your editing system to replace your reference audio with a 24 bit recording.

[Peter Wiley]

I would consider hiring an experienced mixer who has worked on similar projects and who can help you think through the workflow. Such people are not hard to find (http://www.trewaudio.com/frames/ramps.htm).

Failing that talk to staff at either of these places and ask them what you are trying to do and listen to their recommendations:

http://www.trewaudio.com
http://www.locationsound.com

You are making the right decision to shoot double system.

[Seth Bloombaum]

Originally Posted by Andrew Peterson Padilla

...That leaves me with the plan of purchasing a SD 702T and setting it and the camera to time-of-day TC for a rough approximation. Then I can use a dumb or smartslate to sync in post. The thing is, I keep reading here and on dvxuser.com people stating that a smart slate isnt neccesary - that you can just slate once on your first take then again every few hours to adjust for drift.

How? Now obviously the first take is easy matching slate...but after that, how are you syncing clips without a slate to match? Is it just noting the TC difference between the audio and video tracks and simply using that knowledge for all the others?...

I've been a big proponent of TOD timecode into rough sync, note that I've sprinkled in the words "on long takes" in many of my posts.

On doc project interviews and pickup footage a smart slate is an excellent idea. Not that you couldn't sync from what untethered TOD will give you, and perhaps you will end up doing some of that, but having a shot of that TC slate at the head of every clip will make things go much, much faster in editing. Plan now to make the post process manageable.

Regarding post usage of TC in general, yes, noting the TC offset, as its called, and applying that offset for later syncs is a common method in TC syncing workflow. Many good NLEs will give you keyboard shortcuts allowing quick precision in adding or subtracting offsets.

Originally Posted by Jim Boda

Dual system syncage...

Send a reference "TONE" before each take to your dual recording system. You can even develop a tone code that will help you identify the shot if you are that organized.

Send a unique tone and line them up visually by using the waveform in your editing system to replace your reference audio with a 24 bit recording.

This is a method that can help, especially because it applies equally to devices that don't record timecode. We used to use a "bloop" box with pre-timecode nagras. Clip the box to the shoulder strap, roll film, roll tape, up to speed, point the cam at the audio op, who hits the bloop button. A tone is laid on tape, simultaneously a light bulb comes on on the box.

The (non-timecode) problem with this sort of method is keeping takes sorted out (I have some painful memories of this!). We'd hit the button twice for take 2, etc. Hold up some fingers in front of the camera and say "scene 32 shot 4 take 2" into the mic. By the time you've done all this you're back on a dumb slate. Regardless, editors spent a *lot* of time syncing this stuff.

The reason that you might still care about these ancient methods is that you may well end up with non-timecode multitrack audio recordings of your concert footage. With 6+ cameras, someone's going to blow their timecode setup. Shooting P2 cards for an hours-long concert would be particularly painful, you want to eliminate camera starts/stops as much as possible.

Originally Posted by Peter Wiley

I would consider hiring an experienced mixer who has worked on similar projects and who can help you think through the workflow. Such people are not hard to find... You are making the right decision to shoot double system.

Hear, hear!

[Martin Saxer]

Just some options and thoughts:

- why not use an ambient lockit box and record TC to one of your audio channels on the HVX? Jam sync every 4 hours or so and you should be fine. If you edit on Avid you'll have the option to read the TC from one of the audio channels and sync all footage at once. On FCP there is no such easy solution at the moment but a guy in the Netherlands is working on a tool to do it (www.VideoToolShed.com).

- or try this: Time-of-day TC on 702T and HVX and just shoot the running TC on the 702T's display once in a while in order to keep drift under control.

- Low Cost with a no-TC file based audio recorder (e.g. 702): Use Rec run on the camera and shoot the running counter on the audio recorder at the beginning or end of each take (just as you would slate). It's not TC but 10th of seconds and you have the moment when the counter jumps. I then do the following in post: whenever I encounter a running counter on my footage (with the file name of the audio file) I use a little AppleScript that calculates the offset of my rec-run TC from the Camera against the filmed counter of the audio recorder (= the rec run TC of the camera at the point where the counter of the audio file is 00:00:0). In FCP add a TC track (AUX TC 1) to your video file with the audio TC and sync using AUX TC 1.PM me if you go down this road. In a hurry, sorry.

- and: give a mixer a second thought. Working with a cable between camera and boom/sound person is manageable. And you might find yourself wanting more than 2 channels quite quickly. If you have the money give the Zaxcom wireless camera link a second look. With a mixer and a Zaxcom you'd have all: TC, sound on Cam, a backup on solid state memory in case of RF interference, etc. Its expensive but you could rent it or buy it and sell sell it after your project.

Martin

[Peter Wiley]

I meant also to include a link to "the letter"

http://www.coffeysound.com/pages.php?pageid=32