Syncing audio and camera on set?

Joost

Joost

Well-known member
Hello dear collegues,

Normally I just record my audio on board, and sometimes I sync 2 camera and zoom/PA recording with music registration.
You know the drill, just some interview, setnoise, or PA out recordings.

Now I'm going into a different ball-game. I am producing a indi low-budget drama/commedy show. Script is ready, time for producing and place budgets.

Doing 6 years of one(few)-man band work I know allot by doing it myself (writing, lighting, camera and audio) but still reading in deeper.

Probably we will recrute a propper soundguy (hard to find). But still I want to know the options on workflow myself in this part.
Found out that a good way is to, give all the actors wireless-lav's, boom and hide some dynamic mic(s).

This is giving me a lot of tracks to work with. :(

For a recorder I am aiming on the portable recorder Tascam DR-680 wich can record in 8 tracks and zoom h4n. Or 2x zoom h4n and the onboard audio inputs.
For camera's i'm doubting for:
-3x Blackmagic Pocket
or
-3x FS100
or
-2x 5DII and 1x 5DIII

I edit in FCPX


MAIN QUESTION:
What is the best way to record sync ready. 3x video and like 2x audio files (10 tracks. I would like to avoid plural eye's, because of the lav's it might not sync to good. Plural eye is not alway's functioning to good.
Going to use a clapper ofcource.

I have no clue but my ideas on this:
1-Give all the recorders and camera's the exact same timecode. Can this be done super precise on the mentioned camera's? What is the best way? In edit I then search time-codes together (still a slow process, but precise).

2-Can I start al recorders with a lanc or something. Something that start al device at the same time so I can drag audio and video sync in timeline.

3-Sync on the clapper. Most simple option but very manual/time consuming.

A red-link kind of a solution would be awsome. RED is no option for budget reason.



What are other good options?

Thanks, hope you can think along.


Joost de Jonge
 
None of the cameras you're considering can accept or send time code. The easiest way to do this is not the cheapest way, and it requires a time code slate with master generator, a recorder that has time code (Sound Devices 788T or similar), and cameras that can sync time code to an external device. The slate is the master clock, and all devices are jam-synced to it. The camera may also function as the time code master on some productions.

Without any of that, there are two options. If you can at least get a smart slate and something like the 788T, you'll at least have time code markers wit your audio, and visual time code on the slate for easier (but still manual) alignment. As the most low-tech option, you can do what you originally thought and rough in the time code manually on each device and hope for the best (not exact, and you may experience some drift), relying on a clapper for sync marks.
 
Thank you very very much. Clear to me.

Another possible options pops up in my mind.
What about live recording in FCP(X) on multible tracks at a time?
Can this be done on a video-audio-combi mixer/switcher via thunderbolt/usb 3.0 or such?
The BMCP and FS100 got clean hdmi out so that a start.

Not the best option in the outdoor scenes, but doable in the studio scenes.

Wonder if an iMac can handle that or/and a video-audio switcher can buffer and/or output that.
Or maybe there is a recording device that can do it without computer, straight on ssd's.

I'm goin to explore that. But if someone know it is or isn't their, let me know.

http://www.blackmagicdesign.com/products/hyperdeckstudio/models
like this but for multible tracks

or maybe this? Or is it just for one track recording, and live switching?
http://www.blackmagicdesign.com/products/atem/models/
 
Last edited:
Sure, you could go through a switcher and record an already-switched show. I'm not sure what FCPX's capabilities are considering multiple audio inputs, though it was possible in earlier versions with the right interface (like Matrox, AJA, BlackMagic). But that's asking for troouble in my opinion. Not only are you stuck with switched video (which will be a real pain to fix from the camera's source files later), there are too many other variables and opportunities for the system to crash.

Or, you could look at something like THIS. You'll have to purchase the HD interface cards to use it (no HDMI, either)... and you'll need hard drives. It ain't cheap.

When you get into video switcher/mixer units, you're in the area of needing SDI out from your cameras. Honestly, once you need to sync a multi-cam shoot for editing later, or even to record a multi-cam event switched "as live," you're getting into a process that takes a bit more specialized equipment.

The Panasonic AF100A has SMPTE Time Code i/o and can sync to other devices. Just a thought.
 
Yeah AF100a is perfect for time-code and sdi out, so good option. But for me the most expensive. Got the FS100's and 5D's. BMC pocket is raw and cheap so I would like to spent on thoose.

Life editing is no option indeed.
After looking on internet, I found the FCPX multicam option http://www.kenstone.net/fcp_homepage/fcp_x_multicam_basics_stone.html

It is like the Plural eye's option. But later I thought, if I bundle the lav audio it wil have not so much problem syncing the video. And it's free, giving me more budget for further production (food for cast :).
All kind of things still passing my mind, compremising costs, workflow, speed, quality so thanks for thinking with me Alex.

I got the options. Now I just have to make a choice.

Thx
 
Remember that with current file-based recording techniques, whether audio or video files, time-code can provide a single position reference in the file but does not provide any speed references, thus it's great for quickly aligning the audio to the video but doesn't do anything to prevent so-called "drift." Functionally it is identical to a clapper slate on each shot, only faster and more convenient to use. For the situation described, the easiest and most economical way to use it is, as Alex suggested, to use a smartslate with a timecode-capable recorder for the audio. Slate and recorder are both set for continuous run code and jammed togther, take your pick as to which is the master. This way the slate and recorder always have sync'ed timecode. At the start of each shot, slate for each camera with the slate's timecode visible. In post, drop the video into the timeline and align it to the timescale so the marker frame with the slate's visible code matches the timeline's time value, drop the audio file in and the imbedded code in the file header will line up with the timeline as well, thus syncing the audio and video files.
 
Hi

Take a look at new Zaxcom Maxx. Scheduled for June delivery I think. It has built in time-code slate. Also has Never Clip, it apparently can't be clipped because it's dynamic range is greater than the microphones. It has an option you can order that makes it 4 channel ISO recorder plus mixdown. Not too many $$$ or Euros for what you get. I have one on preorder.

You might also take a quick look at technology that allows you to record timecode onto non timecode cameras via an audio track. Zaxcom has a version of it. Although at their price you could just pay more for a time-code enabled camera. But I believe other companies do it for less. I hear varying things about success. So check and test carefully.

To add one more opinion, I would recommend just renting cameras that jam sync and be done with it.

Good luck with your show.
 
The FS-100s can generate time code, but you can't get at it until you edit. But if you have audio with timecode, you can link the two timecodes. Also, it might be possible (through a LANC connection) to get all the FS-100s to start at more or less the same time. You would need cables between them, which might get in the way.

Incidentally, many comedies are shot live, and then edited later. Are you planning to do that? In that case, you might get all cameras to stay in sync for 30 min.

In terms of live switching (if that's of interest) you might check out the Blackmagic ATEM switchers. (Starting at $1000.)
 
People have been aligning multiple tracks for years with a clapper hit - and I've never found sync a big hassle. Zoom in, slide, next track, next track - repeat till done? The only thing I have trouble with is clip ID - so I make sure that somebody always calls the shot number very loudly so it's on the header of each video and audio clip. Plural eyes is great, but I've never had a project where I wished I'd had it?
 
paulears said:
...The only thing I have trouble with is clip ID - so I make sure that somebody always calls the shot number very loudly so it's on the header of each video and audio clip.
Click to expand...

In my limited short film making activity, this is the single most significant 'life saver' to do, both the slate visual and the call out. For a nolo budget production with limited 'script supervision', having the shot numbers visually, and the audio call out, may be the only way to maintain sanity when setting up for the initial spotting and editing activities.

The only experience I've had with "Plural Eyes", and this was just as a 3rd party observer, was for a 48-Hour project last year... the clips that were marked by me, along with the audio were 'aligned' within 'minutes', but shots and audio that were expected to use Plural Eyes, and there were many... seemed to have more hassle than it was worth, especially with the short time limit. Perhaps the people didn't know how to use the tools, but it seemed like they had 'matches' with the wrong audio track more often that not, and calling the shot was not 'consistent'... lead to difficulties...

While the script person and I agreed on a convention for video and audio logging, a couple of the camera operators were inconsistent about what they 'deleted', before having me (acting as data wrangler...) transfer the video and audio to the archive disks. So despite the script person getting audio track's unique 'numbers', the video had 'missing' clip 'numbers' added to confusion when setting everything up for editing...

The short that I directed last... now receding into the distant past of last September... I had slates and callouts for all shots, including MOS, and was able to align things pretty quick... just life has gotten in the way and I've not finished editing... I'll be working on that in the next few weeks...
 
I've done double system on just about everything and using a clapper is by far the best, most secure way you can do it. I've done this on Nagras, Tascam DA-P1s, HHB Portadats, Zooms, and currently into a Sound Devices 552 or a Tascam DR-680 if I need iso tracks. If you can shoot camera reference and use Plural Eyes, more power to you! But syncing from the DR-680 isn't really that much of a chore--that's what interns and assistants are for anyway--there's always somebody young that is willing to spend a week synching footage when you tell them it is for a feature.

If you just use a clapper you shouldn't have to worry too much about drift--that's usually not a problem for dramatic feature work, but for longer form production like concerts and plays. It can happen, but it's usually a relatively easy fix--that's why god invented the razor blade tool!

Just make sure the scene info is called out loudly and that all rolling cameras actually see the clap when it happens. Mostly importantly, get a sound guy who knows how to do this already!
 
j1clark@ucsd.edu said:
The only experience I've had with "Plural Eyes", and this was just as a 3rd party observer, was for a 48-Hour project last year... the clips that were marked by me, along with the audio were 'aligned' within 'minutes', but shots and audio that were expected to use Plural Eyes, and there were many... seemed to have more hassle than it was worth, especially with the short time limit. Perhaps the people didn't know how to use the tools, but it seemed like they had 'matches' with the wrong audio track more often that not, and calling the shot was not 'consistent'... lead to difficulties...
Click to expand...

I have been using PluralEyes lately, and it has been a big time saver. In addition to aligning audio and video, it also puts all files in chronological order, which comes in handy when using multiple cameras. For PluralEyes to work well, it requires clear audio from the camera for reference. If the camera's audio is crap, PluralEyes has a hard time analyzing the audio waveform for alignment. Also bear in mind that if your camera and audio recorder mics are a substantial distance apart, there will be a time discrepancy that is easily fixed by nudging the audio a few frames. This isn't a flaw in PluralEyes, just the physics of audio delay due to distance.
 
Alex H. said:
As the most low-tech option, you can do what you originally thought and rough in the time code manually on each device and hope for the best (not exact, and you may experience some drift), relying on a clapper for sync marks.
Click to expand...

Drift is an issue unrelated to TC. The only way you won't get drift is to lock everything together with wordclock or some other sync setup. And the cameras mentioned can't do this.

OP,

Drift will probably not be an issue if your takes are relatively short, as in under 3-4 min. Some point after that you will probably have one or more devices drift away from each other.

You can't set the clocks accurately enough on the devices to do more than get you in the ballpark. You can use what Alex described and the TC stamps will all be the same, but this is really just a high tech clapper these days because TC is not striped continuously on files, it's stamped at the start. Though actually you would be recording it on an audio track with the cameras so it would be continuous but bot tied to video frames.
I don't know FCP well enough to know if matching takes based on audio TC would even help, maybe.

Also a bunch of features in FCP were left out of FCPX. I would avoid the X if you can, because a lot of those orphaned features are ones very helpful or necessary to doing post that isn't all inside of the FCP suite. This looks like a more complicated project and with X you won't be able to do basics like send your project to sound post.

I would probably go with a mix of setting the clocks on the devices to the same time (as close as possible) and then doing the fine sync with the clapper. I would also record camera audio and send that with everything else to sound post so they can fine tune the sync.

The reason for the last is that (with out using Pluraleyes) your sync is only going to accurate to about a 1/2 frame (with out taking a LOT of time) because NLE's work in frames. Sound post doesn't have that limitation and they can "phase" the dialog to the camera sound and get much tighter sync.

I don't think you can record more than one video feed into FCP but? I wouldn't record direct to FCP, it's not stable enough and that isn't what it was designed to do. If you want to record to computer I would use "Vscope Live", it's designed to record video and has a bunch of things that will make your life easier. And it doesn't cost much, I think it was under $100 when I got it.

If you need a lot of audio tracks and you can use laptops (I'm assuming so if you want to record video to one), then I would go with BoomRecorder. It is extremely reliable is great for production recording and it and some decent interfaces will cost a lot less than a couple of SD recorders. I ran a test and had it recording 64 tracks of 48/24 audio to a USB drive. I ran for 10 min and it never glitched. Plus you can take notes and it will generate sound reports. And it will jam to incoming TC if you want.

Boom Recorder

V Scope Live
 
You must log in or register to reply here.
Back
Top