View Full Version : Best in-the-field documentary shoot practices for easier/efficient Plural Eyes syncin
January 14th, 2011, 11:26 AM
Doing a very traditional (odd to say it) HDSLR-audio workflow. So double system - Rode mic on the camera (typically Canon 5D or 7D) to record better than internal mic reference and then "master" recording on Zoom H4n (typically a lav and shotgun/boom inputs, 48 kHZ 24 bit WAV files.) Then mass sync.ing with Plural Eyes in FCP. But unfortunately I'm getting VERY mixed results. Probably only successfully sync.ing up about a 1/3 of the clips for any given shoot after the first PE pass (lining up all audio & video clips chronologically, and checking of the sequential/chronological and try really hard functions.)
Here's the important thing to know about my production. We’re straight up doc shooting, ie shooting A LOT of footage thru out course of the day. Averaging 200-300 individual video files, 20-30 individual Zoom WAV files per shoot day, without any kind of scripting.
So I understand there's some inherent complications in this, that I'm going to have to do a certain amount of more targeted PE sync.ing, as well as straight up old school manual syncing in FCP.
But anyone have any suggestions on what I can do in the field to make sync.ing life a little easier in the edit room? Right now I’m slating in the field – as much as possible, and that's difficult given the unscripted nature of what we're doing – as well as very carefully saving the WAV files in separate individual Zoom folders per the CF cards. So all CF card 1 associated audio is in Zoom Folder 1, CF2 audio in Folder 2, etc. (I know they generate duplicate names, so I rename later.)
Should I do more than slate - maybe an audible countdown? Should I record more individual Zoom WAV files? Is there some way I can "game" Plural Eyes, ie produce in the field that audio characteristic that best helps PE to sync?
Any suggestions would be really helpful. Or even just pointing me to pre-existing online resources that deal with this. The footage is really staring to pile up.
January 18th, 2011, 09:30 AM
I never used an HDSLR, a Zoom or Plural Eyes but this is what comes to my mind: if you had a second receiver you could try to feed the camera with the signal coming from the lav, this way Plural Eyes could work more comfortably (having to deal with two virtually identical tracks).
January 18th, 2011, 12:25 PM
The best way is to split the output of your recorder such that you have outputs running to your dslrs audio inputs as well as outputs for monitoring. By recording the output from the recorder on your dslrs, pluraleyes will have a much easier time syncing up the footage, even if the automatic gain will screw it up a little bit.
January 18th, 2011, 02:01 PM
By recording the output from the recorder on your dslrs, pluraleyes will have a much easier time syncing up the footage, even if the automatic gain will screw it up a little bit.
Exactly what I was saying, don't know why I didn't think about getting the signal straight from the recorder.
January 18th, 2011, 07:13 PM
Thanks fellas. Actually, for a # of shoots that's what I've been doing - splitting out of the Zoom's headphone jack, one cable line to HDSLR directly, one to monitoring headphones. But haven't noticed dramatically different sync.ing results. And this method is inherently very risky - if for whatever reason cam is not getting audio signal, you get NO audio (camera won't automatically start recording on internal camera), and therefore no reference whatsoever. Nick.
May 19th, 2011, 10:29 PM
Hey Nick I feel your pain. Just finished a Doc that was about the same size and number of clips and audio files and had trouble a plenty with Plural eyes. In fact it was quicker for me to sync manually using the camera audio and Hn4 stuff than to fiddle about waiting and watching plural eyes.
My system is actually a Hn4 AND a Beachtek into a 7D. I split the headphones out of the HN4 but at least with the Beachtek I can see through the level meters and phones that Audio is reaching the camera to as much an extent as with just the 3.5 jack into the camera.
I also use a splitter that is 3.5 male to 3.5 female to 2 channel xlr adaptor cable for a little more security and ruggedness. The XLRs go into the Beachtek and stay there, pretty much. It means you can either monitor out of the HN4 and still pass audio to the Beachtek or you can just monitor the Audio on Both via Beachtek headphone out.
I also run AGC control on the Beachtek so that I am double backed up - if for instance I forget to roll the HN4 I still have useable audio on the camera. Always run my main audio - Talent Mic for eg. on Ch2 and on-camera on Ch1. On top of that have the on-camera as good a quality shotgun as possible and always have it on. This has saved my ass on several occasions.
I also have an SD card locked with 5 minutes of 1k tone on it that I play out of the HN4 to set the beachteck up and check the system before i shoot.
Just need to make sure the 9V in the beachtek is OK throughout the day too. Although I think it will just let the audio pass through you just loose Phantom power and AGC control if batto dies. Never tried to be honest....
I have tried many things with PE and only had limited success. I assume I'm doing something wrong as it won an Emmy! I'm so fast at syncing now It's immaterial and I get to preview footage anyway!
So I'll be interested to see any other tips on improving PE Performance on this thread!
May 19th, 2011, 11:12 PM
First, I use Dual Eyes, which is the stand alone version of Plural Eyes. Essentially your load tell Dual eyes to synch file you have shosen, and you can tell it to make a second file with the the separately recordes sound, while preserving original files. I only have Vista 32 bits and 3 gig of memory, and it seems that if I put too many files in the hopper, Dual Eyes can hang up. Consequently, I try to have a good idea that the files I am putting in a batch synch will match up. I try to limit it to to about 40 files.
Second, I shoot my 5 D or my T2i with camera mic only. I leave AGC on in these cases, so we can have a good level to work with. I have tried using another mic, with Juice link or not, and in doesn't seem to make a difference. So I jsut let Dual Eyes do the work. And since I don't monitor the DSLR sound, using other sources increases the potential of not getting.
I record to a simple Tascam DR-07 attached to and fed by my Sign Video Eng44 and get a pretty clean sound. I shoot mostly narrative film.
May 21st, 2011, 03:47 PM
I'm so fast at syncing now It's immaterial and I get to preview footage anyway!
Ian - don't mean to hijack the thread, but I'm having trouble with PluralEyes too but I'm very slow at syncing in Final Cut. I was wondering if you could describe exactly how you do the sync manually... Thanks! -JP
May 25th, 2011, 08:58 PM
Ian- Thanks for the detailed report on your HDSLR audio practices.
Sounds like you're really doing it as pro-style as you can.
Funny, heavy duty in post and fortunately haven't had the middle of the night, OMG this camera just screwed me moment yet. But I'm waiting for it.
When/if we start up again, seriously considering Sony F3 or F100. Or if Canon gets their audio act together with pro style inputs, during recording levels monitoring, and headphone outs.
May 26th, 2011, 11:35 AM
Clap twice at the beginning of each take.
Beyond that, I have nothing more gear-wise to advise... but the above has worked without a hitch (thus far).
May 31st, 2011, 09:25 PM
Hi, sorry for the time between posts, trying to sell a film. A wise Executive told me the other day that actually making Television was the easy bit, getting someone to give you the money for it is the tester.
Anyway I usually make sure that out in the field I use the stop and then record instead of pause on my Hn4. That is good for two reasons, if the battery goes dead then you only loose the current grab and also it means that all your audio files are of a similar size and number to your video files.
The in the edit well it's pretty simple, as I said the camera has a guide track split off the Hn4, and I first set up a new timeline, and dump all the interview footage and audio from the camera files onto that, then all the audio files from the HN 4 onto ch 3 and 4. If you followed the previous step there should be a similar amount and they should be of similar but not exact length.
Hit Option Apple and "W" to get wave form up and loosely match the separate audio to the sound on the camera files. You roughly get them as close as you can visually and by dragging with your mouse, - you'll know because your voices will have an echo instead of a the gobbledygook of two random audio streams playing at the same time. None of this is particularly hard or time consuming, and when the echo is fairly close, I just select the Hn4 audio and option left and right arrow and slip the grab until the echo goes and they are perfectly in sync. Trust me, it is startlingly lear when they lock up. Also just make sure that right from the start you have linked and locked the video and audio tracks off your camera.
If I am getting confused about which way to go, I find the highest and sharpest peak, and put a marker on it in each track, then you know which way to slip for closer sync.
All I do then is link the clips and or delete the junky Camera sound. You can export them our as new clips but I always keep those timelines, just in case!
One last thing, if for some reason the camera tracks are mute and I haven't used a clapper, I look before during and after for an easily recognisable moment where there might be a sync point. I even watch the interviewees expressions to see if they are waiting for me to get a light stand right or someone has come in a door, that can get you close to your sync marks quite quickly. Then I watch their lips for clear "B"s or "P"s and use those for close sync!
Is that very clear. With a bit of practice you can get stuff down very quickly. Trick is the matched camera and Hn4 Audio and using the echo and a bit of jungle warfare to get perfect sync!
The other thing I do is just pay for someone to turn up with a proper camera and a sound recordist and stop mucking about with these stupid fake video cameras. ;-)
June 4th, 2011, 09:09 PM
Sony, Canon, and Panasonic just came out with a brand new product for shooting documentaries! It's called a video camera. The incredible thing is that it can record video and synced sound simultaneously! It records hours and hours of video and audio together. Then when you have to edit, there is no transcoding necessary, just drop it on your timeline and everything is good to go. Why waste time working twice as hard and twice as long with old fashioned DSLRs? DSLRs are best for B rolls, not documentaries. Check out a brand new video camera, and get on the bandwagon!
June 15th, 2011, 12:11 AM
Thank you Warren for that very helpful and informative post.
June 15th, 2011, 02:54 AM
Thank you Warren for that very helpful and informative post.
He has a point...
June 19th, 2011, 05:48 PM
Who does? me or him?