Should I try to use Time Code. . . or just forget it?

[Stephen Pruitt]

Hi all. . .

I'm facing a quandry. We'll be shooting a feature film with the RED and recording our audio on a double system using the new Sound Devices 788T (8 tracks of audio).

The question I have is whether the time code nonsynchronicities consistently discussed here mean that I shouldn't bother trying to use time code in the production at all, and just rely upon the old mechanical clapper system for synching sound.

What advantage do we have using the time code (we'll be using FCP 6)?

If there IS an adavantage to using time code, should be try to synch everything with a Denecke time code slate or just synch off the camera or synch off of the 788T?

Thanks very much.

Stephen

[Brent J. Craig]

Time code worked perfectly in our tests. Jamming the camera from a Sounddevices 422 and shooting a slate to test. It was always within 1 frame.

I suspect a lot of the mentions of timecode problems on RU are because there are many frames where the timecode slate numbers are in the middle of updating as the rolling shutter records them. It is very easy to mistake the transition between 3 and 4, for example, to be an 8. In that case, you simply advance one frame and you can read clean code.

To state it another way, the rolling shutter does not record the image of the timecode slate at a single instant in time, it takes a small amount of time for the shutter to scan the image and therefore a single frame can contain the top of one timecode frame number and the bottom of the next.

[Stuart English]

Time code worked perfectly in our tests. Jamming the camera from a Sounddevices 422 and shooting a slate to test. It was always within 1 frame.

I suspect a lot of the mentions of timecode problems on RU are because there are many frames where the timecode slate numbers are in the middle of updating as the rolling shutter records them. It is very easy to mistake the transition between 3 and 4, for example, to be an 8. In that case, you simply advance one frame and you can read clean code.

To state it another way, the rolling shutter does not record the image of the timecode slate at a single instant in time, it takes a small amount of time for the shutter to scan the image and therefore a single frame can contain the top of one timecode frame number and the bottom of the next.


Good point. In my tests timecode is accurate, however you can make a recording of a slate where the numbers look to be in transition.

So you might like to experiment with SHUTTER PHASE to sync the video frame capture of the camera to the slate display refresh rate.

[Stephen Pruitt]

Thank you for your responses. . .

Would you jam the time code off the camera? Of the 788T? Or off a time code slate???

Thanks very much.

Stephen

[Emery Wells]

Ive had problems jam syncing at 29.97 (i.e. would not sync) in build 15. 23.98 has worked perfectly for me.

[Cail Young]

Thank you for your responses. . .

Would you jam the time code off the camera? Of the 788T? Or off a time code slate???

Thanks very much.

Stephen

The camera does not currently output TC, so the 788T or the slate would be your master clock.

[Mark L. Pederson]

We shot our first two features on RED the old fashioned double system way. All sync was manual in FCP. It was really not that big of a deal, but it was a few hours every day.

Now that we have audio enabled on camera, it is so much faster and easier for dailies and off-line edit media transcodes to just transcode the QTs the camera makes if you don't mind being tethered. We find most productions tether anyway for monitor, etc.

We have a feature shooting now that is doing a very specific hand-held style of shooting and they insisted on NOT being tethered at all. For that film (Toe To Toe) we put a wireless comtec on the camera to get "scratch" audio for dailies and editing into the camera - we are jammin' - but to a Lock-It box attached to the camera. So in final audio post, we can do an auto conform of the original BWF files to the final edit.

Bottom line, IF you JAM - you should use a Lock-It box (or similar device) so you don't have to re-jam every time you change battery or turn the camera off - that is madness.

If you are running audio to the camera all the time - there is no real reason to jam.

And obviously always keep your BWFs from your mixer in case you need to replace the audio recorded by the camera for any reason.

http://www.gothamsound.com/main.shtml?page=catalog&scmd=DISP&itemid=SAMBLOCKIT

http://www.pro-sound.com/Merchant2/merchant.mvc?Screen=PROD&Product_Code=SDENSBT&Category_Code=S

[martijn76]

hi i own a lockit and might do sound for a short flick shot on red i saw that there is a sort of lemo connector for the red tc in? i used to connect 2 bnc's from lockit to camera?

i use my 744t to jam my lockit via a lemo>xlr female>xlr male>lemo

can i ue the same cable to hook up the lockit to the red?

cheers
martijn

[Stephen Pruitt]

Mark Pederson is a true fount of knowledge.

Amazing.

Looks like we won't jam. . . here's our plan. . . tell me if it make sense:

With the 442 and the 788T, we'll have up to 8 XLR inputs: several lavs, a few plants, and a boom or two. All of these signals will be recorded on the 788T.

Our plan is to tether the camera to the 788T via the four mini-XLR outputs on the 788T, with the four most-likely best audio sources coming out of the 788T and then directly in (line in) to the RED. If we are very happy with those four audio signals, we'll be in like Flynn.

If, however, we feel that we didn't get the best audio into the RED, we can then manually link the (up to four) extra audio files inside the 788T into FCP in order to get the audio up to snuff.

Thus the 788T will serve three purposes:

1) Providing 4 extra XLR inputs.

2) Recording ALL audio signals all the time and providing back-ups just in the case the RED's audio glitches.

3) Giving us more than four channels of audio "coverage" should we not like the four channels recorded directly to the RED.

Make sense, Mark? Anything you'd add?

Thanks so very much.

Stephen

[michael zaletel]

Thanks for these detailed discussions.

-shooter

[Cail Young]

i use my 744t to jam my lockit via a lemo>xlr female>xlr male>lemo

can i ue the same cable to hook up the lockit to the red?

Should work, it's the same pinout.

[Stephen Pruitt]

Mark Pederson is a true fount of knowledge.

Amazing.

Looks like we won't jam. . . here's our plan. . . tell me if it make sense:

With the 442 and the 788T, we'll have up to 8 XLR inputs: several lavs, a few plants, and a boom or two. All of these signals will be recorded on the 788T.

Our plan is to tether the camera to the 788T via the four mini-XLR outputs on the 788T, with the four most-likely best audio sources coming out of the 788T and then directly in (line in) to the RED. If we are very happy with those four audio signals, we'll be in like Flynn.

If, however, we feel that we didn't get the best audio into the RED, we can then manually link the (up to four) extra audio files inside the 788T into FCP in order to get the audio up to snuff.

Thus the 788T will serve three purposes:

1) Providing 4 extra XLR inputs.

2) Recording ALL audio signals all the time and providing back-ups just in the case the RED's audio glitches.

3) Giving us more than four channels of audio "coverage" should we not like the four channels recorded directly to the RED.

Make sense, Mark? Anything you'd add?

Thanks so very much.

Stephen

[Philippe Vandendriessche]

Stephen,
with your method, you only "capture" sound sources without "location mixing".
Are you working alone or with a sound engineer without arms?

"Martijn76": I think you can use your lemo/xlr adapter to send the TC to the RED. But remember there is no TC at TC out of the RED (for the moment), and the link has to be permanent to avoid TC lost when changing battery on the RED.

[Stephen Pruitt]

Actually, I will have a location audio expert on site, but, coming from a recording studio background, I'm not at all in favor of mixing on site. In my opinion, there are few things worse than a mix done "in the moment." I'd much rather capture as much relevant audio information as possible, and then put it all back together at a later time. Once something's mixed and then recorded, it can't be unmixed. But that's just me. If I had more experience, I might try to go another way. I'm just wanting my sound designer to do one thing: Capture great audio. . . and then. . . yep. . . you got it. . . "we'll sort it all out in post."

:-)

Stephen

[Mike Paul]

I would strongly suggest jamming the TC from the 788T into the Red. It is a simple and painless procedure that may save you hours of hand-syncing the BWF files in post. Set the 788t's generator to 23.976, jam the Red, and re-jam every time you power down the Red. If you use a Lockit Box (always a good idea), and you plan to use a Lemo-Lemo cable, the fact that the Red does not output TC will actually save you. The Lockit uses the Lemo input to jam itself to external TC, which it does automatically. If the Red were outputting code, and you connected a Lockit to it, the Lockit would instantly re-jam to the Red's TC output - an it would re-jam every 20 seconds or so, rendering your TC sync unuseable.

In the future, when the Red can output TC, either make sure that the TC output is turned off, or use the BNC output of the Lockit to feed the Red TC input - that way, it won't re-jam.

The TA3 outputs of the 788t are variable, and are set from the factory at a +4db (Pro) level. The Red Line inputs seem to be more of a -20db (consumer) line level, so you will have to adjust the TA3 output of the 788t down a little bit to avoid clipping the input section of the Red.

Oh, and by the way - the 788t is not actually shipping yet. It will be released in limited quantities later this month (and there is already a hefty "Red-Like" waiting list to get one). So, unless you're guaranteed that you'll have one in time for your shoot, you might want to come up with a "Plan B".

Mike Paul
Technical Director
Location Sound Corp.

[Stephen Pruitt]

Hi there, Mike. . . thanks for the tips.

I actually will be getting my 788T VERY early. . . so no worries on that score.

:-)

Stephen

[Julian Banos]

I am very interested in your results. I will be looking foward to any news. Please post.