F3 3D genlocking issues

[Forest Finbow]

Hi folks,

We're shooting a 3D feature with three F3 based 3ality rigs and we're experiencing weird genlock issues on the cameras.
Cameras seem to have trouble getting in sync : when we plug in the reference cable, the camera takes a more than reasonable time to clock, sometimes up to 30" er even a minute if we're unlucky.
The worst part in that is that we also have experienced genlock losses that completely freeze the camera (even the power off button) until we remove the reference cable.

It happened on all 6 cameras so it doesn't seem to be related to a specific camera, all F3 seem concerned. I should also mention they are running v 1.1 firmware.

The reference comes from a Copper Head, fiber channel linked to the tri-level generator. This configuration has been running smoothly on different other cameras this is the first time we experience these problems.

So far, all we can do is be patient for the cameras to clock and if they fails to get in sync unplug and replug the reference cable. It hasn't really got in the way of our shooting but it did raise questions in prep time, did any one else experience comparable issues?

[Thierry Humeau]

To the risk of sounding stupid, all your cams are set to exact same HD raster and frame rate right?

Cheers.

[Nate Weaver]

I think a link to Sony's techs via 3ality is going to be your best bet.

I'm actually surprised 3ality doesn't have a beta version of FW 1.3 running on your cams.

[Forest Finbow]

There are no stupid questions ;) but I can certify all settings are perfectly the same on all three pair of cameras.

The production began with firmware 1.1 before the release of 1.3 which is as you mentioned still actually beta so that felt kind of risky to change the firmware in the middle of the shooting.

We tried without the copper head (with a direct BNC connection from the generator) and the problem seems the same.

Sony techs from france and japan are already involved, this seems to be the first time anyone experienced this and this is the reason why I wanted to ask the community, to gather info in case someone else had the issue.

[Alister Chapman]

That's a strange one. I've been doing lots of 3D with F3's as have many of the Hurricane Rig users and no one has reported Genlock issues.

[Nick Hiltgen]

What happens when you take the y out of one camera and feed it to the genlock of the second camera?

[Forest Finbow]

I know, it's weird it's the first time I see that too.

We haven't tried to feed the ref from one camera to the other but that would not change much since the master camera still needs to receive the reference from the tri-level generator.

[Nick Hiltgen]

Are the cameras being cut together live and that's why you need a master genlock source?

In my experience if the cameras are not being intercut live you they will be ok as long as the left and right eye are genlocked.

[Forest Finbow]

It's not live so I would have to ask the reason we need all the different rigs to be clocked together, what you say does indeed make sense but even when shooting shoulder mounted D and E cams (Sony TD300) we receive genlock and timecode through ambient so it seems important indeed.

There is a bunch of different carts that get the signals passed through: recording, 3d rig control, 3d director and production monitoring and finally video village so having a single clock for everybody does make sense.

Another thing I should mention is that the rigs were built in america, running on the 110 V 60 Hz electric power system and are now running in France on 220 V 50 Hz power outlets. All the devices involved should be 100-240V 50/60 Hz compatible but I was also thinking that could be a source of problems since clocking could also get unhappy due to power frequencies conversions. DC produced from AC is after all not really flat, that could also get in the way couldn't it?