GenLock EX3 and EX1: Is It Possible?

basspig

basspig

Veteran
A member of the board of directors of an orchestra that I work with e-mailed me today and announced that they're in the early planning stages of doing some type of multimedia thing with live projection in the concert hall.

Basically, two of our four cameras would feed a live switcher, the output of which would feed a couple of Christie CP series projectors, firing 100' diagonal sized images on the side walls.

In a rough outline, I think I need a switcher capable of HD-SDI, possibly a repeater for the balcony camera, as there's over 100' of HD-SDI cable from that camera to backstage, a couple of monitors for the switcher, a HD-SDI to DVI converter, and possibly repeaters to feed the Christie projectors. The projectors need 208V 3-phase power, which is something the concert hall electricians will figure out.

Now the question I have is whether it's possible to genlock an EX3 to an EX1? I had a vague recollection that this was possible, but since there is only one SDI out on the EX1 and no TC out, and the EX3 documentation talks about using reference video plus the genlock out, I'm left with some uncertainty.

Another possibility is to rent a switcher that has a 10-bit frame synchronizer, like the DataVideo SE1000, but that depends on what's available and what the orchestra committee will pay for rental.

Has anyone used these two cameras in a genlock/live switching broadcast situation? If so, how did you wire the two cameras and what menu setups were required?
 
I haven't done it, but it can be done. If you have the manual for the EX3, genlock info is found on pages 127-136. Maybe you can download it in PDF file.
 
Jay, I'm looking at the manuals for both EX3 and EX1 and nothing is mentioned in the EX1 manual about genlock. The EX3 manual suggests that it can genlock to another EX3, not an EX1. There is no mention of how to connect either camera's time base to the other.
I don't want to leave discovery of this to the rehearsal night before the concert, if at all possible. :)
 
From what I'd read on some forum a while back, it appeared as if it were possible to slave the EX3 off the EX1's time base. Is this not true? If so, then a live multi-switched shoot between an EX1 and EX3 would not be possible without some external means of time base correction. I need to determine these things before I can figure out what the concert folks will need in terms of switchers.
 
Just for curiousity sake, I fired up my HP 1740A oscilloscope in the shop and I connected a BNC cable between the scope and each of the three EX3 outputs, one at a time, to see what the signals looked like. Obviously, the analog video looked as expected. The TC OUT looks like a PSK modulated square wave, as expected. The SDI output looks like 30MHz broadband noise with random blanking occuring.
Obviously, these are all vastly dissimilar signals. One could not expect the TC IN to make sense of the SDI out of the EX1. Imperical evidence suggests that to do a live switched multicamera shoot, one needs to be running all EX3s. Perhaps I'm wrong about this and I hope I am!
 
you take the green component output or composite output and all sync signal are here to allow connection on an Ex3.
the problem being that the distance between camera is limited and the connectors are just a nightmare.
 
You should be able to just connect the HDSDI out of the EX1 to the Genlock IN connector on the EX3 and the EX3 will be locked to the EX1. You would have to make sure the EX1 HDSDI output was set to the same framerate as the EX3. This is pretty simple Old School stuff. The Y (green) channel of the component output should work too if it's set to the right frame rate and resolution, but the cable length would have to be shorter

For external monitoring, simply use a monitor with an HDSDI loop through and send the looped signal to the genlock in on the EX3, or use an AJA/BlackMagic 1 in 3 out HDSDI box.

Synchronous timecode would be another matter. You would have to use a Clockit timecode generator on both cameras recording onto one of their audio tracks.
 
nosys70,

I don't see how connecting an analog synch signal to a digital synch input would work.




ZumaDog,

the HD-SDI signal is of a completely different type than that of the synch signal. They are not even remotely similar when viewed on a scope. Did you actually try this, or is this an assumption? The synch out from the EX3 is 2.1 volts p-p digital square wave that swings both negative and positive, whereas the SDI output is positive swinging at 0.8 v p-p.
I'm not even certain I'd want to risk connecting the two in this manner as the possibility of damage exists if signal voltages are incompatible.

If this did work, I would need to be able to split this HD-SDI output to feed a switcher input, along with the EX3. The cameras only need to be synched to work with the live switcher. Timecode for any other reason is unimportant to this application.

Remember, we would be feeding at least two EX cameras into a switcher, which would be feeding a couple of Christie CP2000 projectors in the concert hall. I'm pretty certain that, like the analog counterparts 20 years ago, the digital video cameras need to be synched in order for the switcher to display both images correctly and switch between them.
 
on the EX3, there are 2 inputs , one for the genlock, another one for the timecode.
Genlock is nowhere digital, it is by definition a video clock signal, the same one you find in composite video signal when video is black. Most of genlock input accept video signal and strip the clock signal out of it.
You take absolutely no risk of feeding a 1Vpp video signal into a Genlock input.
The biggest problem with digital camera is sometime, the composite or green component signal could be slightly delayed due to some conversion.
In that case genlock would not be perfect or possible. I do not think it is the case with the EX1.
the only problem with the EX1 is the connectors used for composite and component . They really sucks. weak, misplaced and unreliable.
 
Last edited:
I apologize for my the inaccuracy of my earlier post. I forgot to include the crucial step of converting the HDSDI output of the EX1 to analog component via an AJA HD10C2 or similar conversion device utilizing the Y green channel.

The straight analog component output should also work as long as it's set to output at the same frame rate as the EX3 is set to record to.
 
Last edited:
I did an experiment, taking the green channel of the component out of the EX1 and fed it into the EX3's Gen Lock In. When the EX1's comp output was enabled, the EX3's viewfinder showed a green flashband momentarily, like it was trying to synch, but did not display Lock in the display, nor did the timecodes of both cameras match. Any ideas?
 
Yup. Same 24p rate, and same resolution. Also tried HD and SD outputs for the YcBr outputs. Neither produced an ext lock.
 
strange, a guy on DVinfo mentionned he synched an ex3 with an Ex1 doe 3d shoot using this method.
http://www.xdcam-user.com/?tag=3d

something that sounds strange to me is you mentionned 24p and this frame rate does not really exist.
from the camera output , it is probably a 30fps made by repeating some frames, so you should try 30fps first on both camera, because i am not sure the camera can output a component a real 24pfs.
 
Last edited:
I don't shoot 30p because it's not compatible with Blu-ray.

The EX3s would be fine for live if used with other EX3s or cams with real genlock output. The issue is when trying to genlock off an EX1.
 
The timecode will be unaffected by only genlocking two cameras. On broadcast 2+ camera shoots we have always run a genlock cable out from one of the cameras (hero cam) and also fed it's timecode out to the other cams. Genlocking and synching timecode are two separate feeds using two separate cables. Syncing timecode to an EX1 is not possible due to the lack of a timecode IN connector that the EX3 has. There is a possible work-around which is to take the EX3 timecode output and (converting the BNC to a 3-pin male XLR) feed that signal into one of the two available audio tracks as an audio recording. Avid's have no trouble substituting the non-sync timecode for the audio track recorded timecode - but I suspect doing this in FCP might be more difficult.

I don't know what indicators or metering Mark & Mary are using to determine locking, but the green flash is probably a good sign that the cameras are indeed locked together. I am unaware of any "lock" indicator occurring in the viewfinder of these cameras.

I won't have my EX3 back until next week so I can't do a test to determine, but usually there are no obvious indicators of genlock visible.
 
Okay, that clarifies that the timecode is optional (and really unimportant to a multi-camera live shoot).

I did get the green flash while using composite from EX1 to EX3 genlock in, as well as with Component Green channel from EX1 out to EX3 genlock in. However, I was expecting to see the LOCK indicator in the display of the EX3, but did not. I suppose I could determine if the two cams were in synch if I still had my old FX console from 1987 to take the composite outs from each and drive the console and do a cross dissolve. Failing that, perhaps taking the composite outs and feed each to channel A and B of my oscilloscope and observing if they are in step with eachother after the connection is made from EX1 to EX3. If they are out of step before, and in step after the cable plugin and green flashband occurance, then that would be assurance enough.
Why the lack of the LOCK indicator in the display though, is worrisome.
 
I conducted an experiment, using a dual trace oscilloscope and watching the synchronicity of the two camera's outputs.

I took the composite BNC out from EX3 and fed to channel A.

I ended up using a splitter for the component out on EX1, because composite and component outs cannot be active at the same time as it turned out. So I took the split out from EX1 and fed one to channel B on the 'scope.

At this point, I could see the pedestal waves were something like 90° out of phase. Then I plugged the other BNC off the EX1's split Component out cable into the Genlock IN on the EX3. After the displays flickered a few times, both pedestal waves were now in lockstep, starting and ending at the same time.

So, this verifies that the genlock works. Only question remains as to whether it would work with a 100' cable between cameras located stage right and balcony positions. Guess next test is to obtain 100' of BNC cable and try it. But this does look promising. At least we can do a two-camera live projection with switching/dissolves between cameras.

If I unplug the Genlock in cable, the pedestals for cam A starts to lag over several minutes. So they have to stay connected the entire time of the live broadcast.
 
Last edited:
You must log in or register to reply here.
Back
Top