View Full Version : Canon XF105's and Zoom Axis Shift


Alister Chapman
February 27th, 2011, 09:38 AM
I have been playing with a pair of Canon XF105's on my mirror rig and with the 105's you can use the optical image stabiliser to compensate for lens axis shifts. It's very clever and takes only a couple of minutes to calibrate the cameras using the built in routine. You zoom in on a target, then zoom out and use the calibrate function to shift the on screen image up/down, left/right until the image is centred on the same point and press save. Once both cameras were set I was able to do sync zooms that tracked almost perfectly using the IR remote. Very impressive.

Bruce Schultz
March 5th, 2011, 08:21 PM
Here's a shot of two Canon 105's on your Hurricane Rig here in LA

Alister Chapman
March 6th, 2011, 01:26 PM
We should have news on some Nun's knickers for the 105's at NAB as well as a handheld kit for the rig. Were working on the motor kit and oversize hood as well!

Tim Dashwood
March 9th, 2011, 03:56 AM
Does the XF305 have the same built-in routine?

Bruce Schultz
March 11th, 2011, 11:47 AM
Tim, surprisingly, I don't think it does. Geoff Boyle on CML posted a beta test review some months back about these features for 3D, but indicated then that they were confined to this 105 model. Maybe that's changed with a firmware update, I don't know, but I was pretty impressed with how easily these little guys lined up on the Hurricane Rig the day we tried them out utilizing the lens alignment features. Also, recording a pretty robust 4:2:2 (albeit in AVCHD codec) was impressive - but like all the rest, it's best to record out to I-Frame Nano Flash for optimum results. It's also only a single CMOS chip, 1/3" I believe.

Tim Dashwood
March 11th, 2011, 06:22 PM
Hi BRuce,

I have two XF105 rentals in my hands (just picked them up) but I'm working on genlocking one to the other. Did you use an external genlock or did you take the component Y out and feed it into the other camera's genlock in? This is what I had to do to get genlock between them. Unfortunately that means I can't also constantly slave TC. I can slave some free run TC on both and check periodically.

The little hitch we noticed right away was that the built-in mic is too far forward to put these in any mirror rig without the lens hood attached.

Bruce Schultz
March 11th, 2011, 08:43 PM
Tim, we just sent Y from one cam to genlock in of the other. The whole setup was together for maybe an hour just to check it out on the Hurricane. I've been genlocking 2 F3's this way also, but I'm ordering an AJA sync generator box because I think that's a more robust solution - at least until the 4:4:4, S-Log, 3D firmware comes out, after which the genlock will be internal as I understand it. The old way still works fine though.

Curious to hear about image quality on these 105's when you get a chance to post some results.

Alister Chapman
March 12th, 2011, 04:40 PM
Tim, we are working on a solution to the Mic issue for the Hurricane-rig. There will be two options. The first is to use some Genus Nuns Knickers to close the gap between the rear baffle and camera lens. This gets the cameras closer to the mirror than with the canon hood. The second option will be an optional hood with a 140mm aperture in the rear of the hood, big enough to get the entire front of the camera into the hood. However we have found that the microphone gets in to the shot when fully when the cameras are very close to the mirror, the optimum distance appears to be close to where the cameras end up when you get them as close as they will go to the existing hood/baffle.

Fergus Kennedy
March 24th, 2011, 07:25 AM
Tim, we just sent Y from one cam to genlock in of the other. The whole setup was together for maybe an hour just to check it out on the Hurricane. I've been genlocking 2 F3's this way also, but I'm ordering an AJA sync generator box because I think that's a more robust solution - at least until the 4:4:4, S-Log, 3D firmware comes out, after which the genlock will be internal as I understand it. The old way still works fine though.

Curious to hear about image quality on these 105's when you get a chance to post some results.

Hi,
I'm just thinking of hiring two of these cameras (xf105) to play around with 3D shootings. I'm new to genlock etc, so would love to know exactly which cable I need to get to genlock- it's a BNC at one end, but what type of plug do I need at the other? I'd like to have the cable ready before I hire the cameras, and I can't find the information as to what type of connector I need. Any help would be much appreciated.

Thanks in advance

Fergus

Arnie Schlissel
March 24th, 2011, 09:01 AM
You need to connect both cameras to a single sync generator. Either the AJA Gen10 or the Black Magic sync generator will do fine.

Fergus Kennedy
March 24th, 2011, 09:15 AM
Ah, OK Arnie, thanks, that makes sense. I understood from the conversation above that you could avoid that by connecting component Y out of one camera to Genlock in on the other.

Bruce Schultz
March 24th, 2011, 10:38 AM
The XF105 has a component output and you can use it to genlock to the BNC genlock in on the other camera. Usually you have to use the included component cable to connect to the component connector and then use an RCA to BNC connector to finish the cable run to genlock in of the other camera.

All that being said, it's more robust and reliable to use an external sync generator as mentioned.

As far as timecode goes, you might have some problems there with synchronous code as the genlock input is shared by timecode on this camera. If you use external recorders like a Nano Flash 3D then you can output one of the camera's timecode to both them. Someone else with more time with these cameras might chime in to offer a synchronous timecode solution.

Fergus Kennedy
March 25th, 2011, 04:41 AM
Thanks Bruce. That's what I needed to know. I'll order in one of those adapters before I hire the cameras.

Fergus

Bruce Schultz
April 4th, 2011, 12:28 PM
Further work with the XF105 cameras on a 3D rig has forced me to ammend an earlier statement about genlocking the cameras together using component Y out to genlock in of the other. What appears to happen is that the two pictures are slightly offset vertically when you do this.

The problem disappears when you use an external sync generator or you can also make it work by using a Lanc Shepherd controller, but I have no direct experience with this and have only heard it works from others who have done it.

So the obvious answer seems to be that an external sync generator is mandatory for any 3D work with these two cameras.

As an aside, we were very impressed with how accurately these tiny cameras lined up picture-wise on the rig.

Tim Dashwood
April 4th, 2011, 02:23 PM
The phase can be adjusted in the genlock menu. Of course you won't know if the signal is in phase without outputting on screen characters.

Jesse Blanchard
July 10th, 2011, 10:18 AM
Anybody use a pair of XF305s for 3D? or a XF305 paired with a 105? I'm especially interested to see if you can get good results with these two given the 3 vs 1 sensor. Plus, if just the 105 has lens alignment, is that enough to match the 305? Can the 305 control the 105? These seem like the best 3D camera options right now but they don't seem to be getting much use.

Tim Dashwood
July 10th, 2011, 08:35 PM
The lens systems are very different so I wouldn't pair a 105 with a 305. However, it is possible to pair a 100 with a 105 or 300 with a 305 by using the component out from the 100 or 300 as the genlock source for the other camera. This works best with 60i, 50i, 30P or 25P, but is hit or miss with 24P because of the pulldown pattern.

I have a couple XF305 cameras this week for testing and will report back.

Jesse Blanchard
July 11th, 2011, 11:13 AM
Thanks Tim!

Based on the specs alone I'm considering getting a pair of 305's (or 305/300 combo). But I would feel a lot better knowing your opinion before I plunk down that much cash.

Alister Chapman
July 12th, 2011, 01:02 AM
Just consider that if you want to use the 305's on any type of mirror rig the protruding microphone and viewfinder assembly will block the view of the opposing camera restricting you to longer focal lengths, which is not what you want with 3D. Also on mirror rigs the poorer low light performance compared to an EX means you need to either use more light or put up with noisy pictures.

Jesse Blanchard
July 12th, 2011, 10:48 AM
If the mic doesn't go past the lens barrel, how does it block the other camera? I understand you won't be able to get it as close to the mirror accounting for the 45° angle. But it seems like it should be usable for most of the zoom range. Have you tried a pair and run into this specifically? Any chance removing that piece like some have done with the viewfinders for EX1 for SbS rigs?

Tim Dashwood
July 12th, 2011, 03:33 PM
On both of the XF models the microphone does extend out past the front of the lens (with lens hood detached.) This causes many issues in most beamsplitters.
Unfortunately, the microphones are not detachable.

Alister Chapman
July 21st, 2011, 01:06 AM
The XF105 is very useable with just a small amount of zoom in as the mic only protrudes by about 15mm. On the XF305 the mic sticks forwards of the lens by about 50mm and this creates big problems.

I have used these cameras on my Hurricane Rigs and it is a very real problem. I know that some people (with help from Canon) are working to find a way to remove the mic assemblies without damaging the cameras.

Genus have just completed the first prototypes of a new hood for the Hurricane Rig specifically for the XF105's.