View Full Version : 3D Lanc controllers. Which is best?
April 21st, 2010, 03:34 AM
OK I am using a couple of sony cameras without Genlock so need something to get them running together as near as possible. Every time I press record, and as I am on my own, I can't keep clapping my hands or use a clapperboard in front of the cameras.
So there are some lanc controllers out there for 3D rigs. Does anyone use them or has tried them to any degree of success?
First off is the Lanc Shepherd.
Lanc Shepherd (http://www.berezin.com/3D/Lanc/index.html)
Price is $435.
Second is the Colorcode 3D Lanc.
ColorCode 3-D Lanc (http://www.colorcode3d.dk/group.asp?group=42)
at Euro 262.
Third is the Digi Dat.
digi-dat 3D Lanc (http://digi-dat.de/produkte/index_eng.html)
at Euro 453.
The Digi Dat looks the most professional, but I like the shape of the Colocode as it can be mounted easier ( and is half the price)
But at the end of the day it's which one which works best.
April 21st, 2010, 07:21 AM
i dont think only start recording with the latest 2 lanc controllers, you get perfect
sync. I mean, maybe it can start the camcorders at the same time, but
what about image sensor reading difference?
You are always on luck.
The first one can show this difference?
We made a circuit, that show the sensor readout difference with
red, yellow and green leds. It uses the two composite output from
camcorders, and works well!
April 21st, 2010, 07:46 AM
You certainly won't get perfect sync with any of thos listed because there is no genlock. But what you should get is no more hassle with clapper boards. It would be down to the two cameras to be well made enough to stay "on the beat" for a period. certainly a few minutes should be possible before one drifts away. And with most shots only lasting seconds rather than minutes they should work well.
Of course although the signal from the lanc would go to both cameras at the same time, one could start 1/2 a frame after the other so would not lok great on moving shots. But the lanc also gives the ability to zoom in and out rather than just wide open which is the problem I currently have when using two cameras side by side.
April 21st, 2010, 07:54 AM
"You certainly won't get perfect sync with any of thos listed "
So what millisecond difference can show the 1st one?
I can get perfect sync in post with help of our circuit,
but its true i make a handclap at the beginning of every start.
And not only for few seconds!
Yes, zooming is good, now i make every shot in wide angle.
April 21st, 2010, 12:15 PM
But at the end of the day it's which one which works best.
All these devices work on the same principle, so it really depends on what quality of finish, features and costs you prefer.
They all work on the principle, that two camcorders of the same model will power up very close to one another, and because the camcorders use fairly precise crystal oscillators for master clocks that are used to generate the syncs, the cameras will maintain the initial low skew between the syncs for fairly long time. The power needs to be cycled by the controllers first, to get the initial low skew. Right after power-up, the syncs are free running, and all internal functions, including the sensor samplings are slaved to the vertical syncs. The trick is to get the syncs as close as possible together in phase in the first place. The actual activation of the record button is fairly irrelevant, because the response (recording), will be slaved to the syncs, and any full frame skews can be easily aligned in post. It is the skew within a half frame period that is causing difficulties.
I measured on my camcorders the syncs better than 500 microseconds of mismatch for the first 2 minutes after power-up, and within 2 milliseconds for about 20 minutes, after simultaneous power up. After that you simply cycle the power with the LANC controller to get the initial alignment back. These LANC controllers indicate the skew of the two syncs, so you can make the decision.
Using random power-up and clappers, in the post, I cannot align any better than to half a frame of skew. That means skew will be anywhere from zero to 21 ms (24p), 20 ms (50i), 10ms (50p), 16.7 ms (60i), 8.34 ms (60p), and I cannot do anything about it, unless I can start with near zero skew. That is why many people like to capture at 60p or perform per-field deinterlacing for 60i, and then (if needed) reconvert back to 24p, to minimize the sync phase mismatch. Of course this is still bit higher than what you can get with the LANC controller. Of course genlock is the best long term solution, but my camcorders don't have access to them.
If the temporal match is not good enough, especially when filming fast vehicle motion, sports, or even something as simple as hand waving in 3D, the result is false depth cues, vertical misalignment, headaches, complaints, etc. That has been my experience.
I just bought the ColorCode controller, because it is small, light, I can easily integrate it on my stabilizer rig, and it is inexpensive. Unlike my own controller, it indicates the sync skew, so I can decide to record or to cycle power first. It works well but looks cheap, so I painted it flat black.
April 25th, 2010, 04:39 PM
I made a video showing the modification I made to my play rig with the ColorCode 3D controller in charge. The video can be seen either here: YouTube - yt3d: ColorCode LANC Controller, Sony AVCHD and Steadicam Merlin Amateur Rig (http://www.youtube.com/watch?v=Ddb0G8upXSs) , or here: ColorCode LANC Controller, SonyAVCHD and Steadicam Merlin Amateur Rig - Red & Cyan Anaglyph on Vimeo , if you are interested to see it.
April 26th, 2010, 01:30 AM
Thanks for that clip Pavel, very informative indeed. The worry I have with the Colorcode lanc was how to attach it to the cameras, but you seem to have got over that problem without too much issue.
April 28th, 2010, 12:33 AM
I have the Colorcode controller and it works very well, the coloured LED's indicate the accuracy of the camera sync so you can tell if they have not started in sync or are starting to drift. I hope to get a pair of NX5's to use with it.
April 28th, 2010, 02:05 PM
That was great video on the Colorcode box Pavel. Have you found that the 3 minutes is consistent for in-sync usage or does it sometimes last longer?
I wonder if a box or D.I.Y. schematic exists for a circuit that simply measures the time discrepancy between two NTSC or PAL signals? I would be happy with the green/yellow/red LEDs for preset tolerance levels but I really like how the LANC Shepherd shows you in actual milliseconds.
April 28th, 2010, 03:08 PM
The time for the "in-sync" is consistent in my case, because the powerup is consistent and the clocks on each camcorder are consistent, but I am sure it would be different for different camcorders/crystals. If you can replace the crystals in the cameras with better matched pair, the in-sync time could last lot longer. I had a correspondence with Heinz Bihlmeir of this forum once, and he managed to use the clocks from one camcorder to drive another one, which would of course eliminate any subsequent sync drift all together. That would require service manual, schematics and courage to modify the cameras.
As far as the sync difference indicator, I actually did see drawings on the web in the past, but don't right at the moment remember where. I can look around. It is a fairly common function to implement with microcontrollers such as the PIC from Microchip, or Atmel. We once used the PIC chip inside our TV's and I took a development class on it. They had a development tool for it, including boad, display, compiler and libraries. Many hobyists use it and post the implementations. There seems to be a lot on it on the web, like here: PIC Projects with schematics and source code. (http://www.best-microcontroller-projects.com/pic-projects.html) . I bet someone has that exact implementation of vertical sync difference comparator with the display. Prech on this forum seems to mention they developed a 3 LED box, and Heinz probably has something as well. I can look over the internet as well, to see what I can find.
Did you mean something like this? http://www.ledametrix.com/syncshep/index.html That is not as precise as digital counter, but it will indicate sync phase error. Rob Crockett is the designer of the LANC Shepard.
May 10th, 2010, 12:06 PM
Hi Pavel, I have two sony consumer cams ( HDR-HC5's) mounted on a sliding bar with two Manfrotto 577 quick release mounts. Even though I eye them in they are out of line when I watch the footage back.
Now I need something to attach to the two small hot shoes on top of the cameras, and then a bar or something which locks them completely in place. These hotshoes are much smaller than the hotshoes on the EX3 or more pro cams, so I am struggling to find a solution.
Any thoughts would be appreciated..
May 11th, 2010, 01:18 AM
The shoes you want for those are called Sony cold shoes (so that they don't short any contacts, that are hot on the Sony camcorders). I will post couple of links: PS901 Adorama Adapter to Mount any Shoe Mount Flash on Sony DV Cameras (http://www.adorama.com/FASPCP.html) , shoe adapter sony , and SA1 Bescor SA-1 Universal Swivel Shoe Mount Adaptor, with SONY AIS Shoe Mount Included . I hope you can find some of these in the U.K.
May 11th, 2010, 06:18 AM
Well have traced a single shoe adaptor for the Sony camcorders.
Not cheap but will do the job. Now I need to source a way of joining up the two cams. I already own one of these and think one at the top as well will make a very rigid rig.
Dual Camera Bracket Mount 2 Cameras 3D Stereoscopic NEW - eBay (item 200380282030 end time May-31-10 12:25:56 PDT) (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=200380282030)
All then that is needed is a standard sized cold shoe with a longish thread that will fit in the slot of the above bracket mount.
June 28th, 2010, 08:30 AM
Well I thought I woould revive this thread to say I picked up a used Digi dat version ( branded Inition) on Ebay for half retail price. It should be with me in the next few days so I will give a report on how it works.