So I am interested in none other than what the title suggests. What are some ways you can keep your DSLR focused during a crane shot?

My solution is to fly the operator with the camera:
http://regina.co.nz/crane.jpg

But the more modern approach is a remote follow focus, which is a motor and gears that connect to the lens and is controlled on the ground by someone with a remote. They aren't cheap, but they do exist.

And the super low budget solution is to either keep the subject the same distance from the camera or close down the lens until almost everything is in focus.

Hey Andrew, The link wont work, but I think I've seen your crane before.


I reaaally wish I could hoist my focus puller and camera operator but I only have something similar to the Kessler Crane that I built...


As soon as I posted this I found the Varizoom LANC ...it appears to only be a few hundered dollars... But I wasn't sure if I was looking at the right stuff because it had designated equipment for certain camera companies and I wasn't sure they'd work for DSLR cameras...

Use a wide angle lens 11-16 tokina(i used), and put it in infinity.

or
f=big number

LANC controllers won't work with DSLR lenses.

As Andrew noted, the "proper" solution is wireless lens control (Preston, Bartech, Scorpio etc). These start at around $4500 for a single channel controller and motor. There are other systems that are aimed at this market but the products are not yet mature or even available (Viewfactor, Redrock).

A "budget" solution would be a hardwire controller with external motor--there have been a few made over the years but I'm not aware of any current ones that are packaged for this market, i.e. less than the cost of the camera!

Expect to see solutions arriving within the next year.

I've actually been discussing with my buddy who flys a 5D on a Pilot building a remote lens focus system based on really inexpensive parts. I have a ton of experience with R/C heli's and planes and though I'd choose a car-style transmitter (has a wheel instead of joysticks) I think a reliable, very accurate system could be built for under $500. What we were missing up til a couple of weeks ago was a way to remotely monitor the output of the camera so there could be a true "focus puller" away from the Steadicam rig. A guy emailed me about a wireless monitor he's been building mainly to give to a client to view but why not someone controlling the focus!

On a jib, matters get a lot simpler as it's a piece of cake to build a wired servo/controller and feed a wired remote monitor.

even easier IMHO would be to use a long whip on a follow focus. then you don't have electronics to deal with.

Until I get time though, the best solution has been mentioned above...set aperture in the f10-f16 range, bump the iso up a bit and light the mess out of the subject!

Well, you can view this as a challenge, but suffice to say that I have seen a number of people attempt what you are describing (wireless lens control via RC parts) over the years and most have been unsuccessful. The resolution and accuracy required for lens control is (as I understand it) much more demanding, unless you don't have an issue with the focus suddenly zapping in and out during a shot, or being unresponsive. Pulling focus on DSLR's is extremely demanding, and any kind of latency (delay) or inaccuracy is a big deal. The lens throw is so short, a little goes a long way so the resolution has to be high.

Of the folks who have vowed to bring a $1000 system to market (some who have bragged about it), none are truly available yet as a commercial product and all have increased their pricetag, often significantly. It's simply not as easy as it appears.

As I said earlier though, a wired system should be much simpler and less expensive and I'm surprised that no-one has brought one to market as it would be ideal for the crane market, and useable for Steadicam (although not desirable, but then again few things are about the setup with DSLR's).

I will take the challenge as soon as I get a bit of time. One difference on how I plan to attack the issue is my experience as a competition level heli and aerobatic pilot (r/c) There's a lot more setup and attention to detail with how i am used to dealing with servos and controllers. The fine control you speak of (and I feel from doing manual follow focus) is possible and absolutely required. In a normal upright hover, it takes between 8-10 commands per second to keep the hover steady. Imagine upside down an inch off the ground with a $3000 chunk of carbon fiber and metal buzzing at 2000rpms! Precision servos and controllers are essential.

Along with being able to Electronically setup curves for servo travel, the size of the final output gear of the servo and maybe the use of multiple gears before hitting the lens gear can greatly improve precision.

I'm not saying it will be easy by any means but I do have a different perspective from the "weekend warrior" who spent $100 at Walmart on a r/c plane or heli.

Wish me luck!

No, understood and I wasn't meaning to imply that you weren't serious about your level of participation with the RC stuff--but I've seen guys devote years and a lot of money to coming at it from that angle. I do wish you the best of luck and honestly it would be great to see someone finally succeed at cracking this nut; I hate the idea of Steadicam being limited to "shoot wide and stop down" as some of the best Steadicam shots are normal to telephoto, and this kind of forced situation can tend to cement an aesthetic into people's heads (i.e. Steadicam should always be wide). Back when the 35mm adaptors were in vogue, people struggled with the same concept and I always recommended just pulling the adaptor off and doing the Steadicam shots with the camera and supplied lens (which would deliver the deeper DOF necessary to shoot set-and-forget focus).

I think we are a very short time away from a functional revolution in focus pulling, in that the age-old techniques may finally be replaced by adaptable automation. In other words, a very skilled focus puller has always relied on years of practice in judging distance by eye, knowledgeable use of marks and measured distances, tools such as the Cine/Panatape and a highly personal ability to combine all of the above. For newcomers these days, almost all of this has been replaced by watched a monitor and racking the knob until the image appears sharp, which almost always results in a "hunting" effect. Many of the skilled guys I work with are using monitors now also, but as an addition to all of the above. The technological advancement I'm seeing on the horizon is an interface where the puller uses a touch screen displaying the camera's image to pick and chose the area of focus, and powerful software using face and shape recognition endeavor to keep that subject in focus wherever they go. Focus racks would be achieved with a multi-touch type of interaction (placing a finger on the second subject, for instance, and drawing a line between the two at a rate which would dictate the speed of the rack). This is very different from autofocus, which doesn't allow for human interaction which is absolutely critical for narrative focus-pulling. Many people are not aware just how many decisions a top focus puller will make in the course of the day, some of which are highly critical to the storytelling (at least once a day the AC will have a discussion with the director on where to place focus during a piece of coverage, when to pull to the foreground and how quickly etc).

Totally get what you're saying and I've seen a few home made systems too. The real expense in getting one of these systems to market is packaging it in an easy to use compact way. If I can make an "ugly" system, I have a few friends who are skilled industrial designers who can spruce it up and get it to a prototype phase. But that's a LONG way off!

On your other thoughts about a "semi-automatic" electronic system...
I just got a 7d (upgraded from the T2i) and one of the coolest features in still mode is the selectable focus points which are very easy to use. I think that's the basic tech behind what you describe. How to control that during video shooting is the challenge. I say let's aim high and get someone to adapt the gunsight from the Apache helicopter which tracks eye and head movement to aim the gun!! We will all look like Borg! Ha ha!

Thanks everyone for the help!

Robert, if you ever get that controler finished, let me know!