What camera should I use?

[Wiliam Cardoza]

Hey Tyler, riddle explained here... from Jonathan Flanagan's "head's up everyone" the other day... Doug Milsome was the focus puller on "Barry Lyndon", he explains racking over the BNC with the Zeiss .07, about half way through the interview or so. Pretty interesting interview.

 

‘Barry Lyndon’ Q and A with Doug Milsome BSC ASC

 

I guess anything is better then nothing.

So riddle me this one. It was a non reflex camera at that point, so how did they check the composition and focus? by removing the lens, moving it over to a viewfinder and back again? I don't understand how the mathmetics of that work.

[Tyler Purcell]

Yea I saw that interview, but I don't understand how the parallax viewfinder could be in any way shape or form, similar in focus to the film plane.

[Stuart Brereton]

Here's another picture, this time of Kubrick's actual camera, with the parallax finder clearly visible.

 

https://skiffleboom.wordpress.com/2013/06/29/the-complete-stanley-kubrick-exhibit-at-lacma-skiffleboom-com/kubricks-barry-lyndon-mitchell-bnc-camera/

[Tyler Purcell]

Yea, Yea, Yea, but that camera was prior to understanding how to make a quiet pull down mechanism. The Mitchell guys were amazing engineers, but the amount of precision necessary to make a parallax viewfinder system even remotely close to what the camera was actually capturing, is nearly impossible. Even by todays technology, one would be hard pressed to make that work.

 

Which I guess is my question... how did they make it work technically. Did they build a special parallax viewfinder which could focus the lens properly? The tolerances were quite amazing, for again... a company who couldn't figure out how to make their cameras quieter... which we take for granted today.

[Satsuki Murashige]

It was a non reflex camera at that point, so how did they check the composition and focus? by removing the lens, moving it over to a viewfinder and back again?

 

Straight from the horse's mouth, posted earlier upthread: http://www.cinematography.com/index.php?showtopic=71801&view=&hl=&fromsearch=1

 

Edit* I see William beat me to it. Doug Milsome explains it all in the interview. The Mitchell BNC has a rack-over viewfinding system where you can you view the image through the taking lens with parallax error. You can normally check focus this way, but not with the special f/0.7 Zeiss lenses because the flange distance of the lens was modified to an extremely tight tolerance and would only be correct on film.

 

You then rack the lens over into the shooting position and use a separate viewfinder that has frame lines corrected for parallax, similar to a Leica M stills camera viewfinder. As Milsome explains, this method was imperfect but 'more or less' close enough to compose from.

 

To set focus, Milsome explains that he personally scaled the lenses in prep and scribed his own marks on the lens. He says this was the easy part, he could work with a tape measure as normal - the difficult part was ensuring the flange depth stayed within tolerance when the lens was set into shooting position. He says they built a special tightening mount and that he then manually checked the flange with a depth gauge every time the lens was seated. He, Alcott, and Kubrick had no idea if the focus would be good until dailies the next day, principally because the flange depth of the system was so sensitive.

 

Milsome explains it all in great detail in the interview, it really is worth watching in it's entirety.

[Stuart Brereton]

Yea, Yea, Yea, but that camera was prior to understanding how to make a quiet pull down mechanism. The Mitchell guys were amazing engineers, but the amount of precision necessary to make a parallax viewfinder system even remotely close to what the camera was actually capturing, is nearly impossible. Even by todays technology, one would be hard pressed to make that work.

 

Which I guess is my question... how did they make it work technically. Did they build a special parallax viewfinder which could focus the lens properly? The tolerances were quite amazing, for again... a company who couldn't figure out how to make their cameras quieter... which we take for granted today.

You don't check focus through a parallax finder, it's for framing only.

[Tyler Purcell]

Then how do you check focus?

[Stuart Brereton]

You watch dailies the next day.

[Satsuki Murashige]

Then how do you check focus?

With a tape measure and accurate focus marks on the lens.

[Tyler Purcell]

Sure with regular glass, with stuff made specifically to fit motion picture cameras. However, with this glass, stuff that was never designed to work with motion picture cameras... I mean it astounds me they even got an image close to focus, let alone in focus most of the time. With a spinning mirror reflex and laser based focus tools, this becomes a bit easier. Back then, a tape measure may not have been good enough for something where the position of the lens was a 1/8th of a turn from working or not. Its quite remarkable and musta used A LOT of film to test. I gather they must have had same day dailies AND must have done a bunch of more scientific tests before figuring out how it was going to work.

[Stuart Brereton]

The lenses were rehoused and engineered to give 2 full turns from Infinity to minimum. Focus Puller Doug Milsome says in the interview that they never saw dailies until the next day. Although the f0.7 lenses were highly unusual, it was historically normal to not be able to see focus. Reflex viewing was virtually unheard of before the early 1960s

[David Mullen ASC]

This is the way Hollywood movies were made from decades -- the parallel side finder was just for framing, and even then due to parallax on wide angle lenses, the operator had to rack over to see through the actual taking lens, frame the shot, and then rack back over for the take and operate even when part of the subject was now cut off on one side of the frame. Not easy, especially on an Orson Welles movie with a lot of camera movement.

 

The focus puller relied on tape measurements and the lens markings and at the end of the take they would freeze and sometimes rack back over to check focus through the actual lens.

 

But even today if a film camera is on a Steadicam or remote crane, no one is looking through the viewfinder and you are relying on the focus puller because you can't judge critical focus on a video tap signal.

 

Reflex cameras, other than Arri-2C's for handheld MOS work, or the Eclair CM3 by the mid-1950's like on "Touch of Evil", also MOS, weren't used in Hollywood films until the Mitchell BNCR in the late 1950's and then the Panavision PSR came along in 1967.

[Tyler Purcell]

Explain this process of "rack over" if you can. In my mind, I think of something mechanical which moves the lens between the viewfinder and film gate.

 

I really want to borrow a Mitchell BNC to learn more about this non-reflex way of shooting. It would be a lot of fun to learn this and share it with my students.

[David Mullen ASC]

More like shifts the body sideways so that the camera w/ viewfinder either lines up behind the taking lens or a separate smaller lens parallel to the taking lens (incorrect, sorry, the body and viewfinder does move over though but the viewfinder doesn't slide behind the lens of the sidefinder... ) See

▶ 3:24

Jump to 2:50

[Tyler Purcell]

Ohh brilliant, what a genius device! I never understood how that worked, but now I do! :)

 

Thanks for posting David!

 

Ohh and if you know of anyone with a BNC that's just sitting around, I really wanna shoot some stuff with one during my class. I think the kids would really get a kick out of assembling such a massive kit and learning how difficult it was to shoot back then.

[Satsuki Murashige]

Sure with regular glass, with stuff made specifically to fit motion picture cameras. However, with this glass, stuff that was never designed to work with motion picture cameras... I mean it astounds me they even got an image close to focus, let alone in focus most of the time. With a spinning mirror reflex and laser based focus tools, this becomes a bit easier. Back then, a tape measure may not have been good enough for something where the position of the lens was a 1/8th of a turn from working or not. Its quite remarkable and musta used A LOT of film to test. I gather they must have had same day dailies AND must have done a bunch of more scientific tests before figuring out how it was going to work.

Milsome also mentioned that minimum focus was something like 5'. But that even at that distance on the 50mm, the depth of field was razor thin. He kept focus by setting up a closed circuit tv camera with grid lines 90 degrees to the action so he was looking at the actor in profile. He placed his monitor above the follow focus wheel and got the actor to give him some marks, which he placed on the monitor grid lines. He could then easily see when the actor would lean in or back and by exactly how much.

 

Again, he says the hard part was making sure the flange was set correctly. He found actually pulling focus challenging but not impossibly difficult. Until Kubrick decided that they should add dolly moves since the dailies were turning out so well...

[Stuart Brereton]

He kept focus by setting up a closed circuit tv camera with grid lines 90 degrees to the action so he was looking at the actor in profile. He placed his monitor above the follow focus wheel and got the actor to give him some marks, which he placed on the monitor grid lines. He could then easily see when the actor would lean in or back and by exactly how much.

 

 

I worked with his son Mark, who was a focus puller when I was a 2nd AC. He used the exact same system on difficult shots.

[Satsuki Murashige]

I worked with his son Mark, who was a focus puller when I was a 2nd AC. He used the exact same system on difficult shots.

Wow, cool! It's really a brilliant solution. Much easier to do nowadays with wireless follow focus systems.

[connor denning]

They left the shutter fixed at the maximum opening possible for a Mitchell BNC, which I believe is 200 degrees like a Panaflex. However that only gains you something like 1/6th of a stop more exposure over a standard 180 degree shutter.

maybe I'm reading it to this wrong but that would explain why (along with what the scene called for) if you have a longer shutter speed, you're going to have more motion blur, so moving slower to hid that makes sense.

[David Mullen ASC]

200 degrees isn't going to make enough of a difference to the motion to worry about. Many movies shot on Panavision cameras opted for 200 degrees, including the "Rocky" movies shot by Bill Butler.

[Brian Drysdale]

Ohh brilliant, what a genius device! I never understood how that worked, but now I do! :)

 

Thanks for posting David!

 

Ohh and if you know of anyone with a BNC that's just sitting around, I really wanna shoot some stuff with one during my class. I think the kids would really get a kick out of assembling such a massive kit and learning how difficult it was to shoot back then.

 

If you don't want to move around a 100lb camera, you can use a non reflex Bolex. Although it doesn't have a rack over, they have a separate non reflex ground glass V/F for focusing the lens in the top turret position (sort of poor man's rack over). You also have the side viewfinder with the parallex adjustment..

[Mark Dunn]

I don't expect there are many BNCs sitting around. Only about 365 were ever made and they don't wear out. I believe they cost $5,000 in the 1930s- no economies of scale making only 10 a year. Aardman (the people who make Wallace and Gromit) famously owned 10% of them.

These folks have one

http://owyheesound.com/mitchell-bncr.php

but it has been reflexed- I don't know if that meant the removal of the rackover; it certainly wouldn't be necessary anymore. It still has the operator's eyepiece but the t-handle seems to have gone.

Edited June 12, 2016 by Mark Dunn

[Tyler Purcell]

Well, how many students get the opportunity to work with such a famous camera in an old-fashion way of shooting? It gives so much more respect for the filmmakers of the period. I'd like them to assemble the camera as a group and perhaps shoot something with it. It's called hands on history, which is very important when educating minors. You can show them a picture, but when they're forced to use it, then they'll understand.

 

In my course, the older equipment shows up later in the semester, so they will already have experience with smaller/newer cameras. When they see the boxes show up full of BNC, they'll freak out.

 

I see BNC's for sale on ebay all the time, so there are many out there, I just don't think spending $7k on something I'm not going to be shooting with, is wise. But hey, if I have to buy something to teach the way I want, I'll have to do it! :)

 

Ohh and the kids will have some experience with the BOLEX, but my class will be using SR's and LTR's for most of their shooting.

[Stuart Brereton]

These guys have a selection of Mitchells for sale, and they're in Panorama City. Maybe they'd let you borrow one for a class.

 

http://www.cinemagear.com/default.html

[Tyler Purcell]

Ohh yea thanks Stuart. I should go up there and have a chat with them. :)