Craziest Idea to Date

[John Sprung]

Could I just fudge the normal layers of the top strip in post to dodge the special stock costs?

 

Alas, probably not. All silver halides are blue sensitive. The top sensitive layer is for blue, and it's followed by a yellow filter layer that keeps the blue that passes thru from contaminating the green and red. Going the other direction, you'd have blue light exposing the red and green layers, and getting blocked before it gets to the blue layer.

 

Edit: Well, duh! That just about kills the whole idea: The first film's filter layers would kill off any chance of getting a reasonable exposure on the second one. I shoulda thought of that. Oh, well -- back to the beam splitters....

 

 

 

 

 

-- J.S.

[Leo Anthony Vale]

Alas, probably not. All silver halides are blue sensitive. The top sensitive layer is for blue, and it's followed by a yellow filter layer that keeps the blue that passes thru from contaminating the green and red. Going the other direction, you'd have blue light exposing the red and green layers, and getting blocked before it gets to the blue layer.

 

Edit: Well, duh! That just about kills the whole idea: The first film's filter layers would kill off any chance of getting a reasonable exposure on the second one. Oh, well -- back to the beam splitters....

 

You could try a B/W neg for the front layer. That would be used for the shadows. It would desaturate them, but give more shadow detail.

 

One of the "problems" with bipack color processes was that the front ortho neg also acted as a diffusser, causing the red record to be rather soft. Since the green layer is the sharpest to the eye, the soft red record was tolerable.

[Ben Syverson]

Paul, negative film is already "HDR" compared to digital -- I say just shoot with the unmodified Mitchell and do a nice scan on your rig! :)

[Dominic Case]

You could try a B/W neg for the front layer. That would be used for the shadows. It would desaturate them, but give more shadow detail.

That's a great compromise solution, and in a way it fits with human vision in that low light levels are seen with less saturation anyway (scotopic vision). And it gets around the colour filter layer problem and the remjet problem. But it is combining two different emulsions with different curves, and it doesn't really extend the range of colour information that's recorded.

 

I think the two-perf solution is the way to go. Have you looked at the vertical offset 3D projection system, used briefly n the '80s and reintroduced by Technicolor last year? It's a 3D projection system in which the left-eye and right-eye image pairs are each a 2-perf image, one above the other. The projector pulls down int he normal 4-perf mode. The magic is in the beam-combining optic on the front which pushes the two images (polarised) out through a single lens, but offset so they are superimposed.

 

You simply need the exact reverse of this (or really, the same thing with the light going the other way), and instead of polarising filters, you need the heavy ND in one light path, behind the beam-splitter, in front of one halfof the gate.

 

As an aside, (since we've buried the flipped-over colour emulsion idea): even if it was going to work, and even if Kodak would consider reversing the order of emulsion layers in one coating run (I doubt if even James Cameron could get this), the tonal detail would be all to pot: each emulsion layer is manufactured to develop at a certain speed, given that the developer takes significantly longer to permeate to the lower layers, and is turbulated less when it gets there. Bringing the red-sensitive layer up to the top would be like bringing bottom-feeders up from the ocean floor to feed in the oxygen-rich, bright-light surface layer, and sending the surface fish down to the Deep. They wouldn't like it at all.

[Paul Bruening]

From what I'm getting here, a person could rent an old Technicolor camera, run color through two paths (intruding a little bit of ND into a path) and B&W through the top strip of the bipacked side and not have to build anything. It would be already ready already. Where is one? Maybe a museum, somewhere, is getting short on cheese and could stand to rent it out with some matching lenses. Since Technicolor is still alive, might they have one?

[Hal Smith]

From what I'm getting here, a person could rent an old Technicolor camera, run color through two paths (intruding a little bit of ND into a path) and B&W through the top strip of the bipacked side and not have to build anything. It would be already ready already. Where is one? Maybe a museum, somewhere, is getting short on cheese and could stand to rent it out with some matching lenses. Since Technicolor is still alive, might they have one?

 

The only Technicolor 3-strip camera in private hands (from Samuelson's collection) was auctioned off at Christie's in 2006. It went for around $15K if I remember correctly. The Smithsonian has at least one Technicolor camera (I've seen it on display there) but I doubt if they'd let you borrow it. :(

 

Smithsonian Technicolor Camera

[Paul Bruening]

The only Technicolor 3-strip camera in private hands (from Samuelson's collection) was auctioned off at Christie's in 2006. It went for around $15K if I remember correctly. The Smithsonian has at least one Technicolor camera (I've seen it on display there) but I doubt if they'd let you borrow it. :(

 

Smithsonian Technicolor Camera

 

Thanks for the pic. That's a biggun'. And I like bigguns'. How many were made? Are these the only two now in existence? Would Christies release their successful bidder's name?

 

It's probably a better path for me to jam these two Mitchells together and be done with it. I was thinking about others who might like to try it with something ready to rent. They'd have to go 4-perf. But, at least, it is all ready to go.

 

I still need help with my graphs. If they're wrong and this won't work on that end of the mind-spectrum, then there's no point in turning the first screw on this. I don't like building crap that doesn't work. That makes me manifestly dumb instead of only theoretically dumb.

[Paul Bruening]

C'mon, fellers. Hep me. I need hep.

 

As I was driving into town it dawned on me that it's only been in about the last 5 years or so that the scan, computing and storage improvements and costs have made this approach even worth trying. We may be the first people to even consider this thing. Please, splatter some brain on the graphs. I have to know the real limits and gains possible. Do it for the industry on both the film and digital side if not to help a friend.

 

I can imagine all of us standing on the stage at the Oscars with s**t eating grins on our mugs.

[Paul Bruening]

I think the two-perf solution is the way to go. Have you looked at the vertical offset 3D projection system, used briefly n the '80s and reintroduced by Technicolor last year? It's a 3D projection system in which the left-eye and right-eye image pairs are each a 2-perf image, one above the other. The projector pulls down int he normal 4-perf mode. The magic is in the beam-combining optic on the front which pushes the two images (polarised) out through a single lens, but offset so they are superimposed.

 

You simply need the exact reverse of this (or really, the same thing with the light going the other way), and instead of polarising filters, you need the heavy ND in one light path, behind the beam-splitter, in front of one halfof the gate.

 

Hey Dominic,

 

The only way I could make this work was the leap frog thing. Both paths have to share the same FFD. One dog legged path goofs that up. But, that was only what I could come up with. I'm all ears.

[John Sprung]

How many were made? Are these the only two now in existence?

 

IIRC, they made 27 cameras (there's a book in my office that'll have the exact number), not sure how many blimps. The ASC has one, and in the 1970's, UCLA had one (not sure, it might be the same one). Steve Gainer at the ASC would probably know more.

 

 

 

 

-- J.S.

[Paul Bruening]

As an aside, (since we've buried the flipped-over colour emulsion idea): even if it was going to work, and even if Kodak would consider reversing the order of emulsion layers in one coating run (I doubt if even James Cameron could get this), the tonal detail would be all to pot: each emulsion layer is manufactured to develop at a certain speed, given that the developer takes significantly longer to permeate to the lower layers, and is turbulated less when it gets there. Bringing the red-sensitive layer up to the top would be like bringing bottom-feeders up from the ocean floor to feed in the oxygen-rich, bright-light surface layer, and sending the surface fish down to the Deep. They wouldn't like it at all.

 

I haven't thought it out yet. But, would there be any way to use wavelengths and wave angles of light through polarization so that separated images fell at the appropriate FFD and image parsing when the bi-pack is running with both emulsions facing forward? I guess we're back to coating washing and adding a polarizing coating to both surfaces of the film. I hope no one at Kodak is reading this thread.

[Paul Bruening]

I'm piling dumb ideas on dumb ideas, here. All I can come up with for the FFD problem is some kind of mixed polarizer/micro mirror kind of thingie to both stretch and parse the light beam. It's these "thingies" that usually make the whole idea come crashing down.

[Paul Bruening]

Hey! What about polarizing the film's grains? Split the light by angle, stretch one of them for FFD. Then, the top strip of film (vertical angle grains) will catch its light and the bottom strip (horizontal angle grains) will catch its light. All I'd have to do is snap the "thingie" into my Mitchell and it's all Rock n' Roll.

[Paul Bruening]

Dang. I goofed the film's thickness. Imagine each strip as twice as thick. Oh, and the mirror angles aren't right either.

[John Sprung]

Dang. I goofed the film's thickness. Imagine each layer as twice as thick. Oh, and the mirror angles aren't right either.

 

Polarized or not, the light to the back of the bi-pack still has to get thru the front piece of film. That still means getting some special kind of stock made for this process.

 

 

 

 

-- J.S.

[Paul Bruening]

Polarized or not, the light to the back of the bi-pack still has to get thru the front piece of film. That still means getting some special kind of stock made for this process.

 

 

 

 

-- J.S.

 

http://en.wikipedia.org/wiki/Redscale

 

http://www.cinematography.com/index.php?showtopic=39645

 

I thought we were all groovy with shooting through color stock if it didn't have anti-halation coating. The bottom strip needs NDing anyway. How much would the top strip stop it down? I could go for anything between 6 and 11 stops.

 

EDIT: or is it the blue layer thing on the top strip?

[Paul Bruening]

This one popped up quick. Instead of an anti-halation coating on the top strip, what about a blue filter layer? That restores the color balance to the second strip and dodges the chunkiness of anti-halation coating.

[Paul Bruening]

Let Tiffen make the thingie and you could use it as a lens filter. If it were on a spinner ring like an ordinary polarizer the AC would make sure it oriented right and tape it down. That way, the normal diffusion that occurs through the lens would spread the light out enough to fill-in all those tiny platelet sized holes from the thingie that manifest on each strip of film with no more effect on the resolution than normally occurs with that lens.

[Paul Bruening]

Changing the grain density of each strip could manage ND. More density on the top strip and less density on the bottom strip means higher differential of stops between them. You'd pick your stocks in daylight or tungsten as a group with subgroups of density + grain size (ASA) for stop differential.

[Paul Bruening]

Hey John,

 

I wasn't trying to buck you on this bipack thing. It's just that if a solution could be found any camera from 16mm to IMAX could do cine HDRI if it can handle two strips or be slightly modified to handle two. But, I couldn't find a solution under any circumstances. So, I'm back to the two heads are better than one approach. I'm hoping I didn't get a response to my nagging for graph scrutiny because it indicates that they may be okay as they are. Though, I wouldn't mind a few people still looking them over and making their best guess and posting it.

[Leo Anthony Vale]

IIRC, they made 27 cameras (there's a book in my office that'll have the exact number), not sure how many blimps. The ASC has one, and in the 1970's, UCLA had one (not sure, it might be the same one). Steve Gainer at the ASC would probably know more.

 

Many had their innards bebuilt to 8-perf horizontal for VistaVision and Technirama.

 

 

 

 

 

Note the 2000' mag and the humongous Delrama/Technirama anamorphic prism.

 

Also Disney has a couple of three-strip cameras which were used for their sodium vapor travelling matte process. The original prisms were replaced with ones which would seperate the sodium vapor light on the yellow backing from the normal light on the foreground.

[John Sprung]

Many had their innards bebuilt to 8-perf horizontal for VistaVision and Technirama.

 

That was mainly for the two color cameras, they were single strip 8 perf pulldown with over/under separation. It was easy to just take out the prism and open up the two apertures into one. I don't see how that conversion would work on the three strip cameras. They would be a good starting place for the matte idea, though.

 

 

 

 

-- J.S.

[Paul Bruening]

We haven't even addressed what averaging two strips of film means in terms of resolution. No one bothered to manipulate grains even by category since there was nothing to replace the loss in mass. But with two strips you have twice the analog data to work with. Grains could be reduced in situ, thereby increasing resolution with no overall loss in grain mass (of course, when I say grain mass, I mean dye mass).

[John Sprung]

Lowry Digital has done something like that -- tracking objects from frame to frame and averaging grain. They showed us quite an impressive demo once. Much less grain, much more resolution.

 

 

 

 

 

-- J.S.

[Paul Bruening]

I haven't totally given up on the bi-pack, just yet; benefits of being well loved by the Muses, ya' know. Never hurts to be tight with Zeus, either.