Understanding negative size

[Shawn Nelson]

I'm trying to understand negative size in regards to 35mm, super35 3perf vs 4perf & how all that relates to Red. Beyond the technical wizardry of it all, i'm more interested in how it relates to FOV on optics.

In my research I discover that super35mm 3perf is 24.89mmx13.87mm, which is roughly 1.77 or 16x9 aspect ratio. Additionally, I find that James Cameron uses super35mm to achieve a 2.35:1 aspect ratio spherically without going anamorphic per
ASC Article: http://www.theasc.com/magazine/dec97/titanic/pgs35/pg1.htm
Titanic Tech Specs: http://www.imdb.com/title/tt0120338/technical

So what is he doing? Is he cropping super35mm down to 2.35 to use a negative of only 24.89mm x 10.59mm?

It seems that that process is also what Lord of the Rings did
http://www.imdb.com/title/tt0120737/technical

Ultimately I'm trying to figure out if Red's 4.5k is closer to the 2.35 super35 approach or more like the old 2perf Techniscope as used by Sergio Leone in his flicks. So far it seems like it's almost s35 since Red's 4.5k is 24.18mm x 10.36mm.

Interestingly, looking forward to Mysterium-X which will be 30mmx15mm (though I have yet to read what the readable area is). But if we can get the full 30mm wide then a 2.35 ratio would yield a negative sensor size of 30mm x 12.76mm, which is roughly 20% bigger than super35. Which means all our lenses have a 20% wider FOV, and I'm pretty excited about that.

 

[David Mullen ASC]

You have to keep in mind the difference between the captured size and the area used for the final presentation format.

James Cameron starting using 4-perf Super-35 starting with "The Abyss" through "Titanic", and all of those movies were framed for cropping to 2.39 and released in anamorphic ("scope") release prints. 4-perf Super-35 is basically the old Silent Era format, aka Full Aperture, and is 1.33 : 1 unless you use a hard matte.

It's confusing to some people because early 4-perf Super-35 movies used the wider Super / Full Aperture area in order to compose a 2.39 image, but in terms of the vertical height of the negative, some early Super-35 movies used a camera gate that cropped the vertical height of 4-perf so that the image on the negative wasn't 1.33, but varied from 1.50, 1.66, etc. But soon people started just exposing the whole 4-perf area (1.33).

So yes, only a 2.39 area was used off of the original Super-35 negative in order to blow-up & squeeze the image to anamorphic.

And yes, 4.5K is Super-35 in width (approx. 24mm), not Academy in width (approx. 22mm). 2-perf Techniscope generally only exposed a 22mm wide gate, simply because there was no reason to expose Full Aperture (24mm).

If 4-perf 35mm Full Aperture is 1.33 : 1, then cutting it in half to 2-perf would create a frame that is 2.66 : 1, which is wider than needed in order to get a 2.39 image (2.35 back in the 1960's) for blow-up to scope, so they just used a gate that was 22mm wide. The resurrected 2-perf cameras like at Panavision use a Full Aperture gate though.

I'm not sure why it's an advantage to get a wider FOV with the same focal length because of the 30mm x 15mm sensor -- today when shooting Super-35, you can easily just use the next shorter focal length when you want a wider view so why do you need a larger sensor just to get wider-angle shots? The main advantage of the 30mm x 15mm sensor, besides any improvements they make in DR, will be the increased number of pixels so we can get 5K or 6K RAW. But I don't see a FOV advantage.

Have you visited this site?
http://www.widescreenmuseum.com/widescreen/lobby.htm

The 4.5K RAW area is closer to the 3-perf 35mm format than anything else, size-wise. But you can compose any aspect ratio you want within that, just as with 3-perf.

 

[Shawn Nelson]

Whoa, why would James Cameron expose 4-perf when he's going to crop all the way down to less than 3 perf? Wouldn't it make better sense to just do 3 then crop? I'm not seeing the benefit to exposing a 1.33 image and using all that extra film, granted he can afford it, it just seems pointless.

I guess I'm more excited about wider FOV since my arsenal is the five RPPs plus a 14mm, so getting wider is more interesting.

I'll go check out that site now.

Thanks!

 

[David Mullen ASC]

3-perf was proposed off and on for decades, I have articles from the 1960's and 70's proposing it, but it didn't really take off until Super-35 blow-ups were done digitally instead of in an optical printer.

Most 35mm movie cameras were built with 4-perf movements, and were switchable from having the lens centered for Full Aperture (for Super-35) or Academy Aperture (for standard 1.85 and anamorphic).

All 35mm print projection is 4-perf. Almost all 35mm projectors are limited to 4-perf, only a few labs have projectors even today with 3-perf movements.

So since most 35mm movie cameras until after 2000 were 4-perf and almost all 35mm projectors were 4-perf, people shot 4-perf for Super-35. Back then, big-budget movies still made print dailies, so it was easier to contact print and project 4-perf.

Also, many post houses and labs had trouble dealing with negative conforming using keycode & timecode when the negative was 3-perf, it took awhile to make 3-perf more foolproof for neg matching (matters less now that the neg is rarely cut because of D.I.'s). Remember that "Titanic" was made in a pre-D.I. era. The 4-perf Super-35 negative was cut, answer-printed on film, then a color-timed IP was made, and then that IP was cropped and stretched and blow-up to a 4-perf 35mm 2.39 anamorphic IN, and then release prints were made off of that IN.

Also, TV was almost entirely 4x3 standard definition back then, so people shot 4-perf Super-35 to make it easier to make the 4x3 pan & scan version for TV, even when the movie was composed for cropping to 2.39.

And even today, big efx movies prefer 4x3 Super-35 4-perf over 16x9 3-perf because they want more negative area for reframing purposes.

3-perf first caught on for TV work, where the negative was telecine transferred to video and there was no optical printer blow-up to 4-perf involved, nor any film print dailies.

Once D.I.'s became more and more commonplace, more and more people wanted to shoot in 3-perf instead of 4-perf, so labs and D.I. facilities started working out the issues for matching the 3-perf negative to the offline cut, even though at that point, the whole system of keycode, etc. was based on 4-perf. It took awhile, and it didn't help that not only were there fewer 3-perf cameras out there to rent, but most of them were booked for months at a time on TV series work.

Just four years ago, I found it hard to convince a producer to shoot in 3-perf instead of 4-perf, despite the 25% savings involved, because of horror stories from post houses and lack of options for rentals. It didn't help that with hardly any MOS 35mm cameras around are in 3-perf, like Arri-2C's, Arri-III's, Eyemos, etc. -- all of those older cameras tend to be 4-perf -- it invariably meant that some 4-perf footage would creep into the 3-perf edit, which was something for editors and post houses to deal with when conforming. And with hardly any labs having 3-perf projection, the other problem was how to screen tests and dailies.

So 3-perf has taken awhile to catch on for features, really only in the past four years. Now D.I.'s have become the norm and more producers budget for 3-perf automatically rather than 4-perf, which is a problem when you are trying to convince them to shoot in anamorphic. And now you can project your tests digitally (if you scan the footage or telecine it to HD) so the lack of 3-perf print projection is less of a hurdle.

 

[Craig Ryan]

Also, TV was almost entirely 4x3 standard definition back then, so people shot 4-perf Super-35 to make it easier to make the 4x3 pan & scan version for TV, even when the movie was composed for cropping to 2.39.

And even today, big efx movies prefer 4x3 Super-35 4-perf over 16x9 3-perf because they want more negative area for reframing purposes.

Ahhh so that's why there's way more vertical space on the Titanic VHS I have compared to the letter boxed DVD. And after all these years of explaining to the folks that "You're actually seeing more with the black bars", I was wrong. At least in some cases anyways.

 

[Andrew Walker]

Back when use to run dailies on film I would encounter a lot of Super 35 shows. They would always have some target film before the daily rolls so that the framing they decided on while shooting would still be there. Here's some frame and focus from two shows. One is Super 35 target film and the other is Vista Vision target film.

So as you can see they would still shoot the entire frame but then would crop it for the final release prints and print it in anamorphic. "The Perfect Storm" only used Vista Vision for the VFX plates, not the entire movie.

 

[Dan Hudgins]

Shoot high in the frame for Super35

Shoot high in the frame for Super35

You can see in those leaders that the "used" area in the ~1.33:1 full area for the Super35 is off center up and down, it is normal to shoot in the top part of the full area for the Super35 crop so that you do not get too much over the tops of the actor's heads area when you crop for 4:3 release.

You use Super35 rather than 'Scope for FX because 'Scope lenses do not focus close, have distortion, and do not match geometrically with the computer animation or plate shot with a flat field lens.

EDIT: See link in my post on next page about vertical offset.

 

[Stephen Williams]

Hi Dan,

It's not normal for the 2.35 to be offset vertically, I suspect they were framing for comon top 1.1.85. Not sure why it's not centered left & right & slightly bigger either. Thats why framing charts get used on 'non standard' formats.

Stephen

 

[David Mullen ASC]

This was Fincher's GG markings for the 3-perf "Panic Room", with raised 2.40 inside 3-perf Full Aperture:

 

[Shawn Nelson]

Ahhh so that's why there's way more vertical space on the Titanic VHS I have compared to the letter boxed DVD. And after all these years of explaining to the folks that "You're actually seeing more with the black bars", I was wrong. At least in some cases anyways.

Wait a second, does this mean the VHS or full screen shows the complete exposed 1.33 negative image?? So instead of cropping to be full screen it simply "un crops"?

 

[David Mullen ASC]

Wait a second, does this mean the VHS or full screen shows the complete exposed 1.33 negative image?? So instead of cropping to be full screen it simply "un crops"?

Within reason. Even when the full 4x3 4-perf Super-35 negative is exposed, it doesn't mean every bit of it is usable, free of mics and dolly tracks, etc. So some zooming in, and thus some panning & scanning, is necessary when that happens in order to make the 4x3 full-frame TV version. Also, occasionally a 2.40 composition opened up to 1.33 just looks too loose and wide dramatically, and the actor too small on TV, so if nothing important will be lost on the sides, they tighten up the shot for the 4x3 version.

The last thing is that rarely are digital efx finished out to fill 4x3 because that would entail more time and money to work on areas of the frame outside of the theatrical 2.40 frame. So the output effect is letterboxed / hard-matted onto 4-perf Super-35. It varies how much the efx shot is larger vertically than 2.40. For example, a framing chart for "Terminator 2" shows that ILM was supposed to deliver finished efx work matted to 2.00 : 1, so a little more vertically than the 2.40 theatrical frame, but less picture information than non-efx production footage.

Hence why making the 4x3 pan and scan version still took a lot of work and decision-making.

But this is nothing new (the 4x3 TV version showing more picture than the theatrical frame) -- most standard 35mm 1.85 movies are shot open-matted (unmatted) so there is more picture on the 4-perf 35mm negative than is shown when the print is properly masked off to 1.85. But on the unletterboxed 4x3 TV version, that area outside of 1.85, normally masked off, may appear. Although, again, a telecine operator is supposed to zoom in when necessary to remove any mics that dip into the shot, when making the 4x3 full-frame version for TV broadcast.

35mm anamorphic is one of the few formats where the exposed negative is barely larger than the projected area. 3-perf framed for 1.85 also uses most of the negative for the theatrical frame.

Though a non-standard practice, some standard 35mm movies framed for 1.85 have used in-camera hard mattes in the gate to mask the negative off, leaving black borders permanently in the 1.33 frame. Most of the standard 35mm movies shot by Allen Daviau, framed for 1.85 theatrical (such as "E.T.", "Empire of the Sun"), used a 1.66 hard matte in the camera gate.

Some of the earliest Super-35 movies, composed for 2.40, used hard mattes in the Super-35 gate to reduce the height of the exposed image from 1.33 down to anywhere from 1.50, 1.66, or more. But that practice soon was dropped in favor of just exposing the 4-perf 35mm Full Aperture (1.33).

Here's an example of a movie ("Full Metal Jacket") shot in standard 4-perf 35mm where the 4x3 full-frame TV version shows the excess headroom that becomes visible when the image is not being cropped to the theatrical frame:

Here's the same image cropped by me to show how it looked in theaters:

With Kubrick's later movies, it's a bit of a controversy because he insisted that his TV transfers be full frame, no letterboxing to the theatrical frame, though a movie shot with a camera hard matte (like "Barry Lyndon" and "Clockwork Orange") could be transferred to 4x3 with the hard mattes visible. So this meant that most of the versions on home video for "The Shining", "Full Metal Jacket", and "Eyes Wide Shut" were only available in 4x3 full-frame. Kubrick had said that one reason he liked that version was that it reminded him of the old 1.37 Academy format of older movies. But this has led to some people erroneously claiming that these later movies were composed to be seen in 1.33 instead of matted to widescreen. But just looking at the 4x3 TV frames, these movies were clearly composed to be cropped top & bottom. Plus all his former crew people have said that these movies were composed for widescreen theatrical (and "protected" for 4x3 TV) and edited with the widescreen croppings displayed on all the editing equipment.

 

[Dan Hudgins]

See link,

See link,

Hi Dan,

It's not normal for the 2.35 to be offset vertically, I suspect they were framing for comon top 1.1.85. Not sure why it's not centered left & right & slightly bigger either. Thats why framing charts get used on 'non standard' formats.

Stephen

Please see the vertical offsets in this link, I have seen other ground glass markings with vertical offset also for shooting 4:3 with Super35 crop,

http://www.camrock.nl/formats1.html

It only makes sense that you frame with the head in the upper half of the frame for 4:3 so you need to crop high to avoid having the eyes at the top of the Super35 crop. If you frame for "centered" crop you have way too much room over the actors heads in the 4:3 version, it would also be bad for the boom mic to be that far away.

 

[Alexander Ibrahim]

With Kubrick's later movies, it's a bit of a controversy because he insisted that his TV transfers be full frame, no letterboxing to the theatrical frame, though a movie shot with a camera hard matte ...

Jeez Dave, you are SO full of information like that. Sometimes it boggles my mind. I remember talking with you on USENET in 2001 about 2001. I had just read Cinefex 85 (April 2001) and was yapping on and you came out with the old school info from better sources, and added so much to the conversation.

I've learned a lot from you over the years.

Thanks for giving so much time to this sort of thing.