David Mullen ASC

I think I see the mistake on Wikipedia — the Super 8 projector gate is 5.46 x 4.01mm (1 : 1.36) and someone mistyped 5.46 as 5.79 at some point. Unless there was some enlarged-width projector gate made at some point…

But this site says 5.63 x 4.22, which is 1.33. http://www.gcmstudio.com/filmspecs/filmspecs.html

My calculator says 5.79 divided by 4.01 equals 1.44389… but maybe these specs in Wikipedia are wrong or these are camera specs not projector gate specs. (EDIT: looks like Wikipedia is wrong.)

Since you're just cropping in one direction, vertically, to get to 1.78 : 1 from 1.44 : 1 -- about a quarter vertically, maybe less (you're cropping 5.79mm x 4.01mm to 5.79 x 3.25mm), it's hard to talk about crop factor since the horizontal view isn't changing. But you'll probably find yourself widening out a little to compensate because of the feeling that the vertical crop is making the shot feel tighter. Technically it's a 1.23X crop factor but only vertically, so a 40mm would be changed to a 32mm to compensate -- but then you're also widening out horizontally even though there is no horizontal crop.

Kodak Vision 3 stocks are designed to match each other very closely so the main difference is that the faster stocks are grainier than the slower stocks. The 50D is so fine-grained in 35mm that it can stand out a little compared to the next fastest stock, 200T.

Most video back then would have been something like a 2/3" 3-CCD camera like a Sony Betacam unless it was a studio show using cameras on pedestals, but that would have also been a 2/3" 3-CCD camera, but maybe an Ikegami. So you have the look of a 2/3" 3-CCD camera plus the look of interlaced-scan capture (60i in North America), 60 fields per second, usually with no shutter, so 1/60th exposure per field. Sawtooth edges on moving objects is one artifact of interlaced-scan capture. An older station might have still been using 3-tube cameras instead of 3-CCDs, where you got that problem with hot highlights trailing a moving object. And of course we're talking about pre-HD resolution, 720 x 480 pixels for NTSC (if digital.)

Yes, in a room with no other color temperature sources to worry about, you can use them in daylight mode for daylight film and tungsten mode for tungsten film. Or use them in daylight mode in a room with natural daylight and tungsten mode with a room at night with tungsten lamps in the frame. I won't get into the color reproduction issues with LEDs versus a more continuous spectrum source like a tungsten lamp. An 85 filter on a lens while shooting tungsten stock in daylight though is hardly a disadvantage, but if you really hated using a filter, you could shoot uncorrected and then time the blue cast out in post. But in terms of matching to the room's natural color temperature, bicolor lights are useful, even if you mixed daylight with daylight LEDs but shot on tungsten film with or without the correction filter.

Most cinema aspect ratios, unlike video (4:3, 16:9), have traditionally been expressed with the vertical length as "1". I don't know anyone who takes the time to convert 1 : 1.85 or 1: 2.35 / 2.39 to whatever that would be without the "1" on one side. And if one is going to talk about 1 : 1.85, it seems odd to use 4:3 then for earlier formats just from a parallelism standpoint. I don't mind using 4:3 if only talking about early cinema, or TV, or still photography -- but I were writing a textbook on cinema aspect ratios, I would stay consistent in style at least if covering more than the early days of cinema. And cinema wasn't all 4:3 even in the early days -- there were some other formats and aspect ratios though rare. 70mm Fox Grandeur for example, or Gaumont Chronochrome (which was 3-perf to save on stock.) Academy is .825" x .600", which is 1 : 1.375 -- calling that "4:3" seems like an expression to me, whereas calling it 1 : 1.37 seems more accurate. I mean, if talking about 16mm and Super 16, and you say 16mm is "4:3" then what do you say Super 16 is?

I never heard the term "Open Gate" until ARRI started using it for the Alexa, but I suppose it might have been used in Europe or by projectionists to describe a 4-perf 35mm gate with no masks in it (1.37, 1.85, anamorphic), i.e. Silent / Edison / Full Aperture / Super 35. There are standards for projection masks but in terms of making a Full Aperture gate in a movie camera, there can be slight variations since that whole area is not a projection format. You see different specs for 4-perf 35mm Super 35 from Panavision and ARRI for example, since the main purpose was the use the maximum width in order to compose a widescreen image inside Full Aperture. I don't mind if cinema decides to use different definitions for formats compared to still photography, so calling IMAX "large format" is fine because it is the largest cinema format made. It's just that by that logic, maybe you could call the Alexa 65 "large format" and then the Alexa LF "medium format" but on the other hand, these sorts of naming conventions are not enforced anyway. Remember back when Red started making sensors wider than 24mm and calling that Super 35? Then Alexa did the same thing, because there was no term for a sensor that was 26mm or 28mm wide. Now the term just covers a range.

If the Mitchell Standard mount was designed with a shorter flange depth for a non-reflexed camera, I'm not sure how the lens could be adapted for a reflex camera with a spinning mirror shutter. It's easy to go the other direction, put a lens designed for a longer flange depth on a camera with a shorter one just because all you need is a mount extension. As for adapting a Mitchell Standard to 2-perf, I would assume the same movement from a Mitchell BNCR would work, but I don't know who is selling just the 2-perf Mitchell movements plus how easily that could installed in another Mitchell. It seems you have to first start with finding the 2-perf camera and then finding the lenses for it, and accepting that perhaps your Mitchell standard mount lens won't work on it. There is a discussion here about Mitchell flange depths: https://cinematography.net/edited-pages/Flange_Focal_Distance_For_BNCR_Lenses.htm

All a silhouette means is a dark cut-out foreground subject framed against a lighter background. You have many variations to play with. There is no "silhouette police". However, I do find myself clarifying with directors what they mean by "silhouette" because sometimes they mean a backlit subject with minimal fill, which is not really a true silhouette. I've suggested we silhouette an actor but after I'm done lighting and show the director, they often say "the actor looks dark, can we put some light on them?"

I think you're confusing Full Aperture with Full Frame. Full Aperture refers to using the maximum exposable area in any film format. Full Frame refers to the 8-perf 35mm horizontal format, which is VistaVision / Technirama in cinema terms. "Large format" for the Alexa LF is a bit annoying since Full Frame isn't even Medium Format, let alone Large Format! IMAX is technically Medium Format!

Most shoots would change out the tubes to daylight balanced fluorescents since often these scenes have daytime views out the windows. But if there are minimal windows and the tubes aren’t horrible then you could use the existing tubes (after swapping out any mismatched ones) and gel any tiny windows to match that color. Smaller spaces you could afford to put LED tubes inside the housings if they fit.

Silhouettes aren’t that hard - I just asked my wife to take a picture of me with my iPhone:

By definition, a silhouette is black so I wouldn't meter it, I'd just make it as black as possible -- to help, I'd try to make the background brighter so that the shadows fall off to black faster. I'd surround the foreground with negative fill. If I were shooting film, I'd only meter the shadow if I were in a situation where I couldn't control the lighting as well and wasn't sure it would fall to black. On a film print, a medium-toned object falls to black about 4-to-5-stops under but with digital correction, if you get close enough you can crush the shadows down. I don't meter digital either so in terms of lighting and exposing a shot like in "Vertigo", I'd set the exposure for the neon sign (and if it were really too hot, I might darken it with some ND gel or netting or a neon dimmer) and balance the interior for the level of detail I wanted. Remember, you are lighting digital for the display gamma, generally unless you are going for an HDR release, you have more information in the recording than in the displayed image. I actually had a small set where I asked the art department if I could get a neon sign out the window like the one in "Vertigo" and they did it! But this is digital, you can set all the levels by monitor basically using dimmers on the lights.

Generally with a silhouette shot you aren’t fighting to get enough exposure so there usually isn’t a reason to use a high ISO on a digital camera but I wouldn’t necessarily use a very low ISO either. Anything within a range of the recommended ISO is fine. Just expose for the mood you want and don’t clip any highlights.

Here’s a silhouette shot I did for “Big Sur” - this was digital so I didn’t use my meter but if I did for film, I might spot meter the shaft in front of his face and set it about 1/2-stop or 1-stop over 18% gray.

There are all kinds of brightness levels possible for the background of a silhouette. Silhouettes against a saturated color though tend to be exposed at 18% grey or darker to preserve the color.

If the ceiling is too high, perhaps a wall spreader would work if the width is not too much to span. Or if the circle isn't too large and you aren't framing wide, perhaps a menace arm rig or two, one for each Kino side-by-side, or a goal post rig if you can frame out a combo stand on each side... but anything done safely means hiring a grip who knows what they are doing! Or use a large paper lantern, if you can get a lift to reach the ceiling to rig the cord.

Such an important part of cinema history disappearing!

There are all kinds of eye lights. You could leave the 1/4 soft edge that reflects in the screen left eye and add a dim 3/4 key that reflects in the screen right eye. You could just add a dim light under the lens that reflects in the eye. You could not worry about getting a glint in the screen right eye! The shot looks fine without it.

I didn’t ask but I recall way back that the 70mm prints of “Hamlet” were quite expensive due to the running time, but that was also because of the mag striping needed back then for sound. Maybe Apple didn’t want it… they aren’t big on physical media!

Truth is today if you had to go from a wide shot to a close-up needing more softening, you'd probably do it in post, or half in post, use a very light filter for the start and then soften the face in color-correction as you pushed in. Unless this is more of an effect of a dream sequence beginning where you want a misty look to roll in. Mitchells weren't the greatest diffusion, they tended to blur focus too much. They used a pattern of etched diamonds in a chain link fence design, sort of like the glass equivalent of a net weave. But they didn't allow enough flat areas of clear glass spots to allow sharp lines to pass through -- true diffusion is the overlay of a blurred image on a sharp image. You can see the etched chan link pattern of a Mitchell here in the bokeh, from "Psycho". A net diffusion would have a similar pattern but dark lines not white.

You could use two 6x6 filters and slide them in from opposite directions (or just slide one but the effect would appear to roll in from one side) but they would have to be graduated diffusion filters… the only ones I ever recall being made were the old Scenic Fogs from Harrison & Harrison.

Here’s a sliding Mitchell that Roy Wagner, ASC owns.