David Mullen ASC

Most locations ban mineral oil smoke as does SAG for actors, I believe.

Haze from a machine using a water-glycol fluid is one of the few safe things for actors and crew to breathe indoors. If you can see it hanging in the air, you have to research what it does in your lungs. Considering one can rent a small hazer for parties, they can't be that expensive. There are also some "smoke in a can" aerosols but they don't last long and are more for still photo shoots, not for an all-day film shoot where in the long term, a haze machine is more cost-effective. https://www.amazon.com/can-fog-haze-effects-photography/dp/B09XDCY71F/ref=asc_df_B08N5BTTX7?mcid=9ba5e3b96f2b3047aa6f41006128e1b7&tag=hyprod-20&linkCode=df0&hvadid=693360658756&hvpos=&hvnetw=g&hvrand=9954959439392839413&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9061105&hvtargid=pla-1188167900071&th=1

There have been some TV shows that have shot 2X anamorphic on a digital camera for 16:9 HD. "Fear of the Walking Dead" comes to mind. Not the same thing I know. I think you can imagine the issue of 2X on 3-perf, that it's just a smaller negative area compared to 4-perf 35mm anamorphic so you'd expect a bit more grain and softness, and a bit more depth of field because you'd be using slightly shorter focal lengths.

Yes, using expanded C-series anamorphics on an Alexa 65 would definitely be for the look, the aesthetics. Also keep in mind that with a 2X anamorphic being used for a 2.40 : 1 release, only a 1.20 : 1 area of the sensor ends up being used so the C-series would not have to fill the 2.12 : 1 65mm sensor end-to-end. You only need 30.7mm of the 54.12mm sensor to be covered (and that only uses 3.7K out of the 6.5K sensor width -- which is still bigger than using an XT with a 2X anamorphic lens for 2.40 : 1, which is only about 2.6K.)

I suppose that makes sense if you really want the look of a 2X squeeze but without all the squeezing, but that would be a specially-made rear anamorphic. The trouble with cancelling the squeeze altogether with a 2X rear anamorphic I think becomes that you've essentially desqueezed the image into a very widescreen one, so what are you capturing? One cropped back down on the sides to the sensor dimensions? One that retains more width but has fall-off along the top & bottom, sort of a fuzzy letterbox? I guess it all depends on the image projection.

I don't about "Tár" but maybe they used an expander to fill the larger sensor on the shorter focal lengths. "Nightmare Alley" was 1.85 : 1 so it's possible that the Signature Primes filled that width without going all the way out to the ends of the 65mm 2.12 : 1 sensor (though that would still require the lens reach out to 47mm of the 54mm width, for a lens designed for a 36mm width...)

If the front anamorphic is 2X and the rear anamorphic is 2X, that's a 4X squeeze! Quite a widescreen image once unsqueezed... even if your sensor area used is a square, that's a 4.00 : 1 image once unsqueezed... A rear anamorphic doesn't add much of a look per se.

Yes and no... the 4-perf 35mm Silent / Full Aperture was invented by Dickson & Edison. "Super-35" sort of refers to a process of using the Full Aperture width to expose widescreen images. Typically the entire height of Full Aperture is also exposed but the early Super-35 / Super Techniscope movies sometimes had some masks to reduce the height. I believe both "Greystoke" and "Terminator 2" had a camera mask that exposed a less-tall image. This framing chart from Otto Nemenz Cameras seems to indicate that a 1.50 : 1 camera mask was used ("full camera aperture 24.00mm x 15.623mm"). I heard that "Greystoke" also used a 1.50 : 1 mask.

The 4-perf 35mm "full aperture" / silent / Super area is 1.33 : 1. The 3-perf 35mm full aperture is 1.78 : 1 more or less. So if you compose for 2.40 using spherical lenses, some cropping is involved in post whether you use 4-perf or 3-perf, you just waste less negative with 3-perf. As to why Super-35 would be preferred over using anamorphic lenses, that's because spherical lenses are preferred for various reasons. Even before Super-35 was invented, visual effects people in the 1970s were using 8-perf 35mm VistaVision or 5-perf 65mm, both spherical, for visual effects elements for movies shot in 35mm anamorphic -- mostly for the larger negative to offset the grain build-up from optical printing dupe elements, but also to avoid shooting miniatures and matte paintings with anamorphic lenses. Super-35 was sort of a revival of a concept from the 1950s called SuperScope; when John Alcott was about to shoot "Greystoke" (1984) and was told that a 70mm release was desired, plus 35mm scope prints, he didn't want to shoot the movie using anamorphic lenses; he felt that most were too slow for the low-light work or the fast anamorphic lenses were too soft for the low-light work, etc. Anamorphic lenses also tended to be bulkier, heavier, and focused less closely. Plus most of the anamorphic lenses came from Panavision and he was an ARRI shooter who preferred Zeiss lenses. He didn't remember this old 50s format called SuperScope, which involved using 4-perf 35mm and cropping for 2:1 or 2.35 : 1 and then optically blowing this up to CinemaScope (4-perf 35mm anamorphic). But he remembered the old half-frame 2-perf 35mm format called Techniscope from the 60s... so he shot 4-perf 35mm, using the full aperture / silent width of the negative, composed for cropping to 2.40 : 1, and called it Super Techniscope. Later the process was renamed Super-35.

Before digital intermediates became the norm for post by the mid-2000s, 35mm feature films were finished photochemically. 35mm print projection was (and still is) always 4-perf, either then masked top & bottom by the projector to 1.85 ("flat") or an anamorphic lens was used to unsqueeze an anamorphic image to 2.40 ("scope"). So the easiest way to get a print was to shoot in 4-perf 35mm, whether spherical or anamorphic, composed for the 1.85 spherical or 2.40 anamorphic projection area, and make contact print every generation. If you shot 4-perf Super-35 -- or 3-perf -- you had to go through an optical printer step to convert it to 4-perf to create a dupe negative that could be contact printed. Once digital intermediates became the norm, then the shooting format didn't have to be the same as the projection format because the negative was scanned and could be recorded back out in any film format desired. 3-perf photography could be recorded back out to 4-perf. Cropping 4-perf 35mm to 1.85 goes back to the early 1950s as a simple & cheap way of getting widescreen into the theaters; it started happening within a year of the introduction of CinemaScope (anamorphic 35mm). 4-perf Super-35 had to go through an optical printer step because it used the soundtrack area on the left side of the negative for picture information so it couldn't just be contact-printed with an optical soundtrack added.

I will shade the dome from a backlight but otherwise the light on the key side is the key + fill so it's better to let them both hit the dome. You can shade the key if you want to meter the fill separately.

You could try late dusk-for-night for the wide shots and then a very soft, dim ambient bounce light at night for tighter shots, matching the level of blue from the dusk light. Otherwise, if you don't want any fill, I'd try hazing up the field (I'd probably do that anyway) because that will create more ambient fill and separation of the foreground. Yes, your base ISO will be driven by the brightness of the flashlight so get a bright one. You just have to decide if you can live with a daylight LED one or need to find a bright tungsten halogen type.

Sure... though it's a misuse of the term -- the LF camera is really FF35 and the 65mm camera is large for a cinema camera but technically medium format in still camera terms. "Large Format" really means 4x5 and 8x10 still formats and larger.

IPhones also have a medium focal length lens (2X)... to me, the main visual effect is a very deep focus unless you use "portrait" (stills) or "cinematic" (video) modes to simulate a shallower focus, so try shooting at very deep stops. The other effects are over-sharpening and clippy highlights.

I heard that the smaller ARRI 65 will still use the old sensor -- the problem with the new Alexa 35 sensor is that the outputs run along the top & bottom, so it can't be turned 90 degrees and two stitched to create an LF or three to create a 65 version.