Jarin Blaschke

My tests say Summilux is the closest, although the Primos have more pleasing bokeh in my opinion (Sasaki blames Summilux’s aspheric elements for this). At least on celluloid, performance is very similar : primo is more even performance, while summilux has extra performance in the very center. Cooke S4s have chromatic aberration, splitting green/magenta at high contrast edges. Master Primes have flat highlights and less “dimensionality” on film, but could be an asset for digital formats - I haven’t done a digital comparison. I like Leitz Rs a lot, but they are visibly softer in contrast than Primos and I consider them a different look. J

As far away from the window as possible, which shrinks the source (for sharpness/hardness)and brings the beams closer to parallel.

We actually had convex mirrors made for “The Northman.” I put an 18k into it to emulate sunlight or moonlight in the studio or outside at night, so I can’t speak about bouncing the real sun. It would be much more forgiving when tracking the sun, and it would put out a pattern more wide and desirable compared to the standard 4-foot flat mirror. However there is substantial light loss with the convex. For we light snobs, the convex mirror was a very useful tool because it shrinks the source to a proper point, and the “sun” or “moon” starts looking like the real thing. Crisp, believable shadows. To shrink an 18k to the relative size of the sun, we calculated that we’d have to place it over 250 feet away. The convex mirror allowed us to place the lamp 15 feet away from the mirror, and the mirror 30-40 feet from set. You just had to be sneaky about hiding the diverging angle of the rays. Our mirrors were 1.2 meters wide and 150 or 300mm deep depending on what we needed to do. Light loss was about 3.5 stops with the 300, - you’ve been warned. -Jarin Frank: Large silks defeat the mirror of course, and unless your silk is only 4’ wide, you’re only using a fraction of it. Unless you mean multiple mirrors?

Well, I meant DolbyVision theatrical, which tops out at about 100 nits, but blacks are complete- a spooky effect in the theater when you dolly into a black doorway. For home HDR, I capped highlights at about 130 nits. More than that affronts my eyes. Unless you are watching the film outside in broad daylight in Arizona. j

This is mostly accurate. Our Primos were adapted for us, the main differences are a round aperture rather than the spiky Primo aperture, some added barrel distortion, and subtle “cat’s eye” bokeh. The field also seemed to be deeper than a typical Primo. I basically asked Dan for 70% primo, 30% Cooke Panchro series 2. Who knows what he did and how he did it. We also graded a DolbyVision version, which is THE way to see the movie.

Ok, well, the last time I shot digitally was 2017 so there are a broad spectrum of experiences. I'm shooting a commercial on 35mm in a couple weeks, of all things... There is an unprecedented rise in 65mm production as well. Maybe it's just getting more stratified and we could lose a film technician pool to feed the upper end, but for now there is a different film situation than 5 years ago. Jarin

It sounds like a test is really necessary! I will venture to say that it will be more sensitive to red than a tungsten stock, as an educated guess, perhaps similar to 5219's sensitivity to red... J

*cough* "The Northman *cough, cough...* : )

Thanks. Without divulging much of the film or its aesthetic techniques, I’ll probably post my findings within the next week regarding the portfolio of Kodak film stocks, how I rated them, etc, within the next week or so. jarin

We wrap principal photography tomorrow! For a 87 day shoot, only two days were lost due to false positives in our testing regimen, otherwise no covid hiccups. I was tested three times per week. Rob has a very long post production process ahead! J

Yeah, right nowI'm doing exterior moonlit nights now with only 1 1/2 stops difference between the "moonlight" and the ambient "starlight." I also get costumes and sets to make nothing lighter in tone than the average skintone of the actors. Otherwise, you can't set your fill to anything consistent. You really have to compress tones for night work! J

Yeah, I also light these with the same basic technique as day interiors, but with a little cyan+desaturation and a much lower fill/key ratio. On film, I'll put the fill light at -4.5 while the direct moonlight plays between -2.5 and -3. It looks terrible to the eye on set but falls into place on the film. You can put things further down the scale on the Alexa, the equivalent of -4.5 is probably -5 to -5.5 on the Alexa. There's a new additional trick I came up with for moonlight. If it's any success I can share next year.

Thanks! We've been very fortunate to not shut down from COVID (knock, knock). A couple more months to go.... Jarin

Well, since film naturally sees blue more readily, you could argue that a tungsten stock is a more “native” stock: the longer the wavelength, the more sensitization you need. I would hypothesize that a daylight stock needs extra sensitization for red, and to a lesser extent, green, rather than the other way around. In this theory, a tungsten stock naturally “sees” more blue than a daylight stock, and overexposes blue in daylight conditions. A daylight stock has to work harder To boost red and green to balance the color for a given speed. The fact that 5207 (250D) is a grainier and softer stock than 5213 (200T) might be evidence of this. ? jarin

Sorry - how do murder rates relate to Gone With the Wind? J