M Joel W

Just curious if anyone has any BTS or diagrams from larger shows (I remember Deakins had some but not sure they're still available – would love a link to them if they are)? Curious to see anything from Kaminski, Elswitt, Richardson, or any forum members working on larger shows!

David, just out of curiosity. Let's say you have someone standing in a field with mid to late day (directional) sun. With the sun/key hitting their face frontally (maybe softened by a net), how would you set the aperture relative to the sun/key? With the sun/key hitting half their face and the other side maybe 2-3 stops under – how would you set the aperture relative to key? If they are fully backlit so now the bounce light is the key and the sun is the backlight – how would you set the aperture relative to the light on their face? For a dark underexposed scene in an apartment with diffuse light, like someone in bed at night, how dark would you tend to underexpose a face at most before worrying about losing detail? Also what format are you shooting these days? Mostly 3.2K Alexa? Do you ever set the ISO below 800? I think the tendency with young DPs is to underexpose the Alexa a bit. Also – what kind of sources are DPs using on something like Arrival or Her? It feels like a lot of soft sources, more variation in color temperature, it looks very natural and a bit underexposed. But are these shows using natural light and augmenting with LEDs or rebuilding a natural look with large book lights? On a no budget show how could I best emulate this? Lots of dim LED fixtures to complement natural soft light?

I was using a variable ND filter with a lens that had a rotating front element. I noticed that as I focused, the amount of polarization changed. (As if rotating the lens with a polarizer.) But most of my lenses don't have rotating front elements. I had a few questions. If shooting with a polarizer, and rotating the lens to reduce the amount of polarization to the least possible, is that close to the same as no polarization? Variable NDs are essentially two polarizers on top of each other, right? So does this mean there is always some significant amount of polarization? Would there be a way to take a high quality variable ND like this: https://www.simmodlens.com/product-page/simmod-77mm-variable-neutral-density-0-4-0-8-filter And somehow rotate not just the front piece of glass, but independently rotate the entire filter to reduce the effects of polarization for each shot? Or at least try to keep them consistent? The big issue I have with variable ND filters is the variable amount of polarization. But if there a way to reduce it to a minimum consistently, I might use one more. Is there a way to do this just by adding a polarizing filter with the glass removed before the variable ND filter, and then rotating the variable ND filter to the correct strength, then rotating the empty polarizing filter to reduce polarization? Thanks.

Still looking for one of these. Not looking for the 67mm version if I can help it because I want to try the 72mm clamp on version with other lenses too. Thanks.

My bad re: the OLPF, it looks like the Ursa 12K lacks one. Regardless, I think it's an issue with the camera more than with the lenses. I don't know for sure, but try the same lens on both cameras and see.

I don't have a 12K or Phase One – I've never used either – but the pixel arrangement on the 12K is not a traditional Bayer Pattern and is designed for scalability rather than maximum resolution/sharpness. Which is to say, it's not as sharp as a 12K Bayer video camera would be, and that would be most evident at a per-pixel level. But a given 12K Bayer video camera would likely also not be as sharp as an 80MP still camera anyway; I've found video cameras have much stronger optical low pass filters (to avoid moire – while a Phase One back I suspect lacks one to maximize sharpness) than video cameras do. I remember the Panasonic S1H has an OLPF where the S1 doesn't, for instance. Try the same lenses on both. I don't think your issue is primarily optical. There are exceptions for macro work (and I think the Sigma zooms might suffer there, particularly the 18-35mm), but in my experience once you're stopped down to t5.6 or whatever any modern lens is fairly sharp. The Venice 2 or A1 (poor man's Venice 2?) and either Sigma Art or Tokina Vista (or Otus if you're not thrown off by the mechanics) would probably be worth renting and trying. I'm pretty sure your issue is with the cameras, not the lenses. I've heard the Otus lenses are still the sharpest thing out there (if you can't afford Signature Primes, which I'll probably never even see in my lifetime). Anyway, long story short: the issue is almost certainly with the camera, not your lenses. But the sharpest affordable lens would probably be Otus or a high end macro lens depending on the subject.

I'd guess they're rehoused Rokinon/Samyang lenses with polished front elements to remove coatings. The focal lengths match if you round 14mm to 14.5mm and 24mm to 25mm. The T stops are similar too.

Panasonic S1 with an f0.95 lens, just because that’s what I own. 6K mode, so per-pixel, which helps with low light I believe. I'd like to try an A7S3, I bet the result would be similar though. Fwiw I’ve been experimenting with using the red channel for luminance for day for night. (Plugging in, after a LUT or in a color managed workspace, the red channel for luminosity.) With a bit of grading, it can get pretty close to the effect in Nope pretty quickly. Doesn’t always work. You need a sharp lens because blue/purple CA at high contrast edges (trees against the sky, for instance) gets more saturated and thus deeper black than they sky itself, resulting in dark halos around leaves.

I did end up with flares at the edges of the film stock but as I wasn't shooting S16, they weren't much of a problem. But the first nine or ten seconds of film was flashed or totally white when I loaded in subdued daylight (the flares at the edge only lasted another half second or second maybe but in S16 likely more). Given the relatively short duration of a daylight spool, I see why there's the recommendation...

https://physics.stackexchange.com/questions/111006/how-does-light-bend-around-my-finger-tip I've noticed while comparing blue screen composites with real images that a foreground object will distort the background and also make the background appear more in focus than it otherwise is because the foreground is working as an aperture somehow. Edit: except it seems to be the foreground that is acting strange for you? Try it at other apertures. I think it's optical but am not sure.

never mind, double post

Night for night (no grade beyond a LUT): https://vimeo.com/749452544

*pretty well for me as a viewer, obviously I haven't got the resources to even attempt anything similar

The main difference I've noticed is that moonlight has a warmer color temperature. I compared day for night and night for night with the fastest equipment I could get my hands on. The main difference isn't the shape of the light, it's that moonlight is warmer. The cinematic convention is for a darker sky. (With night for night, the sky is typically black.) The above process seems convoluted in the extreme, but at the same time – it worked pretty well for me.

That makes sense. I was looking at handheld 500T footage and all I could see was more grain texture, but even that is a difference. That's a great kit. I don't have access to quite as nice a set up so I'm content with 2K scans unless someone else wants to pay for it. The reason I posted was figuring out what format to shoot in. But removing my own choices from this (which are based on time and budget) I have noticed that I have a preference for a certain level of sharpness from the entire system. These ultra-sharp large format digital cameras with wide open vintage lenses on them look wrong to me. Too much fine detail, and a mushy image otherwise. Meanwhile Good Time or Suspiria or Her or the Lighthouse I think look great by matching older (but not too soft) lenses to a moderately soft imager. U16s look great with S16 film. I think Greig Fraser has found great techniques to take the edge off of large format high res digital, but they're also expensive and convoluted for the average person to employ. Doesn't really matter, just something I noticed while researching what format to shoot in. As for diffusion, I have some Classic Soft filters. I'm thinking 2.8K ArriRAW, Mk1 standard speeds, 800 ISO but underexposed and pushed a half stop, no post halation or grain, and Classic Soft filters. Wide open for night, t5.6 outside. I feel like the diffusion filters and underexposure will add some grain and halation without having to do a "film look" pass in post.

I agree in principle but not in practice; if I had unlimited resources I'd prefer the 4K scan – but the extra cost at the lab combined with the extra cost in post offers diminishing returns that imo could be afforded to better things. I think streaming benefits from 4K far more than IMAX projection – and yet I'm not sure it was worth it for Netflix to ask for 4K vfx so easy on, the quality of work suffered because of additional render time. The graininess isn't also inherently better looking at 4K, even that is subjective; Bill Pope, for instance, preferred the look for 2K scans to 4K scans because it had a slight "noise reduction" effect, even in S35. In the abstract, though, I like the idea of a 4K S16 scan. I do wish I'd bought a set of B speeds back when it was under $20k. ? There's a separate discussion I think about taking the "edge" off digital footage. Greig Fraser seems to be all about this – the detuned lenses for Mandalorian, the film out for Dune. And that's despite the Alexa LF having a soft look per-pixel compared with Venice or Red. I don't like how Men or Army of the Dead look and I think the tendency to shoot high res with wide open mushy lenses just makes things look like detailed mush. I prefer the look of Suspiria or Her balancing a softer acquisition format with slightly sharper lenses (still Cookes and K35s, so not ultra-sharp) to any of the above. Zeiss lenses and film are both known for (whether accurately or not) lower high frequency detail with more acutance. So, as I wrote before, sort of the opposite MTF curve from what you'd expect the system on Men to have. What I'm shooting with is a bit ridiculous to discuss when I'm just making expensive home movies – although the feedback here has been helpful and I'll probably go with ArriRAW. The more interesting discussion is what people are doing at the high end to take the "edge" off increasingly sharp digital formats and whether they should have to. I remember I was blown away by the Alexa Classic but underwhelmed by the LF. When I go back and look at the Classic footage, it still feels more "organic" to me. I like Greig Fraser's approach to softening the LF but feel like it's a crutch. And he's softening the image despite the LF having a softer image than Venice or current Red cameras. (Consumer cameras are even sharper.) Dune, Mandalorian, Batman, Men, Army of the Dead, etc. all rely on something very expensive (film out or bespoke detuned/unaffordable lenses) just to take the edge off of something that's also very expensive (a Venice or high end Red or Alexa LF). I can't afford any of this but it begs the question: does it even look better? Or is someone paying a lot of money to hit a spec and then someone else paying a lot more money to tame the aesthetic faults of that spec?

Thanks – right now it's between 2.8k raw with Mk1 standard speeds and S16. Why scan S16 at 4K? The only time I worked with a 4K S16 scan it was 500T and it didn't have much more detail than a 1080p scan, however it seemed to have more accutance from the sharper grain structure.

That makes sense. I never understood Arri's "pixel math" justifications for what seemed like arbitrary resolutions, but they're actually pretty simple. 2880 is 1920*1.5 so I'm guessing the sensor is designed foremost for 1080p. 2868 is 2048*1.4 so that's why they crop slightly for 2K. 3.2k is 3840/1.2 so that's why they chose 3.2K for the XT. 3.4k is presumably just the whole sensor (but very close to 4096/1.2) I still don't like the "over-sharp" digital look and it's a mix of contrast and resolution I think. Why is it that something like the new Doctor Strange (Red), Men (Venice), Tusk (ArriRAW), Army of the Dead (Red), etc. all have a "harsh" look to me whereas something like Her (ArriRAW) or Knives Out or even something like It Follows (shot on 2K ProRes) has a more organic look? I feel like with Army of the Dead or Men the DPs are shooting vintage glass wide open to compensate for an overly sharp sensor but you lack micro contrast and still have a lot of high frequency detail. Whereas 2K early Alexa projects (In Time, etc.) with Master Primes had some built in low frequency sharpening I think (which Alexa 2k ProRes pretty clearly has) but not too much fine detail. But also a lower contrast in general. I feel like if you drew an MTF curve for the entirety of the system it would look very different for these different projects.

Thanks – I think I'm just going to transcode 2.8K ArriRAW to ProRes and work with that. I'm overthinking this.

This is amazing. I was speculating on how they achieved this effect and I guessed something entirely different and entirely wrong. Men had some night skies added in post that I wasn't wild about (fantastic vfx overall however) and Northman had night exteriors that had great depth too. But what they did here totally mystified me and I would never have guessed this was how it was achieved.

I think the early generation Alexas output 1080p via SDI. I don't know if the Classic started with the ability to shoot 2K internally, but it was added pretty early on if not. 2K ProRes is downscaled from slightly fewer pixels than 2880X1620, which is what 1080p ProRes is downsampled from. I'm not sure why. This isn't a very serious project so it doesn't really matter. But I was researching it and kept finding I actually liked the look of 2K acquisition better in many cases. I think I like a balance of soft and sharp that you naturally get from 2K or 1080p ProRes. You can get that with higher res formats but it seems like a balancing act.

There's 2K and 1080p. If you wouldn't recommend 2.8K ArriRAW or 2K ProRes what would you recommend? Some of my lenses don't cover 3.2K so it's one or the other. (Or renting another camera.) The file sizes for ArriRAW are not that large.

Yeah, it was digital IMAX vs a 4K tv. I'm mostly trying to justify shooting 2K ProRes rather than 2.8K ArriRAW (I have access to both for the same cost – the only issue is getting the camera up and running and the extra work in post). But I noticed along the way I kind of prefer the look of softer images. And it makes me wonder if I'd rather shoot 2K regardless. The Alexa is also kind of strange. There must be some internal sharpening with 2K ProRes.

I agree that 4K TVs and LCD panels look sharper than IMAX projection. I also agree that the Alexa is susceptible to aliasing, but it’s never been a deal-breaker for me personally. I think there’s a happy medium for sharpness for each of us that’s subjective. I think for me it’s a bit softer than for most people. My favorite DP growing up was Kaminski and his style is full of diffusion and grain. If I had all the resources in the world I think I’d get that softer look with high res acquisition, vintage lenses, grain added in post, maybe selected blurring in post, etc. Yet to me the Lighthouse has a more compelling look than Mank (yes, I know both are 4K finishes) so the extreme of doing it all in camera looks better to me than the extreme of doing it all in post. Regardless, I don't think I prefer the look of 2K to 4K after all.... I just prefer a softer look than most do. And a lower acquisition resolution is one path to getting that look. (One I prefer to shooting 6K Venice or whatever with wide open vintage lenses, or emulating that look in posts they did on Mank – but not one I prefer to whatever they did on Suspiria (2018) or the Lighthouse, both 4K finishes. Again, these are subjective aesthetic choices and preferences so I'm not saying one is "better.")

I read that in Mandalorian they shoot the suit wide open to soften the chrome helmet but shoot the rest a bit more stopped down. Not sure what the story is with Rings of Power but I think it's the same camera.