Forum Replies Created
-
Replies
-
The primary purpose of the 2383 print emulation LUT isn’t to create a film look, it’s to work within a color space/gamut that works for both a DCP and a film-out so the movie would look the same in both versions. It’s less of an issue today with so few film-outs being done.
You have to start with what you want it to look like before you figure out what lights are needed. What effect are you creating outside the windows? Passing streetlamps? Passing lights from buildings? Moonlight? All of them? What light is on inside the bus? What colors do you want? What makes sense for the scene logically? Dramatically? Are you matching sources established in exterior shots? You have to have all those answers before you start talking about Skypanels versus Astera tubes.
I’d also say that if the backgrounds are going to be black, you don’t want to overdo a lot of passing lights from multiple places that can’t be seen, unless you dirty up the windows a lot with rain drops & dust, etc. so that the light washes out the black background. Otherwise if the interactive light gets too busy like the bus is in the city, you’d expect to see a view. So if this is more of a lonely country road or suburban street, there would probably be just passing overhead streetlamps, so given that there isn’t a lot of light outside the bus, it suggests that there are some interior lights on unless you are aiming for a very moody, shadowy scene, which is possible. It all goes back to painting a picture in your head as to how it looks before you can then figure out how to light it.
Most of us work our way up in budgets so it’s partly a matter of scaling up over time in terms of crew, equipment, schedules, ambition, etc. I shot a dozen short films on my own where I was the entire crew, then I shot over 20 short films in film school with a tiny crew, then I shot over 20 small features before I even joined the union. It’s rare that someone makes a huge leap right at the start. But even if you do, you often have an experienced crew to support you.
In other words It’s a LUT for the initial color correction, not for the grading, Is It correct?
Isn’t “grading”, “timing”, and “color-correction” all the same term? The only differentiation today in digital is whether it is for dailies or for the final master. Sometimes VFX elements are pre-graded, or plates for process work.
Camera LUTs are mainly for on set displays and for dailies. Whether they get used as a starting point for the final color-correction just depends on how the colorist and cinematographer like to work. Some might start from scratch from the raw files but then throw on the camera LUT now & then to see how close or far they are.
You need some sort of LUT to convert a raw/log signal for display on a monitor, whether Rec.709 or P3 — most personal LUTs are close to the standard Rec.709 LUT, just tweaked either for personal taste or for a specific look.
Also the color science behind a particular camera design will have an effect on color, as well as any LUTs applied, etc.
Technically you’d have a little less noise at 5600K because sensors are naturally close to daylight-balance – the blue channel has to be boosted after raw conversion to make the image balanced for 3200K.
However, regarding color reproduction, that is more an issue of the lighting technology involved rather than the color temperature. True daylight and tungsten are continuous spectrum sources whereas LEDs have spikes of color at different frequencies, some have more color phosphors / spikes with smaller gaps, creating more accurate color reproduction.
Yes, if you go from 100 foot-candles to 200 foot-candles, you’d have twice as much light so would have to close down by 1-stop to maintain the same exposure. Or cut the ISO by half. Or close the shutter angle by half.
I’ve never used contrast ratios because that requires some conversion math… but in terms of measuring and noting the shadow level, if that’s done, it’s more about matching coverage. You light the wider shot by eye and create the contrast you want, but sometimes when you start covering a scene, you might take notes on the shadow level versus the key for matching purposes, especially if you are shooting VFX elements for compositing.
Of course everything is on a spectrum, it’s not like the choice is between anamorphic lenses with zero anamorphic artifacts versus the funkiest, oldest, more flarey and distorted anamorphic lenses ever made. But it seems to me that one would want some anamorphic artifacts, even if subtle, if you were going to pick anamorphic — if all you want is a 2.40 frame, there are easier, simpler, cheaper ways of getting there.
The only reason I am “split” on the subject, as you say, is that if you reduce every optical indication that something was shot on anamorphic lenses, then why shoot on anamorphic lenses? In the pre-D.I. days of film prints, a 35mm anamorphic production used a larger negative area than Super-35 cropped for a 2.40 image plus every generation was contacted printed. And there was somewhat less depth of field on average compared to Super-35 due to the longer focal lengths used.
But now with digital and larger format cameras, none of those are issues; even with 4K cinema release, the current digital cinema cameras with spherical lenses deliver enough resolution, and shallow-focus isn’t hard to get with fast spherical lenses. So that leaves using anamorphic lenses for their unusual optical qualities.
The focal length you use also depends on how much of the plate you’ll be cropping into. With car scenes, usually there is a 1:1 ratio because the edges of the screen are mostly blocked by the car body, plus in the days of 35mm 1.85, your plate was probably shot and projected 4-perf 35mm full aperture but your camera’s 1.85 frame lines were slightly cropping in on this. So you’d use the same focal length for the plate and the stage shot. But if you need to crop into the projected image more, you should shoot the plate on a slightly wider focal length.
The thing with dynamic range is that you don’t always need a lot of it… until you do! You may be rolling on a second take when suddenly outside the window the sun comes out and hits a white car parked across the street.
It’s just the nature of additive versus subtractive color — if you want to make a primary color deeper and more saturated in a film print, it needs more dye so it gets denser, letting less light through. You want to make a primary color more more saturated with a digital projector you let more of the pure red, green, or blue light through. You can go even further with RGB lasers as a light source.
He looks lit by a softened HMI coming from the right through another window, but whether the window in the middle of the frame needed ND gel is hard to tell – it looks sort of misty/foggy and overcast out there so maybe no ND gel would be needed. In a location like this, you might have hard acrylic ND gel panels made for each window that can be quickly slapped up — or removed as it gets darker outside.
