Forum Replies Created
-
Replies
-
With digital post, it’s sort of the opposite — we not only have control over the gamma (contrast) of every frame, we have control over portions of the frame.
In the most general sense, the concept of “visualization” (or “previsualization”) of how one’s exposure and development choices would translate to the finished product has some loose application in cinematography, but in the specifics, it’s very hard to apply a system that was based on taking single still images rather than a sequence of moving images for a scene that all have to intercut. It’s similar to why it is hard to use ETTR (Expose To The Right) still photography approach in cinematography except for one-shot scenes or visual effects shooting, or just as a general idea of “get a good exposure, just don’t clip detail”.
Ansel Adam’s idea was that knowing the intended contrast of the print, one could expose and develop the negative so that tonal values fell into the areas that one wanted in the print. With the motion picture photochemical approach, this was hard to do because we couldn’t adjust the gamma of the negative on a shot-by-shot basis and we didn’t have a lot of control over the gamma of the release prints either, unless we did something like skip bleach or ENR, and then the whole reel had to have that same technique.
If the camera is recording something like Arriraw Open Gate at a 1.55 : 1 aspect ratio using spherical lenses, then both the 2.39 theatrical and 1.90 IMAX versions are cropped from that larger area. Technically then the 1.90 IMAX version is “opened up”, not cropped in, compared to the 2.39 version.
Generally you frame for one aspect ratio on set — usually the one most people are going to see — and “protect” the area outside of that as best you can for reframing for the other aspect ratio.
Recreating the natural fall-off of skylight in a set with windows is very difficult due to the inverse square law. On the other hand, plenty of movies and TV shows have lit convincing day interiors even if the fall-off rate isn’t like it would be in real life.
Obviously you start by working large and far as possible on stage to get a gentler fall-off rate for the soft light.
In a wide shot that is static and the windows are to one side of the frame, ND grads or ND attenuators are useful. I just used an ND.6 attenuator a few months ago for a shot in a white kitchen where I wanted to light someone at a wall-mounted phone from the side but the wall was getting too hot even with flagging. A similar thing can be done in post with Power Windows as long as nothing is clipped but it’s nice to get it right in camera.
For closer shots, you can sometimes use nets or an additional diffusion frame on the foreground person by the window to darken them relative to the background person who is deeper in the room. But of course once you go in closer, it’s easier to balance things.
Yes, if the fall-off is too steep opposite the windows, you sometimes do things like bringing up the room with a soft ceiling or floor bounce. The softer it is, the less “source-y” it is.
Moonlight is sometimes a realistic source – I shot this still photo in the desert under real moonlight.
If you are recording Arriraw on the Alexa, white balance is only metadata anyway. The sensor is naturally “biased” towards preferring daylight, i.e. it is less sensitive to blue so likes light with more blue wavelengths, so when you convert the Arriraw recording to RGB for a 3200K scene, the blue channel is being boosted in comparison to the others to balance the color for 3200K lighting.
If you record log-C with a color temperature selected, then that color temperature is “baked in”, meaning that if you set the camera to 3200K, then the camera is boosting the blue channel to compensate for a lack of blue signal.
So it is a bit like chasing your tail to shoot in daylight with the camera set to 3200K, record log-C, end up with the camera boosting the blue signal off of the sensor and creating a blue-ish daylight scene, and then in post, lowering the blue channel back down to correct the image to daylight-balance. It may turn out OK, but keep in mind that you’d be taking a recording where the blue information was needlessly boosted and then having to lower it again while boosting the red information to compensate. When you bake-in color temperature, it’s more of a WYSIWYG scenario, you only have the colors you see to work with in post. So you have to ask yourself what you are gaining by recording more blue than you need and less red than you want. Unless you actually want a cold day look.
I think film is similar but also different, even if you shoot daylight on tungsten-balanced film (which is film in which the blue layer is faster to compensate for a lack of blue in tungsten light), you are still recording information in each color layer, so red information can be recovered depending on your base exposure.
It often comes down to Theory and Practice. You read, you watch movies, you shoot, and repeat over and over again.
Activity and Reflectivity are key as well, you need time to digest what you’ve learned. Vittorio Storaro once said that after his first feature, he couldn’t find work and spent two years visiting museums and studying art. At first he thought of those two years as lost time, wasted time, but later he realized that these were the most important years of his professional life because they gave him the foundation for the rest of his career.
What’s a vector set? Is this a day scene? Do you have a grid above the set? How big are the windows? What do you want it to look like in terms of mood?
Keep in mind that the proper or right exposure artistically isn’t necessarily a full or “normal” exposure – what’s more important is consistency of exposure across the coverage. That’s where the tools are handy.
It’s tricky to talk about how to know exactly how your negative is exposed except in obvious cases of heavy over or underexposure. Even when talking about printer lights, keep in mind that the printers were calibrated at each lab slightly differently, printing at 25-25-25 (middle of the 1-50 scale for each color) didn’t yield the same results at every lab with the same piece of negative — though the lab could tell you want something shot “normally” based on LAD values should print at. More or less.
This is why testing in prep was so important, what mattered is that you found a set of printer lights you felt was best for the image you wanted. After that was determined, day by day you could find out if your footage was printing close to what you wanted, knowing that underexposed footage would end up using lower numbers, etc. Or you could do what cinematographers like Richard Kline did, actually tell the labs to use a set of numbers — then you’d find out yourself if your footage was too dark or too bright rather than have to corrected for you. Of course he did that in the days when there was only one Kodak stock to shoot. With the varieties of stocks later, and daily variations in lab chemistry, etc., I was more likely to ask that the roll be timed for a grey card shot at the head.
If you’re shooting film, then having a neutral gray card (and face) — shot in boring, flat “white” light so that there is no creative interpretation involved in deciding what “neutral” and “normal” is — is the best way of knowing if the footage that follows the card was under or overexposed a bit more than intended.
This is best answered by a focus-puller — there are a number of techniques that can be used, larger monitors, zooming-in on the monitor signal, peaking tools, etc.
If your AC is struggling, you may have to give that person a deeper shooting stop as well, though a deeper-focus image actually makes it harder to focus by monitor since you don’t see the fall-off in focus as clearly. But then, you also don’t see the focus mistakes as clearly!
Perhaps a pola filter would help, flags would help… but today there is very good spill correction software that compositors use to deal with that.
You can’t really compare the two because your shot is not in focus. If you zoom into it, you can see that you’re focused on the tip of his nose, not his eyes.
<p style=”text-align: left;”>The lens (and its f-stop setting) isn’t the only factor. You haven’t mentioned what camera you used nor the recording format nor the delivery format. Nor the source of the frame from “The Revenant”. Also in a small image seen on a computer or mobile device, the difference in sharpness between two similar images might just be the amount of electronic sharpening applied.</p>
The majority of cinema has been shot on standard 35mm for a century, not Full Frame. So there is no reason for filmmakers to always convert the focal lengths they are using to Full Frame equivalents in terms of field of view when discussing their movie. When we talk about Orson Welles’ use of the 24mm on “Citizen Kane” or the 18.5mm on “Touch of Evil”, we don’t convert those numbers to Full Frame equivalents.
