- Hollywood Tech Nerds
- Posts
- A Legendary DP’s Tech-Minded Approach to Color Grading
A Legendary DP’s Tech-Minded Approach to Color Grading
PLUS: What Google Learns From Your Photos
Welcome to December Hollywood tech nerds!
In this week’s post:
Subscribe to get Hollywood Tech Nerds magically delivered to your inbox every Tuesday!
A Legendary DP’s Tech-Minded Approach to Color Grading
Robert Richardson is an Oscar-winning, highly-esteemed cinematographer with a resume that encompasses everything from Oliver Stone’s Platoon to Martin Scorsese’s Casino to Quentin Tarantino’s Once Upon a Time… in Hollywood and even to comic book movies like Venom: Let There Be Carnage!
I was super-stoked to see Richardson interviewed in American Cinematographer, where he discusses his thoroughly modern approach to color correction and grading. Richardson truly does meld his vast experience shooting on film with the technological advancements available via digital.
On the difference in color timing between shooting on film vs. digital:
When you’re shooting and timing film, there’s always a question mark. Did you get it right? Did the lab get it right? There are so many issues you have to contend with that are solved by the digital process. …the mystery is reduced by having a high-quality monitor on set. Most directors have video villages, so I’ll try to get a 65-inch or something large for the director and for crewmembers, and I want it tuned by the post house so that the image I have, especially when I’m operating a remote system, is exactly what I want, and the director and the producers see the same thing.
…we’re providing an image that is a lot cleaner and easier to show to networks or studios. We didn’t have that capability with film. You’d shoot the film, it’d be cut on a Steenbeck or whatever it was back then, and then you’d go to a screening room and see all of these flaws because the film was not always properly preserved. I saw a cut of Platoon six months after we shot it, and it was yellow…
For years now, for nearly every digital film I’ve done, I always try to get a colorist on the set or in a room close to the set. That way, I can go through the dailies every day either at lunch or during a break. This is one of the shifts you’re going to see in filmmaking: More color will be done on set.
However, when shooting on film, Richardson still finds exciting ways to manipulate color:
For Snow Falling on Cedars… I was able to work with [dailies colorist Thor Roos] to get a bleach-bypass [process]on all the dailies, and it’s probably the happiest I’ve ever been in terms of working with the director and colorist closely to get ‘the look’ on film.
When Yvan and I were doing Once Upon a Time… in Hollywood, some of our film fell to the bottom of the developing tank. They recovered just a small section of it, and it came out looking so beautiful, like a Stan Brakhage film.
Richardson also has an unsurprisingly practical evaluation of the potential use of AI in crafting a film’s look. Instead of hype-cycle wishcasting, Richardson says:
I think AI is certainly going to influence some productions more than others, and as it becomes more advanced, you’ll see it start to infiltrate at even higher levels of the creative process. Being able to ask AI to come up with a look — something you wouldn’t have thought of on your own, especially on a budget where you don’t have access to people like Yvan, Stefan, Stephen or Élodie — could be revolutionary. You could say, ‘Give me the ballet scene from The Red Shoes,’ and AI would deliver that. Or you might ask for Black Narcissus and Apocalypse Now and combine those looks. It would be a missed opportunity if this technology didn’t evolve in that direction, but ultimately it has to be you, the artist, saying, ‘I want this.’ The final creative vision should always rest with the artists.
What Google Learns From Your Photos
I very much enjoyed/was unsettled by this article in Wired about Vishnu Mohandas, a former Google engineer so disturbed by Google’s military-aligned AI training that he left and started his own photo storage and sharing service as an alternative to Google Photos, which uses uploaded images to train its AI systems.
The resulting service is Ente, but even more fascinating is Ente’s promotional site They See Your Photos:
People can upload any photo they want to the website, which is then sent to a Google Cloud computer vision program that writes a startlingly thorough three-paragraph description of it. (Ente prompts the AI model to document small details in the uploaded images.)
The resulting description is incredibly creepy. Don’t believe me? Just look what happens when I upload my profile picture from this very newsletter!
Here’s what it says:
The photo is a close-up shot of a person on a phone call. The foreground is dominated by the person's face and the phone, while the background is blurred but shows a natural outdoor setting with green foliage suggesting a park or garden. The lighting suggests it's likely daytime.
The individual appears to be a Caucasian male with reddish-brown hair and beard, possibly in his late 20s or early 30s. He looks serious and focused on his phone conversation. He's wearing sunglasses and a ring, suggesting a relatively relaxed yet possibly professional lifestyle. The picture was taken on May 13th, 2016, around 2 PM using a Canon EOS 60D camera.
The reflection in the sunglasses shows a bit of the surrounding environment, mirroring some of the background elements. There's a subtle hint of stubble visible just above his beard line. The person is holding the phone to his right ear, with his fingers slightly gripping the device. His facial expression conveys a sense of concentration and seriousness, hinting that he is engaged in a conversation of some importance.
Horrifying for many reasons, not least that it’s clocking me for using a profile picture almost 10 years old. How dare you! It is not correct that I’m holding my phone to my right ear (I guess AI can’t tell left from right) or “that he is engaged in a conversation of some importance,” because this picture was taken while I ordered a pizza.
Kernels (3 links worth making popcorn for)
Here’s a round-up of cool and interesting links about Hollywood and technology:
How they created the incredible scale of Gladiator II. (link)
The 200 best inventions of 2024. (link)
In a world of distraction, here’s how to reclaim your attention. (link)