“The inherent process of printing an image… you are destroying dynamic range. That’s the job… This is actually part of what makes images work… I actually want these things to clump together into the bottom, the middle, and the top.” — Cullen Kelly
Cullen Kelly recently gave a masterclass on the necessity of contrast and how the compression of dynamic range is essential for artistic impact.
It’s a brilliant explanation of SDR-era image-making.
Only it contradicts the ‘SDR-first’ workflow he promotes with Steve Yedlin. Let’s break it down.
Kelly was describing the art of SDR grading, and in that context, he is 100% correct. With a hard ceiling of ~100 nits, Kelly’s description is gospel:
- Crush shadows to create depth.
- Roll off highlights to squeeze them into the 100-nit ceiling.
This “homogenization,” as he calls it, is what makes an SDR image feel rich and “digestible.” This belief that limitation is a virtue extends to his view of art history itself:
“Go find me a painting where you can discern more than five to six stops of dynamic range… that’s generally not the way we absorb a painting… that’s actually part of what makes those paintings so captivating.” – Cullen Kelly
Rembrandt didn’t choose limited dynamic range—he worked around it. Kelly’s theory of “captivating limitation”—which he presents as objective truth—is nothing but a post-hoc justification for technical limitation. The Old Masters didn’t sit around praising the “captivating” limitations of their pigments.
Elevating limitations to virtues is the core error of the SDR-first mindset.
His advice is perfect for working within a limited system. But applying this SDR-minded philosophy to an HDR system would be a fatal mistake.
- In SDR: You’re destroying range to create contrast because you’ve got no choice.
- In HDR: Setting an arbitrary speed limit of 200 nits neuters HDR’s core strength—preserving spatial contrast relationships that SDR compresses.
Kelly’s advice is to ‘lean into contrast,’ but by demanding that the HDR and SDR grades match each other, the Yedlin/Kelly workflow actually leans away from it.
The Daejeon Chronicles 
The thing that drives me so crazy about all of this. PQ gives people the freedom to do what they want. If they want to create a low contrast SDR looking image in a PQ container they can. If they want to use more of the dynamic range available, they can make use of that.
The most annoying thing, when they use PQ, it actually gives them absolute control over the brightness of the image, so they can basically force their 100nits limited look on everyone. If people are watching 100nit graded work on any normal SDR display, it’s very likely the viewer is looking at it much brighter at 200-500 nits since that has become the modern standard for brightness on consumer devices. They should be praising PQ for being able to maintain their kind of ugly, low contrast, low dynamic range intent.
It’s also funny how they treat this like a new thing. Mad Max Fury Road came out when 10 years ago? A film that has a massive range if imagery from shallower dynamic range low light to super wide dynamic range, and was graded in a way that suits the film beautifully.
I don’t think anything could possibly save Star Wars: The Last Jedi, but not having 200nit limited light sabers sure wouldn’t hurt.