In his presentation Debunking “HDR”, cinematographer Steve Yedlin shows a slide of BT.1886 to argue that it inherently matches human visual perception. To support his claim, Yedlin points to the distribution of code values:
“This has been the theme the whole time, is that the human visual system is relative more than absolute. So it’s not accidental that 18% of scene white—18%, which we think of in photography as neutral—is stored at 50% in the code value. So, you have half of the code values allocated to luminance values above neutral gray and half of the code values allocated to luminances below neutral gray, and it works really, really well. So, the shape of this curve, combined with the y-axis actually working like the human visual system as a percentage, works great. That’s why they came up with it.”
Yedlin’s claim contains an elementary technical and historical error.
First, he implies that BT.1886 was intentionally developed from the ground up to match human visual perception.
It wasn’t.
BT.1886 was explicitly engineered to mimic the electro-optical transfer characteristic (EOTF) of legacy CRT displays, providing a reference EOTF for the flat-panel displays used in HDTV production.
Second, Yedlin misattributes the scene-light-to-code-value relationship directly to the display-side BT.1886 EOTF. In reality, that mapping belongs to the camera-side OETF defined in ITU-R BT.709.
The 18% to 50% mapping is a system property defined at the point of capture, not a characteristic of the display.
The chronological evidence is irrefutable: BT.709 was first approved in 1990, while ITU-R BT.1886 wasn’t published until 2011—more than two decades later.
The Daejeon Chronicles 
Leave a Reply