“Next distinguished cinematographer Steve Yedlin ASC enlightened us all on the subject of HDR. Among the key takeaways were that the whole notion that the SDR version has to look totally different than the HDR version is wrong. Depending what the scene is, there may be no difference between SDR and HDR. Being able to go much further in the highlights will only show up if the scene allows it (a very bright window for instance).”
—Lars Pettersson FSF, ODCC 2025, IMAGO, The ninth edition of the Oslo Digital Cinema Conference, Dec. 11, 2025
What Pettersson is describing is how not to leverage HDR for powerful storytelling. It’s a textbook example of subordinating HDR to SDR and providing cover for filmmakers to avoid having to engage with the format.
Why Yedlin was even invited to speak at the ODCC after his anti-HDR rant last year defies comprehension.
Note: The timestamps below refer to his presentation Debunking “HDR”:
Yedlin Fails to Understand PQ
Yedlin admits, ”I don’t know who came up with this PQ curve” (1:15:03 – 1:15:13). Dolby did. He confesses, “I don’t know why they came up with this PQ curve” (1:14:54). Not understanding why PQ was created is patently disqualifying for anyone attempting to debunk HDR. Per ITU-R BT.2390 Sec. 5.2, “The traditional gamma nonlinearities of Recommendations ITU-R BT.709 and ITU-R BT.1886 are unsatisfactory when stretched to the much larger dynamic ranges desired for future television productions.”
Yedlin Confuses PQ with Log
Yedlin insists that PQ is logarithmic, specifically “inverse log,” at several points in his presentation (31:15 – 31:22) and (32:38 – 32:42). Reality per ITU-R BT.2390: PQ is perceptually quantized (Sec. 5.2), not logarithmic. It uses the Barten vision model to match the contrast sensitivity of the human eye.
Yedlin Cripples PQ
Yedlin invalidly tests PQ at 8 bits and declares BT.1886 superior (1:20:47 – 1:22:15). PQ was never designed for 8-bit; ITU-R BT.2390 Sec. 5.2 requires 10/12-bit pipelines to avoid artifacts (Fig. 14 confirms PQ works at 10 bits, Fig. 13 shows it’s flawless at 12 bits). ITU-R BT.2390 never mentions 8-bit viability.
Yedlin Derides PQ as Inefficient
Yedlin calls PQ a “data hog”: “You should be able to see how much more efficient SDR is at using the bit depth than HDR. When they say it’s 10-bit, they mean ‘it needs 10 bits’ [snickering]. It’s a data hog, is what it means. So that means at any data rate, the SDR curve is gonna be better” (1:22:42). ITU-R BT.2390 proves the opposite: BT.1886 needs >12 bits for HDR (Fig. 13: “rises substantially above the visual threshold”), while PQ delivers artifact-free HDR at 10/12-bits. Yedlin’s critique inverts reality.
Yedlin Distorts Historical Reality
Yedlin implies that BT.1886 was built from the ground up to match human vision. Specifically, he states (1:14:45 – 1:14:52): “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.” BT.1886 was designed to match hardware behavior, not the human visual system. Per ITU-R BT.1886, “The reference EOTF is specified as a simple equation, with exponent function, based on measured characteristics of the Cathode Ray Tube (CRT).”
Yedlin misattributes the scene-light-to-code-value relationship directly to the display-side BT.1886 EOTF (1:14:26 – 1:14:52). In reality, that mapping is defined by the camera-side OETF specified in ITU-R BT.709. It is a property of the overall imaging system, not of the display-side BT.1886 EOTF.
Yedlin Conflates BT.1886 EOTF With a Color Space
Yedlin refers to BT.1886 as a display color space throughout the video (14:57, 24:45, 26:00). “Rec.1886 primaries” / “BT.1886 is a color space.” ITU-R BT.1886 defines solely a reference EOTF. It has no white point or RGB primaries: “This Recommendation specifies the reference electro-optical transfer function (EOTF) that the displays used in HDTV programme production should follow in order to facilitate consistent picture presentation.”
Yedlin Reduces Everything to a Variant of BT.1886
Yedlin reduces everything to no more than a variant of BT.1886: “Display P3 is the color space that Apple computer monitors use. The monitor itself is Display P3, but very often it’s receiving Rec.1886 and just converting it to P3. Rec.2020 is exactly the same as Rec.1886, but with a wider color gamut. sRGB is also exactly the same as Rec.1886, but with a slightly different transfer function.” All three claims are physically impossible. Display P3 is a color space. BT.1886 is a transfer function. One cannot “convert” BT.1886 to Display P3. Rec.2020 is a color space. BT.1886 is an EOTF. It has no primaries. Therefore, Rec.2020 cannot be “BT.1886 + wider gamut.” sRGB is a color space. BT.1886 is a transfer function. Therefore, sRGB ≠ BT.1886 + “a slightly different transfer function.”
Yedlin Alleges HLG Is an SDR Color Space
Yedlin claims, “HLG (in effect if not in name) is just yet another SDR colorspace.” HLG is emphatically not “just another SDR colorspace.” HLG is an HDR format that uses the HLG transfer function and BT.2020 color primaries and was designed to be backward compatible with SDR displays.
Yedlin Claims Rec.2020 Is an SDR Color Space
Yedlin asserts that Rec.2020 is an SDR color space (1:30:01 – 1:30:25). ITU-R BT.2020 specifies BT.1886 as its baseline EOTF, but anticipated an improved one: “if it is shown that an alternative electro-optical transfer function (EOTF) will provide significant benefits without also imposing significant disadvantages, then this Recommendation should be extended to enable use with an improved EOTF [1].”
Yedlin Reduces HDR to a Color Space
Yedlin conflates HDR and SDR formats with color spaces: “When it comes to ‘formats’ (i.e., colorspaces)…” (1:06:33). The HDR format is not a color space. It is an entire ecosystem of which color is but one component.
Yedlin Claims ITU-R BT.2100 = SMPTE ST 2084
Yedlin explicitly states that ITU-R BT.2100 = SMPTE ST 2084. SMPTE ST 2084 (PQ) is an EOTF. BT.2100 is the ITU-R standard for UHDTV (HDR/WCG), defining—among other things—Rec.2020 primaries, white point and EOTF options (PQ/HLG).
Yedlin Claims Color Spaces Are Perfectly Convertible
Yedlin frequently asserts that color spaces are “perfectly convertible” because “any color within the gamut” can be matched (12:38 – 12:54, 14:28 – 14:35). Color space conversion is not lossless.
Yedlin Claims WCG Not Exclusive to HDR
Yedlin insists that “Wider gamut is not a distinction between HDR and SDR.” (1:06:58, 1:30:52). While BT.2020 does technically define a path to SDR/WCG, it’s effectively unavailable to consumers via streaming services, which exclusively deliver WCG content through HDR formats.
Yedlin Claims BT.1886 Preserves Creative Intent Better Than PQ
Yedlin claims that BT.1886 preserves creative intent better than HDR: “SDR is the more reliable system for viewers to see relative contrast as authored by the filmmaker.” Reality per BT.1886 & BT.2390: SDR’s OOTF alters contrast by design (Sec. 2.1). PQ was engineered to preserve absolute luminance relationships.
Yedlin Claims SDR Reproduces Subtler Increments
Yedlin claims that SDR can reproduce subtler increments of luma and chroma at a “given bit depth” than HDR (2:11:27 – 2:12:00). Consumer SDR is typically 8-bit. HDR is 10-bit. 10-bit gives 1,024 steps per channel versus 256—a 75% reduction in step size—and 1.07 billion colors versus 16.7 million. PQ allocates those steps more efficiently, prioritizing more codes for lower luminances where the eye is more sensitive.
Yedlin’s Folly
Yedlin proposes “going back to a relative system with a gamma-style transfer function [and] keeping a wide gamut but only going as wide as is actually used in practice (which is more like P3 than Rec.2020)” (2:05:34 – 2:06:50). His scheme is physically impossible. A pure gamma function would require substantially more than 12 bits to avoid banding in HDR, as illustrated in Fig. 13 of ITU-R BT.2390, a document Yedlin appears not to have consulted. As Charles Poynton noted over a decade ago: “Today’s BT.1886 is not capable of HDR; in order to accommodate HDR content in the transmission chain, we’ll need an HDR-capable quantizer [2].”
- The Historical Intent of Rec.2020
Yedlin’s labeling of Rec.2020 as a strictly “SDR colorspace” is ahistorical. The ITU-R BT.2020 standard (Aug. 2012) explicitly anticipated HDR displays and the need for a new EOTF. Parallel to this, the PQ EOTF was proposed to the ITU-R in spring 2012 and was already undergoing SMPTE standardization, proving the standards body was architecting for HDR from the outset. - Charles Poynton, Stessen and Nijland, Deploying Wide Color Gamut and High Dynamic Range in HD and UHD, SMPTE Motion Imaging Journal, April 2015.
The Daejeon Chronicles 
Leave a Reply