DEBUNKING "HDR"

[Martin Stevens]

So, my take on this is that Steve Yedlin is correct in stating that HLG is essentially the same as SDR in functional terms, but with a less optimized transfer function. Both use relative encoding rather than absolute luminance values, meaning they rely on the display’s capabilities to interpret brightness—unlike PQ-based HDR, which encodes fixed nit levels.

While HLG adds a logarithmic curve to extend highlight range, it offers no real benefit over SDR in most post-production or delivery workflows, and its less refined tone curve can actually degrade image control. Furthermore, HLG lacks wide support in end-to-end pipelines, making it impractical for cinematic or streaming content. Including it as another deliverable would only add unnecessary logistical complexity without improving the viewer experience.

 

[Martin Stevens]

Also, while it may seem counterintuitive, it's not absurd to argue that HLG’s transfer function is inferior to Rec.709 (BT.1886) when judged by practical image optimization and workflow integration. Rec.709’s curve is highly refined for perceptual uniformity in controlled environments and has been tightly integrated into decades of display technology, grading tools, and calibration standards. HLG, on the other hand, was designed primarily for broadcast convenience—specifically for backward compatibility and no-metadata HDR delivery. In real-world grading and mastering environments, HLG offers less precision in highlight control, less consistency across displays, and less support across critical stages of post-production. So while HLG’s curve is mathematically broader, it’s less practically optimized for high-fidelity creative workflows, which I think is exactly Yedlin’s point.

And, the argument isn't that HLG is mathematically "bad"—in fact, it’s a clever compromise for broadcast use—but rather that in a professional post-production context, Rec.709 (via BT.1886) is better optimized for grading accuracy, display predictability, and creative control. HLG introduces complexity (like less predictable highlight roll-off and weaker tooling support) without offering a clear advantage in cinematic or streaming workflows. That’s why it’s fair—and accurate—to say it’s inferior for those use cases, even if technically it's a broader curve.

 

[Jon Pais]

In real-world grading and mastering environments, HLG offers less precision in highlight control, less consistency across displays, and less support across critical stages of post-production.

As Rec.709 is not supported in HDR workflows (since it was defined in the age of the now-defunct CRT) and has a maximum dynamic range of 6 stops, you’re not really going to lecture us about highlight control, consistency across displays or support in post-production workflows, are you? Rec.709 has no highlights. It will clip and suffer from banding, two of the most disagreeable artifacts in video.

HDR content across all of my devices- ASUS ProArt PA32UCDM, Apple iPhone 15 Pro Max, MacBook Pro with Liquid Retina XDR display and Samsung 65” S90C - looks vastly superior to obsolete SDR, and that goes for low bitrate streamed content as well.

For all of his supposed expertise, Yedlin’s Glass Onion looks markedly inferior to countless other movies and TV shows where the filmmakers understood that HDR directly influences how lighting ratios are perceived and where the colorists didn’t lop off highlights.

With the emergence of the piano, old harpsichords were used as firewood in the Paris Conservatory. The same will eventually become of SDR.

HLG… offers no real benefit over SDR in most post-production or delivery workflows.

On the contrary: it’s Rec.709 that has no place in modern HDR workflows and that needlessly complicates on-set monitoring, dailies, post-production and distribution.

 

[Martin Stevens]

Well, I do hate 8 bit banding. : ) And, "will clip" is not the same as saying that it can clip.

I think we're addressing two different things: technical potential versus practical deployment. You're absolutely right that Rec.709 was defined during the CRT era, and yes, its native dynamic range is limited. But the point I was making—and that Yedlin emphasizes—isn't that Rec.709 is technically superior to HDR curves, but that its well-understood, tightly integrated, and predictable behavior across post-production and display systems makes it incredibly effective for preserving creative intent.

HLG, while useful in broadcast contexts, introduces ambiguity. It encodes luminance relatively—just like SDR—but lacks the rigorous post-pipeline infrastructure and calibration standards that Rec.709/BT.1886 enjoys. That ambiguity can lead to inconsistent rendering.

I think Yedlin's broader point is that perceptual consistency matters more than absolute technical range—a film that holds together across viewing environments often communicates more effectively than one that’s brilliant on an HDR reference monitor but unpredictable elsewhere.

And on your Paris Conservatory metaphor—let's not forget that most music is still written and performed on pianos tuned to temperaments developed centuries ago. Sometimes what's perceived as “obsolete” remains because it still works. SDR works well, very well.

 

[Jon Pais]

But the point I was making—and that Yedlin emphasizes—isn't that Rec.709 is technically superior to HDR curves

WHAT??? Yedlin spends like 30 interminable minutes of his 2+ hour presentation talking about how the PQ curve is inferior to SDR gamma 2.4!

 

[Jon Pais]

Yedlin says during the presentation, “the things that are actually different about HDR compared to SDR are all detriments, not advantages.” The man is seriously unwell.

Steve Yedlin’s Technical Assertions Warrant Scrutiny

Steve Yedlin Wages War On HDR “When it comes to “formats” (i.e. colorspaces) as opposed to hardware, with the exception of one thing (how we use the excessive screen brightness availabl…

daejeonchronicles.com

I can assure you that had I written half the nonsense Yedlin says during his presentation over at the lift gamma gain forums, as much as they despise HDR, the colorists would have been all over me like a swarm of angry wasps.

Because cult.

 

[rand thompson]

Here's a TRANSSCRIPT that " ADAM REIGN" feed to CHATGPT that was made into an .rtf file

SendGB | Send Large Files | Free file transfer

Free and fast file sharing. Send large files easily. No registration required. We transfer files up to 5GB. File upload with great functions. Mail big file now!

sendgb.com

 

[Jon Pais]

Here's a TRANSSCRIPT that " ADAM REIGN" feed to CHATGPT that was made into an .rtf file

SendGB | Send Large Files | Free file transfer

Free and fast file sharing. Send large files easily. No registration required. We transfer files up to 5GB. File upload with great functions. Mail big file now!

sendgb.com

Correction: It is not a transcript. What it is is a summary of the presentation. Adam first fed a transcript of the talk to ChatGPT, then he broke it down into sections and emphasized the key takeaways.

 

[Jon Pais]

Fact-Checking A Few of Steve Yedlin’s Claims

CLAIM: SDR Can reproduce subtler increments of luma and chroma at a given bit depth than HDR.

FACT-CHECK: If content is authored and displayed strictly within SDR parameters (e.g., 8-bit Rec. 709), it can appear smooth on compatible displays.

However, HDR is technically superior to SDR in reproducing subtler increments of chroma and luma, thanks to its higher bit depth, wider dynamic range, and perceptual encoding. Yedlin is engaging in pure sophistry. As Yedlin is well aware, an increase of bit-depth is essential to minimize banding artefacts that might arise with the increase of dynamic range.

CLAIM: HDR can display highlights either gently/artfully or more caustically, as preferred by filmmaker but there is community pressure to use only a caustic style.

FACT-CHECK: None of Apple TV+’s recent shows – Murderbot, The Studio, Carême, Your Friends & Neighbors, Dope Thief, Wolfs, Presumed Innocent – have any HDR intent whatsoever and are graded at SDR levels. The same goes for many, if not most Netflix Originals.

The DCI issued a statement years ago declaring that brightness levels are solely up to the filmmaker and colorists have confirmed that studios don’t dictate peak brightness levels.

Although numerous studies of viewer preferences have indicated that highlights much brighter than possible on current TVs were required to satisfy the majority of participants, in the professional community, most advocate for restraint; and Yedlin supplies no proof for the assertion that there is ‘community pressure to use a caustic style’ of HDR grading.

CLAIM: Yedlin writes about SDR that overall luminance (not just little areas of the frame) can take full advantage of a monitor’s maximum physical brightness, while HDR cannot.

FACT-CHECK: We confess to not understanding what on earth Yedlin is talking about here. Why would anyone want to master full screen 10,000 nits, or whatever it is he’s talking about.

CLAIM: Only SDR is capable of faithfully reproducing relative contrast as authored by filmmakers. (if monitor is capable).

FACT-CHECK: HDR is fully capable of faithfully reproducing contrast when the content is natively authored and displayed correctly and can even surpass SDR in representing a filmmaker’s vision for high-contrast scenes.

CLAIM: HDR’s “punch” is optional.

FACT-CHECK: Yedlin greatly underestimates the artistic value of highlights. When used creatively to enhance storytelling, PQ enables this in ways that SDR can never do.

CLAIM: HDR is a grading style only.

FACT-CHECK: HDR is an end-to-end process, from capture and post-production to storage, distribution and display. In order to be successful, the color, contrast and highlight and shadow detail, as well as the compositional choices that make effective use of HDR need to be evaluated on set, thereby ensuring that the look/emotional impact travels through all the way to the final deliverables.

 

[rand thompson]

Correction: It is not a transcript. What it is is a summary of the presentation. Adam first fed a transcript of the talk to ChatGPT, then he broke it down into sections and emphasized the key takeaways.

I stand corrected. I kinda was wondering How CHATGPT would transcribe this. I just assumed this came out as some type of PDF FILE or something.

 

[rand thompson]

Fact-Checking A Few of Steve Yedlin’s Claims

CLAIM: SDR Can reproduce subtler increments of luma and chroma at a given bit depth than HDR.

FACT-CHECK: If content is authored and displayed strictly within SDR parameters (e.g., 8-bit Rec. 709), it can appear smooth on compatible displays.

However, HDR is technically superior to SDR in reproducing subtler increments of chroma and luma, thanks to its higher bit depth, wider dynamic range, and perceptual encoding. Yedlin is engaging in pure sophistry. As Yedlin is well aware, an increase of bit-depth is essential to minimize banding artefacts that might arise with the increase of dynamic range.

CLAIM: HDR can display highlights either gently/artfully or more caustically, as preferred by filmmaker but there is community pressure to use only a caustic style.

FACT-CHECK: None of Apple TV+’s recent shows – Murderbot, The Studio, Carême, Your Friends & Neighbors, Dope Thief, Wolfs, Presumed Innocent – have any HDR intent whatsoever and are graded at SDR levels. The same goes for many, if not most Netflix Originals.

The DCI issued a statement years ago declaring that brightness levels are solely up to the filmmaker and colorists have confirmed that studios don’t dictate peak brightness levels.

Although numerous studies of viewer preferences have indicated that highlights much brighter than possible on current TVs were required to satisfy the majority of participants, in the professional community, most advocate for restraint; and Yedlin supplies no proof for the assertion that there is ‘community pressure to use a caustic style’ of HDR grading.

CLAIM: Yedlin writes about SDR that overall luminance (not just little areas of the frame) can take full advantage of a monitor’s maximum physical brightness, while HDR cannot.

FACT-CHECK: We confess to not understanding what on earth Yedlin is talking about here. Why would anyone want to master full screen 10,000 nits, or whatever it is he’s talking about.

CLAIM: Only SDR is capable of faithfully reproducing relative contrast as authored by filmmakers. (if monitor is capable).

FACT-CHECK: HDR is fully capable of faithfully reproducing contrast when the content is natively authored and displayed correctly and can even surpass SDR in representing a filmmaker’s vision for high-contrast scenes.

CLAIM: HDR’s “punch” is optional.

FACT-CHECK: Yedlin greatly underestimates the artistic value of highlights. When used creatively to enhance storytelling, PQ enables this in ways that SDR can never do.

CLAIM: HDR is a grading style only.

FACT-CHECK: HDR is an end-to-end process, from capture and post-production to storage, distribution and display. In order to be successful, the color, contrast and highlight and shadow detail, as well as the compositional choices that make effective use of HDR need to be evaluated on set, thereby ensuring that the look/emotional impact travels through all the way to the final deliverables.

Yeah I read this on your Blog and it made since to me, But I'm not an HDR expert. Dado Valentic is creating a DCTL for Resolve based around yedlin's Presentation that he is supposed to talk about FRIDAY. So I will see what that will be about . He will release the Video for that for the public on Youtube after that ZOOM Presentation a few days after it has been completed.

 

[Jon Pais]

The following is the final paragraph from the ChatGPT summary.

As colorists, your job is the artful control of relative contrast. Yedlin’s “Debunking HDR” is a reminder that—regardless of whether your client asks for “HDR approval” or “SDR deliverables” —the true creative work happens below the absolute ceiling. Keep that in focus, and your grades will remain consistent, predictable, and faithful in any environment.

The advice argues that colorists should prioritize relative contrast in the low/mid range & treat HDR highlights as non-creative ‘ceiling’ elements and that this approach guarantees consistency across SDR/HDR deliverables. Although well-intentioned, the counsel oversimplifies HDR grading and contains several flawed assumptions. Here's why it's not necessarily good advice:

Highlights Aren’t Just A “Ceiling”: They’re Creative Tools

The core flaw is treating HDR highlights as a superfluous, non-essential, absolute ceiling; as technical ‘overhead’ rather than creative tools. HDR highlights can define a visual language. Think of how highlights on chains and swamp water in Emancipation (2022) reinforce the story’s oppressive brutality, whereas pushing those highlights down to SDR levels would destroy/dilute the artistic intent. Ignoring them relegates HDR to nothing more than ‘bright SDR’, missing out on its enormous creative potential.

Highlight Rendering Directly Affects Color & Texture

We’ve already addressed in great detail how highlight roll off is a critical creative decision (perhaps more thoroughly than in any HDR masterclass!) in the Grading section of our Monster Guide. Treating highlights as merely a "ceiling" ignores the artistry involved in their rendering. Yedlin completely ignores how highlight handling affects color saturation and texture. Clipping (or rolling off, if you will) 1,000 nits to 100 nits SDR isn't just brightness reduction - it's data amputation.

SDR + Highlights

We’re going to keep it brief. Worth noting is the advice assumes an outdated HDR = "SDR plus highlights" way of thinking. The approach underutilizes HDR as a creative medium.

Final Thoughts

We’re just focusing on the final paragraph, which could very well give an inaccurate idea of the whole; which is why it is strongly recommended that you either download the summary or watch the entire 2-hour presentation yourself. To be clear, not once during his presentation does Yedlin say to ignore highlights, nor does Adam Reign’s summary fail to mention Yedlin’s various strategies for handling them. What is not conveyed in the final paragraph quoted above is Yedlin’s undisguised contempt for HDR highlights. Nor does the summary correct the falsehood perpetuated in the presentation that HDR is nothing more than a grading style.

 

[rand thompson]

Jon when I was interested in HDR a few years ago, I took one of those many HDR MASTER CLASSES from someone I won't name. One of the things that was taught was how to set the SDR signal so that later when you wanted to repurpose it for HDR it would look ok. For many years after that class I remembered thinking how is that going to be a true genuine HDR grade. I realize from reading most of what you have posted that most of what I thought I disliked about HDR came from individuals or post houses whom where repurposing an SDR grade for HDR. It makes me want to see how a production that has been shot for HDR in the first place from start to finish actually " SHOULD" Look.

 

[Steve Sherrick]

Steve Yedlin will be on Roger Deakins Podcast coming up and they will be touching on Steve's HDR presentation. I recommend everyone check it out.

Next Wednesday.

 

[rand thompson]

DELIVERABLES & PRESERVING AUTHORIAL INTENT - with Steve Yedlin

 

[Jon Pais]

Yedlin’s treating Rec. 2100 (or more accurately, P3-D65 ST 2084) as just another color space or container really squanders HDR’s potential. Think if, instead of giving us so many immortal piano works, Beethoven had written for the harpsichord, with its limited range and power. HDR is the process of creating for the container’s unique capabilities, primarily dynamic range.

Instead, Yedlin has nothing but contempt for HDR’s expanded range. He talks a lot about relative contrast and how 100 nits on one display can be a perceptual match for 1,000 nits on another display, which is great to know for matching displays, less so for creative grading. A cymbal crashing in an overture by Rossini at full volume may have the same relationship to the rest of the orchestra at a lower volume, but it doesn’t shake me to my very core like it does when performed as written by the composer.

“But you can’t watch HDR in a sun-filled room!”

Well, we don’t watch movies at the theater with the lights on like we’re savages dining at a Denny’s restaurant, either. HDR can be watched that way (e.g. HLG), but it loses all its dynamic range - like listening to Rossini on a car stereo in heavy traffic. FWIW, I only ever watched SDR at home in the evening with the lights off too, since I dislike reflections intensely.

Perhaps the worst consequence is all of these self-appointed experts in the forums, linkedin and elsewhere who distort Yedlin’s concepts beyond all recognition, like this Dolby Vision certified color management expert who mixes up transfer functions with gamuts (!), etc.:

http://daejeonchronicles.com/2025/0...ified-workflow-expert-spreads-disinformation/

 

[Ido Karilla]

Well I could follow Steves theory but it only stands for when your look is first built in SDR.
From my Grading expirience, grading in HDR allows me to achieve looks that I can not achive in SDR. Even if done on the same excuisite monitor. What am I missing?
I never understood why grade the cinema SDR version before HDR and not the other way around, unless you are nostalgic and fearing a change.

 

[Steve Sherrick]

Some interesting thoughts here too.

PQ HDR & HLG

Understanding PQ HDR and HLG

lightillusion.com

 

[Phil Holland]

I've taken a step back responding to this particular presentation. In terms of what I believe Steve's core "issue" is comes from maintaining the creative intent and potentially having one quality master out there. This particular topic has been big in the filmmaking community overall. And it hits on a more interesting level when you monitoring either HDR or SDR on set and where the master is developed from there. Prior to now, a big topic has been if the cinematographer stays on during the color sessions or is apart of the decisions made in post. I'm fortunate as I'm pretty involved if not entirely involved (i.e. doing the work) with all decisions.

All of this hascome up in my world as well as I've been mastering for HDR for over a decade now and generally I am still making unique SDR and HDR deliverable to provide the best of what both have to offer. I should also underline I make different masters for different display mediums as well and in some cases, different viewing scenarios. A bag of tangled yarn to talk about the nuance of this, but if you control the viewing experience top to bottom, you can really create for the best experience.

The educational aspect of understanding that gamut and gamma are two different things regarding a colorspace as a whole is good. However, one again, from a physics perspective the combination of gamut and gamma can produce colors that are impossible to recreate in SDR from an HDR master, the other way around however is very doable. The general concept of relative contrast is fine as it does pertain to our human vision, but SDR is actually dimmer by a whole bunch if it's being implied.

This all being said. I've found that it requires a great deal of care and attention as well as testing to make a compelling HDR grade that showcases the display medium that also doesn't exhaust the viewer. On the SDR perspective, this is much easier as you can't produce "annoying or uncomfortable to view" as easily. Your brain is doing a lot to essentially normalize the image to ingest the information, but there are minimums and maximums. This is really all about the three bears and the porridge being "just right". And many will have an opinion of what just right actually means.

What I will say is before and even during HDR's early development I was pushing for support for broadly released 10-bit SDR content, rather than 8-bit. And I think we'll see beyond 10 and 12-bit HDR in the future. I come from a feature film 16-bit delivery background and much of my personal mindset is based on that metric. I'd also enjoy a slight increase is visual fidelity for web encoded streaming content. Each streaming service provides different quality. Similarly, television and display technology as a whole is also needing to tighten up when it comes to reproducing gamma, particularly in the shadows.

Sidebar. I have been on 4K and 8K displays, I personally find it difficult to enjoy poorly encoded 1080 or 720 material at this point even with a good upscale. It's been a long while since I've had an 1080p display to watch or create with. But I keep a couple panels here to test content on occasionally. My general POV is if the audience is paying for the content, whether it's physical media or streaming, we should be focusing on a high quality delivery.

The initial hope for digital delivery was to harmonize some of the chaos we experienced from mastering for theater and eventually bringing the film to homes via some form of medium. VHS, Laserdisc, DVD, Blu-Ray, HD DVD, UHD Blu-Ray, Cable, Streaming, etc. However, between new delivery requirements, multiple masters, and a medley of display technologies, it's actually become a bit wilder than it was while also dancing with the concept of being better overall.

Tricky stuff. But there's an artistry to it.

Bringing all of this into a more photographically approachable concept. And long time REDusers may remember, I somewhat consider SDR printing a photograph on paper, while HDR can essentially be described as a backlit Cibachrome print. I did and do different masters for those as well. And more interestingly, if you've printed anything to a variety of materials or even different papers, you'll find benefits to tweaking the image to each display medium.

And since Steve mentions fundamentals of art, I share a similar sentiment when it comes to painting/drawing as well as creating a quality motion picture. You don't always need to paint with pure white or black. Often it's nice to work within a limited tonal range all together. It is a bit different and difficult painting the painting twice or more times however for different affect.

 

[rand thompson]

Here are two DCTLS created by Dado Valentic to translate the intent of your SDR Grade to HDR and the HDR Grade to SDR. The DCTLS are currently only available to members of the "ART OF COLOR RESEARCH GROUP" Website. For a limited time the membership if $49.99 a Month. I am not currently a member of that Website.

LINK to WEBSITE

HDR DCTLs - colour training

Two Professional DCTLs for Seamless SDR ↔ HDR Workflow Match Your Look Across SDR and HDR Without Compromise In today’s hybrid delivery world, colorists are constantly challenged to create HDR and SDR versions of the same content—without sacrificing creative intent. That’s why we’ve developed...

www.colour.training

Here are two videos Explaining the method and the DCTL.