If the sunflower and Osaka photos aren’t in HDR, there’s something off with your system. Make sure you aren’t in low-power mode.
Fun fact: the HDR photos are actually looping videos, because that’s way more reliable for browsers and CMS platforms than actual HDR photos. Hopefully this improves in the next few years.
This post came out a few weeks ago. To answer your question in advance, your favorite thing from Mark II that’s missing is eventually coming to Mark III.
I’m right now on a short detour updating our video app, Kino. I couldn’t justify touching it until this Mark III preview was out. After that, look for a few more Mark III preview updates before the big launch.
Not going to lie, it’s been an exhausting 12 months, but I’m genuinely excited for what’s ahead.
> It's about HDR from the perspective of still photography, in your app, on iOS, in the context of hand-held mobile devices.
It's from the perspective of still photography, video, film, desktop computing, decades of research papers, and hundreds of years of analog photography, condensed into something approachable.
> However in the much broader context of HN, a highly technical community whose interests in imaging are diverse, the article's content level and narrow focus aren't consistent with the headline title. It seems written at a level appropriate for novice users.
"On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity"
To be clear, I didn't submit the post, and I never submit my posts. I don't care if my posts make a splash here, and kind of dread when they do because anything involving photography or video attracts the most annoying "well actually" guys on the Internet.
> When I saw the post's headline I thought "Cool! We really need a good technical deep dive into the mess that is HDR - including tech, specs, standards, formats, content acquisition, distribution and display across content types including stills, video clips and cinematic story-telling and diverse viewing contexts from phones to TVs to cinemas to VR."
The post is called, "What is HDR," and the introduction explains the intended audience. That audience is much larger than "people who want to read about ITU-R Recommendation BT.2100." But if you think people are interested in a post like that, by all means write it.
This post is written for people who have heard "HDR" and feel confused. That introduction lists two types of HDR people might think about. "The first" means "the first of two types we're going to explain," not "the first research in the chronological history of HDR."
> Hehe outside is “HDR content”? To me that still comes off as confused about what HDR is.
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
That's from, "Burst photography for high dynamic range and low-light imaging
on mobile cameras," written by some of the most respected researchers in computational photography. It has 342 citations according to ACM.
First sentence: "Tone mapping is a technique used in image processing and computer graphics to map one set of colors to another to approximate the appearance of high-dynamic-range (HDR) images in a medium that has a more limited dynamic range."
> Why did you make the incorrect and obviously silly assumption that I was suggesting a camera’s aperture changes the outdoor scene’s dynamic range rather than what I actually said, that it changes the exposure?
Because you keep bumbling details like someone with a surface level understanding. Your replies are irrelevant, outdated, or flat out wrong. It all gives me flashbacks to working under engineers-turned-managers who just can't let go, forcing their irrelevant backgrounds into discussions.
It's cool that you studied late 90s 3D rendering. So did I. It doesn't make you an expert in computational photography. Please stop confusing people with your non-sequiturs.
You opened this thread arguing that Ansel Adams didn't "use HDR." I linked you to a seminal research paper which argues that he tone mapped HDR content, and goes on to implement a tone mapper based on his approach. This all seems open and shut.
> I’m happy to rescind my critique about Ansel Adams
Great, I'm done.
> and switch instead to pointing out that “HDR” doesn’t refer to the range of the scene
"Surprisingly, daytime shots with high dynamic range may also suffer from lack of light."
"In low light, or in very high dynamic range scenes"
"For high dynamic range scenes we use local tone mapping"
You keep trying to define "HDR" differently than current literature. Not even current— that paper was published in 2016! Hey, maybe HDR meant something different in the 1990s, or maybe it was just ok to use "HDR" as shorthand for when things were less ambiguous. I honestly don't care, and you're only serving to confuse people.
> the aperture can be adjusted on an analog camera to make a scene with any dynamic range fit into the ~12 stops of range the film has, or the ~8 stops of range of paper or an old TV.
You sound nonsensical because you keep using the wrong terms. Going back to your first sentence that made no sense:
> Analog cameras have exposure control and thus can capture any range you want
You keep saying "range" when, from what I can tell, you mean "luminance." Changing a camera's aperture scales the luminance hitting your film or sensor. It does not alter the dynamic range of the scene.
Analog cameras cannot capture any range. By adjusting camera settings or attaching ND filters, you can change the window of luminance values that will fit within the dynamic range of your camera. To say a camera can "capture any range" is like saying, "I can fit that couch through the door, I just have to saw it in half."
> And I’ve used the Reinhard tone mapper in research papers, I’m quite familiar with it and personally know all three authors of that paper. I’ve even written a paper or maybe two on color spaces with one of them.
I'm sorry if correcting you triggers insecurities, but if you're going to make an appeal to authority, please link to your papers instead of hand waving about the people you know.
> The sentence “What if I told you that analog photographers captured HDR as far back as 1857?” is explicitly claiming that analog photographers use “HDR” capture,
No, it isn't. It's saying they captured HDR scenes.
> The result of the juxtaposition is that the article did in fact claim Adams used HDR
You can't "use" HDR. It's an adjective, not a noun.
> Film’s 12 stops is not really “high” range by HDR standards, and a little exposure latitude isn’t where “HDR” came from.
The Reinhard tone mapper, a benchmark that regularly appears in research papers, specifically cites Ansel Adams as inspiration.
"A classic photographic task is the mapping of the potentially high
dynamic range of real world luminances to the low dynamic range
of the photographic print."
> Perhaps it’s useful to reflect on the fact that HDR has a counterpart called LDR that’s referring to 8 bits/channel RGB.
8-bits per channel does not describe dynamic range. If I attach an HLG transfer function on an 8-bit signal, I have HDR. Furthermore, assuming you actually meant 8-bit sRGB, nobody calls that "LDR." It's SDR.
> Analog cameras have exposure control and thus can capture any range you want.
> It seems like a mistake to lump HDR capture, HDR formats and HDR display together, these are very different things.
These are all related things. When you talk about color, you can be talking about color cameras, color image formats, and color screens, but the concept of color transcends the implementation.
> The claim that Ansel Adams used HDR is super likely to cause confusion, and isn’t particularly accurate.
The post never said Adams used HDR. I very carefully chose the words, "capturing dramatic, high dynamic range scenes."
> Previously when you took a photo, if you over-exposed it or under-exposed it, you were stuck with what you got. Capturing HDR gives the photographer one degree of extra freedom, allowing them to adjust exposure after the fact.
This is just factually wrong. Film negatives have 12-stops of useful dynamic range, while photo paper has 8 stops at best. That gave photographers exposure latitude during the print process.
> Ansel Adams wasn’t using HDR in the same sense we’re talking about, he was just really good at capturing the right exposure for his medium without needing to adjust it later.
There's a photo of Ansel Adams in the article, dodging and burning a print. How would you describe that if not adjusting the exposure?
In theory PQ specifies absolute values, but in practice it's treated as relative. Go load some PQ encoded content on an iPhone, adjust your screen brightness, and watch the HDR brightness also change. Beyond the iPhone, it would be ridiculous to render absolute values as-is, given SDR white is supposedly 100-nits; that would be unwatchable in most living rooms.
Bad HDR boils down to poor taste and the failure of platforms to rein it in. You can't fix bad HDR by switching encodings any more than you can fix global warming by switching from Fahrenheit to Celsius.
I don't think that's the colloquial meaning. If you asked 100 people on the street to describe HDR, I doubt a single person would bring up ITU-R BT.2100.
> Does anyone else find the hubris in the first paragraph writing as off-putting as I do?
> "we finally explain what HDR actually means"
No. Because it's written for the many casual photographers we've spoken with who are confused and asked for an explainer.
> Then spends 2/3rds of the article on a tone mapping expedition, only to not address the elephant in the room, that is the almost complete absence of predictable color management in consumer-grade digital environments.
That's because this post is about HDR and not color management, which is different topic.
Human vision has around 20 stops of static dynamic range. Modern digital cameras can't match human vision— a $90,000 Arri Alexa boasts 17 stops— but they're way better than SDR screens.
No. When you simply adjust shadow and highlights, you lose local contrast. In an early draft of the post, there was an example, but it was cut for pacing.
While it isn't touched on in the post, I think the issue with feeds is that platforms like Instagram have no interest in moderating HDR.
For context: YouTube automatically edits the volume of videos that have an average loudness beyond a certain threshold. I think the solution for HDR is similar penalization based on log luminance or some other reasonable metric.
I don't see this happening on Instagram any time soon, because bad HDR likely makes view counts go up.
As for the HDR photos in the post, well, those are a bit strong to show what HDR can do. That's why the Mark III beta includes a much tamer HDR grade.
By definition, a RAW container contains sensor data, and nothing more. Are you saying that Adobe is using their proprietary algorithms to render proprietary RAW formats in Lightroom?
RAW is ultimately about sensor readings. As a developer, you just want to get things from there into a linear, known color space (XYZ in the DNG spec). So from that perspective, interoperability isn’t the issue.
How you process that data is another matter. Handling a traditional bayer pattern vs a quad-bayer vs Fujifilm’s x-trans pattern obviously requires different algorithms, but that’s all moot given DNG is just a container.
> Camera sensors from different companies (and different generations) don't have the same color (or if you prefer, spectral) responses with both their Bayer filter layer and the underlying physical sensor
This is all accommodated for in the DNG spec. The camera manufacturers specify the necessary matrix transforms to get into the XYZ colorspace, along with a linearization table.
If they really think the spectral sensitivity is some valuable IP, they are delusional. It should take one Macbeth chart, a spreadsheet, and one afternoon to reverse engineer this stuff.
Given that third party libraries have figured this stuff out, seems they have failed while only making things more difficult for users.
Raw decoding is an algorithm, not a container format. The issue is every is coming up with their own proprietary containers for identical data that just represents sensor readings.
Fun fact: the HDR photos are actually looping videos, because that’s way more reliable for browsers and CMS platforms than actual HDR photos. Hopefully this improves in the next few years.