Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

HDR Mode...

hmm..but these look too surreal..what happened to the shadows..?

Could be great for certain movie look..but seems abit extreme.

Here's a quick HDR+tonemapping I did... all of the 5 original photos look bad/average on their own. Stuff is blown out or the shadows have noise.

So to take this into context, if I was shooting a movie here in my basement (fyi the drinks aren't mine lol), I could save money and time without having to light the interior so the outside doesn't get blown out. Although it is just concrete in this case haha. This is of course assuming HDR-video would be tonemappable like HDR-photos.

LDR (best of the 5):


HDR (combined 5 and tonemapped):
 
You can basically think of HDR as the ULTIMATE RAW. Latitude and color corrections that you never could imagine are possible with float point 32bpc.

Remember folks... keep repeating this to yourself and to all the 'pop gurus' out there on the internets: HDR is NOT tonemapping. Tonemapping is the only way to SQUASH the massive data set back down to our 8bpc world we live in right now to get a representation of the latitude contained within.
 
With around 13 stops DR in a single exposure, I think some sort of tone mapping with the flexibility to combine several FLUT™ settings into the final image would be extremely useful. The goal for a realistic representation of how we actually see such a range is still around 7-8 stops of DR for current display technologies, even 10 bit. The dynamic contrast settings for current gen consumer displays don't look natural at all. For the big screen maximum theoretical contrast from a projector in a theater is around 1000:1. Most theaters typically fall under 200:1 with film and under 400:1 with digital.
 
A lot of people think of tone-mapping when HDR comes up. We don't. We think of high dynamic range.

Jim

Yes, this is how I think of HDR too. I was very excited when it was announced, but the tone mapping examples that I saw posted threw me off and I thought maybe I didn't know what HDR was so I kept quiet.

The latest example by Chris Clogg is a great example. And this too: http://www.cambridgeincolour.com/tutorials/high-dynamic-range.htm

HDR in full motion seems like an end game scenario as far as who has the best dynamic range in a motion movie camera.
 
With around 13 stops DR in a single exposure, I think some sort of tone mapping with the flexibility to combine several FLUT™™ settings into the final image would be extremely useful. The goal for a realistic representation of how we actually see such a range is still around 7-8 stops of DR for current display technologies, even 10 bit. The dynamic contrast settings for current gen consumer displays don't look natural at all. For the big screen maximum theoretical contrast from a projector in a theater is around 1000:1. Most theaters typically fall under 200:1 with film and under 400:1 with digital.


Still., we look at the above photo of a sink, and we see that it looks natural on an 8 bit compu display, and better than if no tonemapping had been used.
 
Which has been true in the past, but having a sensor that has less noise means that the range of usage for "Found light" filming will be different now. But, to even take that a bit further, also given the sensitivity and lower noise of the M-X sensor will mean that for a lot of shooting, you wont need 5 Billion Watt lights to setup the scene. Now to even take that even another step further, with HDR capability, you'll have a little more freedom in how a scene is lit even outside since you won't be losing as much detail due to the differences in the scene.

Technologies like the M-X sensor not only improve visual quality, eventually, if they are that good (and it seems to be from what we have seen), they change work flows or requirements in setting up scenes. This is why I am so excited about this technology. It opens up your window of ability in acquisition and gives more freedom to the cinematographer to work outside the bounds of what has been available previously.

Like, when all anyone shot was black and white film. Then along came color. I see these new abilities, while not as massive an enhancement as that, as analogous in terms of new thinking.

Though I will say one thing, nothing will replace needing light in terms of light as a part of the art form of the shot. The light is an illustrative device as everyone knows in painting the scene and putting focus and mood and feeling. Not to mention bringing things together. It's just now, you can do it a little darker if you want :)

Well put. Point taken.
 
It's funny, I am the exact opposite. I feel that most films are lit in a way that makes them look fake, yes "Hollywood Magically" but hardly realistic. I saw the girlfriend experience and I thought that the cinematography was great and I often felt like I was right there. It did not flop because it looked bad but because no one knew about it, it had no notable actors and any press it did get was centered around the fact that it was a semi mainstream film centered around a prostitute played by a real life hardcore pornstar. Anyone who did know about it stayed away from it for the above reasons. The artistic decision to film with available light was one of the films appeal and if you read reviews, critics either liked the look of did not seem to mind. Films like the Hurt Locker and The Thin Red Line prove that available light can be just as good if not better than staged lighting in the right situations. I wish more would experiment with it. I don't know how many times I go to the theaters and say to myself, "In a real situation, the light would never look like that". Again, most don't notice it but I do and it can get a little frustrating. Roger Deakins is the best when it comes to making artificial light look real. He often complements available light with artificial to maintain a realistic look that is somewhere in the middle, not perfectly realistic (documentary) or overly forced (Any film by Michael Bay).

Michael Bay seems to have been successful in getting butts into seats. I am all for choices. It's great that Hurt Locker can get done, but also part of freedom that I neither want to see that film nor would enjoy working on it. Besides, if all films looked that gritty, they would no longer appear gritty, they would just appear ordinary.

Again, your "real" is my depressing and off-putting. I could care less about critics - most strike me as a bunch of out-of-touch academics who do not impress at all, and I almost always disagree with most of them. That too is my opinion. It is because we all disagree that we have such a wonderful rainbow of styles being created every day.
 
Still., we look at the above photo of a sink, and we see that it looks natural on an 8 bit compu display, and better than if no tonemapping had been used.

Exactly my feeling. I hear a lot of people saying that tone mapping and HDR aren't the same, and they are not, but the basic principles are absolutely tied together for any practical application.

Maybe the 'definition' of tone mapping means something different for other people, and how I have seen it used has been varied. Maybe there should be a new term to describe what is needed. Maybe it should be called dynamic range scaling, because that is what you are really doing. It's not about changing the tone, but squeezing that dynamic range into a smaller range.

Call it what you want, but if you have 14-bits of light depth and your display has 8-bits, you WILL be either showing only a portion of that 14-bit depth OR you will scale that data into your visual technology. How you scale is definitely a user choice, linear, non-linear, truncated, whatever, but it's going to happen or it's useless.
 
Still., we look at the above photo of a sink, and we see that it looks natural on an 8 bit compu display, and better than if no tonemapping had been used.

Agreed..the first pic looks like how human eye would see the scene.
 
Agreed..the first pic looks like how human eye would see the scene.

Which also brings up a point, sometimes you want to get a result that even your eye couldn't perceive. It's not that you want the result to look unnatural, but maybe you are out filming wildlife and the conditions of light are just so harsh that that extra DR gets you more detail in shadows you can bring back together later. No, it wasn't what was really there, but you will artistically wrangle in the range to present something more palatable visually for the audience.

Not everything shot has control of the lighting or the different elements of the scene. It's for these conditions that HDR will really help make usable acquisition more frequent for "found footage" if you will. :)
 
Agreed..the first pic looks like how human eye would see the scene.

I think he was actually saying that the 2nd pic looked more like what the human eye would see (at least that's my interpretation). And by the way, the tonemapped-HDR one IS basically what the human eye sees. I took the photo, and the tonemapping made it much closer to what I was actually seeing.
 
Agreed..the first pic looks like how human eye would see the scene.

Well, no, it doesn't. That's just the point. The second pic looks like how the eye would see it. The human eye is amazingly adaptive; zooming, focussing and racking iris so fast and intuitively you don't even know it's happening. This is one of the main challenges for any photographic medium, be it film or electronic, to render a given scene as the eye would - and, so far, it's been impossible to do any more than approach it without lighting.

HDR doesn't mean weird-looking images. Ansel Adams proved that. Sure, like so much other technology, it can be done well, or it can be done badly or it can be used to achieve something that, if not realistic, represents one person's creative vision. It's simply another tool. How an individual uses that tool is totally up to the individual.
 
Both the retinal circuitry, the thalamic relay and visual cortex all perform algorithms to map local luminance into local contrast. HDR and tone mapping approximates more closely what the visual system does than not. The blown out highlights and limited dynamic range of photography are much more of an aberration, yet orthodoxy and our Bayesian priors (our visual experience) teaches us that it is more "normal" in photography. Much as the cadence of 24fps gives a "cinematic" sense that 120fps does not.
 
Both the retinal circuitry, the thalamic relay and visual cortex all perform algorithms to map local luminance into local contrast. HDR and tone mapping approximates more closely what the visual system does than not. The blown out highlights and limited dynamic range of photography are much more of an aberration, yet orthodoxy and our Bayesian priors (our visual experience) teaches us that it is more "normal" in photography.

What he said... I think.:blushing:
 
Both the retinal circuitry, the thalamic relay and visual cortex all perform algorithms to map local luminance into local contrast. HDR and tone mapping approximates more closely what the visual system does than not. The blown out highlights and limited dynamic range of photography are much more of an aberration, yet orthodoxy and our Bayesian priors (our visual experience) teaches us that it is more "normal" in photography. Much as the cadence of 24fps gives a "cinematic" sense that 120fps does not.

The clear solution here is for Red to reverse engineer the organic circuitry of an eyeball. If anyone can do it, they can :) Would give a whole new meaning to "Red Eye" :) I can see the slogans now:

"The only time you want Red Eye in a photo is with Red's new Eye technology!"
 
The blown out highlights and limited dynamic range of photography are much more of an aberration, yet orthodoxy and our Bayesian priors (our visual experience) teaches us that it is more "normal" in photography. Much as the cadence of 24fps gives a "cinematic" sense that 120fps does not.

Yes..thank you..that's what I was trying to say.

We know the sun is the brightest source of light in our world from experience so we are willing to tolerate quite abit of "blow out" in situations like exterior window views.
Also I see the wall in the pic is actually right up on the window(my bad) and not really the traditional exterior view out the window. thanks..it's good discussion on tone mapping in these posts.
 
The clear solution here is for Red to reverse engineer the organic circuitry of an eyeball. If anyone can do it, they can :) Would give a whole new meaning to "Red Eye" :) I can see the slogans now:

"The only time you want Red Eye in a photo is with Red's new Eye technology!"

Would I? Red Eye!
 
Both the retinal circuitry, the thalamic relay and visual cortex all perform algorithms to map local luminance into local contrast. HDR and tone mapping approximates more closely what the visual system does than not. The blown out highlights and limited dynamic range of photography are much more of an aberration, yet orthodoxy and our Bayesian priors (our visual experience) teaches us that it is more "normal" in photography. Much as the cadence of 24fps gives a "cinematic" sense that 120fps does not.

Extremely good point. 3D is also revealing new challenges in the replication of normalcy - in the future we might have new definitions of "3d-viewing normal" and "natural XBox Project Natal interfaces" as we do now for "normal looking" photography.
 
Exactly my feeling. I hear a lot of people saying that tone mapping and HDR aren't the same, and they are not, but the basic principles are absolutely tied together for any practical application.

Maybe the 'definition' of tone mapping means something different for other people, and how I have seen it used has been varied. Maybe there should be a new term to describe what is needed. Maybe it should be called dynamic range scaling, because that is what you are really doing. It's not about changing the tone, but squeezing that dynamic range into a smaller range.

Call it what you want, but if you have 14-bits of light depth and your display has 8-bits, you WILL be either showing only a portion of that 14-bit depth OR you will scale that data into your visual technology. How you scale is definitely a user choice, linear, non-linear, truncated, whatever, but it's going to happen or it's useless.

Tone mapping is like color-correction - pushed just enough and it looks enhanced but natural - pushed too far and it looks like you're a low-budget wedding videographer who just bought his first copy of Magic Bullet.
 
Back
Top