Can we see HDR wider than our perceptions?

[zoossh]

this question is conceptual.

ok, i read somewhere that most sensors and films handle about 6 stops difference whereas our eyes can see about 8 stops. HDR can push the limits up to 12 stops.

ok, assuming that i did a HDR of 12 stops difference in tonal intensity, and print it out, i assume the printer can print that difference, if i look at it, can i see tell the difference of 12 stops difference from another HDR at 8 stops difference visually? since our eyes see 8 stops difference.

if our eyes can't tell 8 stops difference, having a latitude higher than that is not of visual importance but only of detail integrity on post processing, am i rite?

pls correct my understanding if i'm wrong.

 

[jmmtn4aj]

I think the paper would have to be producing light to have that much dynamic range..

 

[SnapJaX]

hahas..funny answer, yet makes sense..

 

[zoossh]

I think the paper would have to be producing light to have that much dynamic range..

it does reflects light rite, isn't it?

if not, how about monitor and displays?

 

[roygoh]

this question is conceptual.

ok, i read somewhere that most sensors and films handle about 6 stops difference whereas our eyes can see about 8 stops. HDR can push the limits up to 12 stops.

ok, assuming that i did a HDR of 12 stops difference in tonal intensity, and print it out, i assume the printer can print that difference, if i look at it, can i see tell the difference of 12 stops difference from another HDR at 8 stops difference visually? since our eyes see 8 stops difference.

if our eyes can't tell 8 stops difference, having a latitude higher than that is not of visual importance but only of detail integrity on post processing, am i rite?

pls correct my understanding if i'm wrong.

I think you are over simplifying the human visual perception. Not sure where you get the 8 stops from, but what I read was that without changing the pupil size (our aperture) the human eye has about 30 dB dynamic range. This refers to the sensors on the retina.

Human visual perception is more complicated. Unlike a photograph that is taken at a single exposure setting, human vision is dynamic. Exposure is constantly adjusted (pupils, sensitivity), and we can only capture details in a very small area of our FOV, so we integrate entire image by scanning. There are also 2 kinds of sensors, like the new fujifilm sensor, except that the high sensitivity sensors in our eye is only monochrome. Also, there is also some kind of dynamic range compression going on in our brain plus all kinds of other processing that in some cases can generate illusions instead of showing us what we are really looking at.

Because of all these dynamics our visual system can definitely see more than 8 stops, if not the eyes alone.

You don't have to get such a print. Just go out on a sunny day and you will notice that you can see details in the sun as well as in shady areas, while a digital camera shooting in 12-bit RAW will still have problem capturing both the highlights and the shadow details.

My favourite optical illusion is the checkerboard illusion:
http://www.popularscience.co.uk/features/feat16.htm

If we can percieve such a big difference in the brightness between sqaurea A and B even when they have the same RGB values on the monitor, then it is really hard to measure the dynamic range of the human visual perception.

 

[michhy]

ok, i read somewhere that most sensors and films handle about 6 stops difference whereas our eyes can see about 8 stops. HDR can push the limits up to 12 stops.

while the HDR process creates an illusionary 12-stop image, the final print/output is just like any other normal image to our eyes, so of course we can see it! you really know what HDR does to a photograph, right?

 

[hazmee]

My favourite optical illusion is the checkerboard illusion:
http://www.popularscience.co.uk/features/feat16.htm

If we can percieve such a big difference in the brightness between sqaurea A and B even when they have the same RGB values on the monitor, then it is really hard to measure the dynamic range of the human visual perception.

Man I had a good read and also great laugh reading that optical illusion. :bigeyes: That was a good one roygoh! Damn, my eyes have been playing tricks on me all along. :bsmilie:

 

[zoossh]

while the HDR process creates an illusionary 12-stop image, the final print/output is just like any other normal image to our eyes, so of course we can see it! you really know what HDR does to a photograph, right?

actually i think i haven't grasp the idea, that's why i ask....

 

[Ansel]

Don't forget our eyes' gamma are different from the sensors too.

 

[Ansel]

actually i think i haven't grasp the idea, that's why i ask....

A HDR file is a high bandwidth file that stores a lot of data for a wide dynamic range image. How the DR is displayed is dependant on your display medium or device. No display device or medium today can fully reproduce the dynamic range that is recorded in a HDR file, or the range that our eyes are capable of discerning.

 

[denmad]

actually i think i haven't grasp the idea, that's why i ask....

michhy is correct. What HDR involves is take 2 or more photos of the same scene of differing exposure level and combine them into a single photo. When people say that a HDR photo reproduces 12 stops, it actually means that the details present in the HDR photo has a 12 stop range, not that the photo itself has 12 stop range; it still has a range of 6 stops.

For example, if you take 2 photos of the same scene, 1 at -3 stops and 1 at +3 stops (middle value), you get 2 photos, one of -6 stops to 0 stop, and the other of 0 stop to 6 stops, assuming that you are using film with 6 stops range. In essence, you have captured 12 stops of details, but in 2 photos. When you combine the 2 photos to form an HDR photo, it is as if you took the -3 stops photo and pushed it up 3 stops, and you took the +3 stops photo and pushed it down 3 stops, which results in a photo still with only 6 stops range.

The easiest way to think about this is that the HDR photo is not what you see in the original scene with your own eyes.

 

[zoossh]

A HDR file is a high bandwidth file that stores a lot of data for a wide dynamic range image. How the DR is displayed is dependant on your display medium or device. No display device or medium today can fully reproduce the dynamic range that is recorded in a HDR file, or the range that our eyes are capable of discerning.

which means that

our eyes+brain sees more than the HDR while HDR sees more than the display or print

eyes+brain > HDR data > eyes alone, sensor alone, printout of anything including HDR output

 

[zoossh]

michhy is correct. What HDR involves is take 2 or more photos of the same scene of differing exposure level and combine them into a single photo. When people say that a HDR photo reproduces 12 stops, it actually means that the details present in the HDR photo has a 12 stop range, not that the photo itself has 12 stop range; it still has a range of 6 stops.

For example, if you take 2 photos of the same scene, 1 at -3 stops and 1 at +3 stops (middle value), you get 2 photos, one of -6 stops to 0 stop, and the other of 0 stop to 6 stops, assuming that you are using film with 6 stops range. In essence, you have captured 12 stops of details, but in 2 photos. When you combine the 2 photos to form an HDR photo, it is as if you took the -3 stops photo and pushed it up 3 stops, and you took the +3 stops photo and pushed it down 3 stops, which results in a photo still with only 6 stops range.

The easiest way to think about this is that the HDR photo is not what you see in the original scene with your own eyes.

this part i understand. the question is how wide the sensors, the printout, the computer display can show. i'm curious about that.

and in real-life, i wonder how wide the subject contrast really differs, assuming we are looking into a light source that our eyes need not squint.

 

[denmad]

this part i understand. the question is how wide the sensors, the printout, the computer display can show. i'm curious about that.

and in real-life, i wonder how wide the subject contrast really differs, assuming we are looking into a light source that our eyes need not squint.

Well, with regards to how much dynamic range sensors can handle, printers can print out, and monitors can display, I guess you answered it in your first post, which is about 6 stops.

As for nature, I do not know the numbers, but I can imagine that it is many, many, many stops. Just take you camera, record a scene with 1/4000s to 10s in 3 stops interval and you can see for yourself how much dynamic range a typical scene has.

 

[theRBK]

this part i understand. the question is how wide the sensors, the printout, the computer display can show. i'm curious about that.

and in real-life, i wonder how wide the subject contrast really differs, assuming we are looking into a light source that our eyes need not squint.

basically, in equation form,

HDR > eye > display, print

of course, when I say HDR > eye, I am saying that it potentially can show wider range than eye can see...but of course that depends on how wide the dynamic range the image is composed of...potentially you can squeeze from infrared to ultraviolet or even more, beyond the range of wavelength that the eye can see... but that is going abit extreme...

visual representation of HDR at the moment is restricted to display and print technology... whatever HDR image we have, the dynamic range has to be squeezed into the range that a display or print can produce... and that is definitely less than what the human eye can see at the moment... so basically, the HDR representation we can see is restricted to what can be displayed on screen or in print

so basically

eye > display > print

 

[denmad]

of course, when I say HDR > eye, I am saying that it potentially can show wider range than eye can see...but of course that depends on how wide the dynamic range the image is composed of...potentially you can squeeze from infrared to ultraviolet or even more, beyond the range of wavelength that the eye can see... but that is going abit extreme...

Hi theRBK, just to avoid further confusing the TS, dynamic range and wavelength range are two separate things. Dynamic range refers to the difference between the highest intensity and the lowest intensity, so it doesn't matter whether it is visible light, infrared light or ultraviolet light we are talking about.

 

[theRBK]

Hi theRBK, just to avoid further confusing the TS, dynamic range and wavelength range are two separate things. Dynamic range refers to the difference between the highest intensity and the lowest intensity, so it doesn't matter whether it is visible light, infrared light or ultraviolet light we are talking about.

sorry, was just trying to describe what the eye can and cannot see and got carried away...;p

 

[Yatlapball]

Our eyes and brain is the most advanced imaging system available in the world today. Heh. You really cannot duplicate its effect. It comes built in with auto-focus, auto aperture, auto sensitivity, auto everything. Heck it even pixel maps on its own (read article on blind spot in our vision). :bsmilie: I look forward to the day there is a camera like that.

I think for excessive HDR might cause the photo might look extremely odd because every part of the photo would be correctly exposed. Which in reality isn't the case. Our eyes just happens to "expose correctly" the exact point that we are looking at.