Go Back   Steve's Digicams Forums > Post Your Photos > HDR

Reply
 
Thread Tools Search this Thread
Old Mar 6, 2010, 3:38 PM   #21
Senior Member
 
VTphotog's Avatar
 
Join Date: Mar 2005
Location: Extreme Northeastern Vermont, USA
Posts: 4,214
Default

[QUOTE=Chato;1061104]Technically, HDR is a 32 bit, as opposed to an 8 bit image (In this case, 8 and 32 bits refers to each channel). Few monitors and no printers are capable of showing the actual image.
HDR Photography explained, definition and realization

Your source is incorrect in his basic assumption that the full dynamic range of a scene cannot be captured by 8-bit sensor. In fact, you could use fewer bits and still cover the full range. The only difference in 8-bit and 32-bit is in the definition of the steps between the highest and lowest levels. In converting from, for example, a 12-bit sensor, to 8-bit jpeg, many cameras will maintain the bit-level values, and clip either highlights or shadows, or both, depending on the scene's range. Some cameras now have algorithms to compress the values to retain the full sensor range in the 8-bit jpeg output.

As I mentioned in another thread, HDR is not viewable on most monitors which we mortals use. Nor can it be printed as such with the technology we have access to. Nor does it fit within the confines of 8-bit jpeg without compression to fit the 8-bit container.

This is what HDR, in the context of this forum, is. A High dynamic range image which has been compressed to fit within 8-bit jpeg parameters.


Sensors now can resolve up to about ten (or more) stops.


Brian
VTphotog is offline   Reply With Quote
Old Mar 6, 2010, 3:45 PM   #22
Senior Member
 
Join Date: Aug 2009
Location: NYC
Posts: 1,990
Default

Quote:
Originally Posted by Bynx View Post
After shooting something as simple as stained glass windows I believe HDR approach will improve any given lighting situation no matter what it is. It will never make it worse. Some situations like a dull overcast day with very weak shadows will not show the effects of the HDR as much as a bright sunny days with deep dark shadows but there will be some improvement. Light is so subtle and so complete from absolutely blinding white the the deepest darkest blacks. And present equipment cannot record it in all its glory.
The above is the position that I have trouble with. I've been working to give a demonstration of my disagreement, but realise that in order to truly give that demonstration I would have to send you my original RAW file - Something I am prepared to do. Even then, it's only a one shot file - One exposure. I worked Photomatrix and worked Photomatrix, and my normal shot was better. Not much better, but better. This is because there is only an itty, bitty tiny place where the highlights are in danger of being blown.

If you like, I'll post the URL of the DNG file, which I cropped in Bridge so that it's about 11 Megs..

Dave
Chato is offline   Reply With Quote
Old Mar 6, 2010, 3:48 PM   #23
Senior Member
 
Join Date: Aug 2009
Location: NYC
Posts: 1,990
Default

[QUOTE=VTphotog;1061755]
Quote:
Originally Posted by Chato View Post
Technically, HDR is a 32 bit, as opposed to an 8 bit image (In this case, 8 and 32 bits refers to each channel). Few monitors and no printers are capable of showing the actual image.
HDR Photography explained, definition and realization

Your source is incorrect in his basic assumption that the full dynamic range of a scene cannot be captured by 8-bit sensor. In fact, you could use fewer bits and still cover the full range. The only difference in 8-bit and 32-bit is in the definition of the steps between the highest and lowest levels. In converting from, for example, a 12-bit sensor, to 8-bit jpeg, many cameras will maintain the bit-level values, and clip either highlights or shadows, or both, depending on the scene's range. Some cameras now have algorithms to compress the values to retain the full sensor range in the 8-bit jpeg output.

As I mentioned in another thread, HDR is not viewable on most monitors which we mortals use. Nor can it be printed as such with the technology we have access to. Nor does it fit within the confines of 8-bit jpeg without compression to fit the 8-bit container.

This is what HDR, in the context of this forum, is. A High dynamic range image which has been compressed to fit within 8-bit jpeg parameters.


Sensors now can resolve up to about ten (or more) stops.


Brian
I don't understand your point.

Dave
Chato is offline   Reply With Quote
Old Mar 6, 2010, 3:55 PM   #24
Senior Member
 
VTphotog's Avatar
 
Join Date: Mar 2005
Location: Extreme Northeastern Vermont, USA
Posts: 4,214
Default

If you are using the article as a definition, and it is flawed, further reasoning from that definition is flawed also.

The absolute levels of light which can be captured are not determined by the bit resolution of either the sensor or the output format.

brian
VTphotog is offline   Reply With Quote
Old Mar 6, 2010, 4:05 PM   #25
Senior Member
 
Join Date: Aug 2009
Location: NYC
Posts: 1,990
Default

Quote:
Originally Posted by VTphotog View Post
If you are using the article as a definition, and it is flawed, further reasoning from that definition is flawed also.

The absolute levels of light which can be captured are not determined by the bit resolution of either the sensor or the output format.

brian
You are going to have to explain to me why a RAW file with 16 (or 12) bits per channel, can reveal the blown highlights that cannot be restored with an 8 bit shot. While no scientist, the explanations I've read simply state that the more bits per channel, the more data is captured. If such is the case, you Really must explain this...

No doubt, some of the limitations are simply the limitations of our present technology - But all things being equal, 16 bits per channel (let alone 32 bits per channel) will always have more data. At some point in the future this will become academic, (since the human eye, from a practical point of view, is the determining factor of what we can analyze in an image) but it's not academic today.

Dave
Chato is offline   Reply With Quote
Old Mar 6, 2010, 5:45 PM   #26
Senior Member
 
VTphotog's Avatar
 
Join Date: Mar 2005
Location: Extreme Northeastern Vermont, USA
Posts: 4,214
Default

Ok, one last try:
A stairway may reach a level of ten feet with 8 steps or with thirteen steps. The difference is not in the absolute level of the bottom stair or the top stair, but the increment. Same with sensors. This is simply an analog to digital conversion, from 0 to x. The finer resolution of a larger bit depth may provide cleaner looking output due to the smaller 'steps' but doesn't have anything to do with the absolute light levels being recorded.

In many ways, what the HDR programs are doing is reinterpolation of the intermediate values contained in 8-bit images, and then re-compressing to 8-bits - all within the original light and dark limits.

brian
VTphotog is offline   Reply With Quote
Old Mar 6, 2010, 8:11 PM   #27
Senior Member
 
Join Date: Aug 2009
Location: NYC
Posts: 1,990
Default

Quote:
Originally Posted by VTphotog View Post
Ok, one last try:
A stairway may reach a level of ten feet with 8 steps or with thirteen steps. The difference is not in the absolute level of the bottom stair or the top stair, but the increment. Same with sensors. This is simply an analog to digital conversion, from 0 to x. The finer resolution of a larger bit depth may provide cleaner looking output due to the smaller 'steps' but doesn't have anything to do with the absolute light levels being recorded.

In many ways, what the HDR programs are doing is reinterpolation of the intermediate values contained in 8-bit images, and then re-compressing to 8-bits - all within the original light and dark limits.

brian
Ahh, I think I now understand your point. From practical experience of working with RAW, whether you are technically "right" or not, in practice yours is a meaningless distinction. You bet, we can capture white to black with our present technology, but the problem with range here is that our cameras, faced with an unbalanced scene, will clip either highlights or shadow. The same image shot in RAW allows less clipping. And the technique of HDR allows the restoration of lost data, even when we use RAW.

You seem to be implying that there is no lost data, but in practical terms, there certainly IS lost data, because larger or smaller steps, is simply another way of saying, "What the hell happened to my highlights? I blew the shot!!! (Whine snivel, moan)...

Now for your information HDR Does have a scientific definition and that means 32 bit (per channel) images.

The link I posted doesn't disagree with you at all, it simply points out that 256 steps is NOT enough to always capture all the graduations...

Quote:
Well, any scene should be divisible in 256 levels, but that depends directly on the dynamic range that the sensor is able to stand, and thus on the technology of the current sensors. For example, if you have shadows in a fully sunny scene, the sensors (even best films) won't be able to "divide” such high variation of contrast into 256 levels. From a technical point of view, while recording the very low number of photons emitted by the shadows of the scene, the photosites of the sensor (or silver grains of a film) should be able to record at the same time an enormous quantity of photons in the high-lights. Today, such a quantity of photons “overflows” of the photosites, they saturate, and thus records a pure white. It's even more obvious on ultra compact cameras, which have very tiny photosites. Maybe one day we will have some photosites able to be emptied and record continuously during one exposure time, but that's still far … Thus, in such a scene, on one hand you increase exposure time to capture the shadows, saturating high-lights, on the other hand you decrease exposure time to preserve high-lights, but then shadows will be black, because too little (or no) photons emitted by the shadows will hit the sensor in such a short time !
Dave
Chato is offline   Reply With Quote
Old Mar 7, 2010, 3:36 AM   #28
Senior Member
 
MartinSykes's Avatar
 
Join Date: Feb 2010
Location: Cheshire, UK
Posts: 854
Default

Quote:
Originally Posted by Chato View Post
Now for your information HDR Does have a scientific definition and that means 32 bit (per channel) images.
While I agree that is a good definition - I'm all in favour of definitions that are objective and unambiguous - language like everything evolves and I think HDR has come to mean something different in practical terms.

If we could display 32-bit images direct on a printer or monitor I don't think they would bedescribed by most people to be HDR. I think a more general use of the term would be to describe image that have undergone any of the various processes that amounts to local tone mapping.

To slightly change the stairs analogy, imagine a ladder which is so tall that you can't see it all in one go. You can label the rungs 'black' at the bottom and 'white' at the top. A 'normal' technique would squash the ladder so that the rungs were closer together and the whole thing was shorter. An HDR technique would cut it into pieces and put them side by side.
MartinSykes is offline   Reply With Quote
Old Mar 7, 2010, 10:27 AM   #29
Senior Member
 
Join Date: Aug 2009
Location: NYC
Posts: 1,990
Default

Quote:
Originally Posted by MartinSykes View Post
While I agree that is a good definition - I'm all in favour of definitions that are objective and unambiguous - language like everything evolves and I think HDR has come to mean something different in practical terms.

If we could display 32-bit images direct on a printer or monitor I don't think they would bedescribed by most people to be HDR. I think a more general use of the term would be to describe image that have undergone any of the various processes that amounts to local tone mapping.
Err, yes...

Steve's Digicams Forums - View Single Post - HDR Philosophical Discussion

Dave
Chato is offline   Reply With Quote
Old Mar 8, 2010, 5:47 AM   #30
Senior Member
 
bahadir's Avatar
 
Join Date: Jan 2006
Location: Izmir, Turkey
Posts: 6,263
Default

@ Ordo: I'd like you to know that I quite appreciated your analogy, which has made me review a thesis I wrote years ago on the unity of sound, form and colour, now with the hope of making an independent article out of it : )

......................................

As for MartinSykes's question '' If you take 3 exposures of a scene and the middle one doesn't contain any blown highlights or clipped shadows because the scene isn't that extreme, then what value do the other exposures add?''
I think the middle exposure (rather than a perfect exposure, which requires awareness about the perfect exposure!) should be exposed in such a way to make the next shot easily blow the highlights while retaining more info from shadows, consequently making the underexposed shot useful for highlight detail while avoiding pitch black shadows.
Well, surely there should an end of every rational act. Now, our gain would be an even finer gradual shading (than a single shot) giving more voluminous (other than the meaning of being huge) feel to objects almost catching up what our eyes are capable of seeing.
That being said, even with a scene containing harsh contrast, depending on the effect I desire, I sometimes suffice with a single shot deliberately! There are times I prefer the light breaking the forms as in baroque painting the following of which is a good example, http://employees.oneonta.edu/farbera..._Carpenter.jpg
and there are times as in renaissance paintings in which I prefer the light unifying and modelling the forms as tangible....All in good time, at least, for me : )

.................................................. ........

As inferred above, I , like the folks here, welcome any method that works for my intended purpose. Yet, when it comes to arranging my toolbox, I'd hardly agree with an opinion which equates or compares 'sharpening' to 'hdr' though I invariably use different sharpening methods for the final output. I don't take shoot a picture to be sharpened, but I originally take pictures to be 'hdr'ed'! In short, all squares are quadrilaterals but not all quadrilaterals are squares ; )
I shoul also say I agree with Bynx in one thing, which is practice makes perfect! There's no magic formula which fits all in HDR through the software only. Before setting one's camera to auto bracketing one should take an initial shot to determine how many frames at which exposure spacing is required for the scene as well as how much lighter should the 'middle shot' be than the perfectly exposed shot, like determining the range of a catapult

Cheers,

Bahadır











bahadir is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 5:46 AM.