kevindar wrote:
and here is one with exposure blending. NOthing HDR about it. allowed me for a brighter looking light house, and selectively not only darkening the sky, but decreasing the reflection off of the water
Kevin, that is exactly what I just posted about. By taking 2 or more exposures and blending them together is HDR. You can`t get around that but it is not tone-mapping. These will produce totally different results
1. Anytime you get more DR than people expect from a photograph, be it blends or ND grads or even a camera with more DR, people will think either HDR or that something has been manipulated. The name matters little unless you are a photographer, it's the perception that matters.
2. The other problem in my opinion is with peoples expectations. The camera does not capture the DR or for that matter the DOF we see in person, but what we expect is based on prior experience with photographs and the limits imposed.
What I think we need to do as photographers is to inch up on this. Be slow to add DR and wait until it is more accepted. And also be careful not to exceed reality unless you are making an artistic statement.
Scott Kroeker wrote:
See for me any time you take more then one photo to capture DR is HDR. The problems is that so many people have confused tone-mapping with HDR. IMHO they are 2 different things but so much misinformation and wrong terminology used on-line we have a backwards perception of the term HDR. HDR today is mostly associated with over tone-mapped and cooked to death images. The to me, they should be categorized as tome-mapped not necessarily HDR as you can tone map any image. But to try and get the masses to change their way of using this terminology is pretty much a lost cause at this stage. Which is sad because HDR has gotten such a bad rap because of grotesquely tone-mapped images. ...Show more →
Scott, that may be your definition, and I can understand your logic, but the term HDR came into existence recently (after exposure blending had been around for quite a while) specifically to refer the the technique in which the software does a sort of luminosity averaging across the scene based on algorithms that operate on the entire image while being weighted towards localized area to increase local contrast.
I'm sure I'll get push-back from at least a few people, but I think that it makes the issue simpler to think of two things:
1. Exposure blending - which is a sort of super-charged version of what we used to do with GND filters.
2. HDR - which is an analytic software process that increases local contrast using data from multiple captures.
This is not meant as a judgment of the value or appropriateness of the techniques, but rather an attempt at clarifying something that often gets confusing.
Scott Kroeker wrote:
But, taking 2 or more exposures for DR and then manually blending (painting) them together in photoshop is still HDR as you have captured higher dynamic range then you could in one image.
But Scott, the use of the GND also lets you "capture a higher dynamic range" than you could in a normal single image.
Let's say I'm shooting the classic "bright sky above darker foreground image." In Case A I use a two stop GND (let's say) with somewhat narrow transition so that half of the image is effectively shot two stops differently than the other half. In Case B I make two exposures separated by two stop and then combine the two images using a mask with a somewhat narrow transition so that the result is the same.
I don't see how one is HDR by your definition and the other isn't.
Of course, if you want to play our your logic all the way, a lot of things that are done to handle large dynamic ranges would also have to be regarded as HDR. Dodge the shadows a bit in the darkroom? HDR. Use a CP filter to darken sky or control reflections? HDR. Burn down the highlights? HDR.
gdanmitchell wrote:
Scott, that may be your definition, and I can understand your logic, but the term HDR came into existence recently (after exposure blending had been around for quite a while) specifically to refer the the technique in which the software does a sort of luminosity averaging across the scene based on algorithms that operate on the enter image to increase local contrast.
I'm sure I'll get push-back from at least a few people, but I think that it makes the issue simpler to think of two things:
1. Exposure blending - which is a sort of super-charged version of what we used to do with GND filters.
2. HDR - which is an analytic software process that increases local contrast using data from multiple captures.
This is not meant as a judgment of the value or appropriateness of the techniques, but rather an attempt at clarifying something that often gets confusing.
Whether or not we can all agree upon the deeper/truer/more accurate definition of HDR, what Dan has described is how most photographers with some experience in this area view things. That not to say it won't change over time as HDR technology advances. But I think this is the current way of differentiating and defining these approaches. There are in fact many discussions, debates, etc... regarding Digital Blending vs. HDR techniques/tool. I might add that 'HDR' normally also refers to a more 'automated' process vs. 'manual' blending.
One thing new which changes the playing field a bit at least is the new HDR function built-in to some higher end cameras like the new Canon 5DIII. I haven't really seem the results yet. There was some initial joking about just pressing the 'HDR Button' instead of needing to perform manual blends. But I don't know how effective this new feature actually is. Someone did mention it looks more natural than what current post processing HDR tools produce. It will be interesting in the coming years to watch this stuff evolve. I think advances in sensor technology will be the greatest improvement as greater DR will be captured natively.
I have GND's but rarely use them b/c it is easier in PS to manually blend RAW files - lately one RAW file. I can usually emulate the GND in the field.
Moving subjects or foreground is where the GND is still useful and so I carry them with me. They are so scratched, though, I need a new set, and they flare badly so I have to decide if it's worth shooting the sun with them. They give me more work in PS to clean them up. And nothing's more disappointing than coming home and seeing your GND line in the wrong place. That's an instant deletion from the computer just like OOF shots. Not worth trying to fix.
I typically shoot the same scene multiple ways for that anticipated PS workup later. That could mean 20-30 shots for the one scene. But I tend to shoot from a few spots and capture the changing scene until I think the show is over.
Hope this helps. I have HiTech GND's that I now hold in front of the CPL and process with CS2 (not a typo).
Scott Kroeker wrote:
Kevin, that is exactly what I just posted about. By taking 2 or more exposures and blending them together is HDR. You can`t get around that but it is not tone-mapping. These will produce totally different results
Thats a somewhat arbitrary definition. to me using a filter is not any more or less an hdr than simple esposure blending using a gradient mask. Besides, now you are leaving HDR definition in part as the form of precessing and sensor technology. would getting 14.3 stops of dynamic range out of a d800 and pushing the shadows not an HDR, but using two images with a 5d2 an HDR?
In any case. I will be more clear for sake of semantics and call it tone mapping.
I am also very much with ben. as far as photography goes, we are very much used to the available technology. so noone minds the exaggerated percpective one can achieve with a 12mm lens by getting very close to foreground objects, since it has been around for a while , but dynamic range that as photographers we recognize is hard to achiever looks fake.
Grads for me first, if I need more control and am shooting a 1 second or longer exposure and only want to block certain areas, I will use a black glove or a black card in front of the lens. (like burning in a darkroom, only you're doing it live, during the exposure)
This method takes practice but can be highly effective.
I used to hold out for the broader, generic definition of HDR against some who maintained that the term only applied to a process using a 32-bit intermediate file. To my mind the concept of "high dynamic range" photography should apply any time more than one source file, each including a different range of scene luminance, were combined in a way to create a pleasing rendition of the larger tonal range in the final output medium. By that definition, "HDR", exposure fusion, manual blending of multiple images, and blending multiple images with luminosity masks would all fall under that umbrella. I finally gave and decided I didn't want to be associated with "HDR" anyway. Now, when I bother, I try to used the term, "extended dynamic range" (edr?) photography.
However, with ether term the key element is that multiple images are involved in gathering good data on the full range of scene luminance, when it exceeds that of the camera system. It therefore doesn't encompass everything used to deal with excessive scene luminance; just particular techniques involving multiple images. Using grads is a method of coping with scene luminance in excess of the film or sensor abilities; effectively reducing that range before it reaches the recording medium. Similarly, a D800 or future camera with even greater dynamic range may make the concept of "edr" obsolete, not by "being" edr but by making the additional steps unnecessary. By the same token, cameras that do "in camera HDR" are still doing HDR (or some form of edr) so long as the internal process involves multiple images.
So while I don't think that having a precise definition of "HDR" is critical, it's still useful to make a distinction between all methods of dealing with dynamic range, and the subset of them that involve the use of multiple images. And I don't care what you call it!
khilleg wrote:
I have GND's but rarely use them b/c it is easier in PS to manually blend RAW files - lately one RAW file. I can usually emulate the GND in the field.
Moving subjects or foreground is where the GND is still useful and so I carry them with me. They are so scratched, though, I need a new set, and they flare badly so I have to decide if it's worth shooting the sun with them. They give me more work in PS to clean them up. And nothing's more disappointing than coming home and seeing your GND line in the wrong place. That's an instant deletion from the computer just like OOF shots. Not worth trying to fix.
I typically shoot the same scene multiple ways for that anticipated PS workup later. That could mean 20-30 shots for the one scene. But I tend to shoot from a few spots and capture the changing scene until I think the show is over.
Hope this helps. I have HiTech GND's that I now hold in front of the CPL and process with CS2 (not a typo).
I often wonder why, in the end, it would matter whether a photograph is the result of two exposures or one.
It can't be simply that "manipulation" is involved, since essentially no photograph is unmanipulated. It can't be that data from a time frame larger than that of a single exposures is involved, since we often use individual exposures that cover more time. It can't be that a separate processing is done on the dark and light areas since that is what the GND does. It can't be that it looks phony, since well done blending of two exposures tends to show fewer processing artifacts than a single one down with the GND.
I read this thread in a way with a sense of humor thinking about what this discussion will be like in say 5 years. With the newer sensors it will become less and less of an issue. Terms will probably evolve with lines being blurred. So it doesn't really matter all that much IMO. I guess it helps us some now in just knowing what one is talking about. But the bottom line is simply the end result - the image. If the final product is pleasing then the methods to get there really don't matter a whole lot, unless of course a technical question is answered regarding how one 'made the image.'
You are in the minority with your opinion of what constitutes an HDR image. I will second everything that Dan wrote. As photographers, we were blending images with layer masks long before the automated process that was then called HDR came along. You can try and insist all you want but this has been discussed here in many other threads, so you can call an apple an orange all you want, yes they are both fruits, but they are not the same thing. (and yes, there are obvious holes in that logic, but hopefully you or the other few will understand the intent).
For this forum's purposes, the following definitions will apply.
HDR: the automated process where by the user can adjust some sliders to adjust aspects of the automated process with 2 or more images. Typically the dynamic range is highly increased, along with a loss of contrast unless sliders are adjusted to counter this effect.
Manual Blending: The manual process of the photographer taking 2 or more images and manually creating layers. These layers are combined by the photographer using layer masks where the photographer decides where the blends take place in the image. This is not HDR, though the dynamic range can increase, it is done specifically by the photographers choice.
I believe in most cases because in general the HDR look is so horrible, that some people want to combine the automated HDR procress and the manual layer mask blending together into 1 term so that they can justify their use of the automated HDR programs.
So for our purposes here, that will be our forums definition of what HDR is, and what HDR isn't. So let's get back to the OP's question...
JimFox wrote:
HDR: the automated process where by the user can adjust some sliders to adjust aspects of the automated process with 2 or more images. Typically the dynamic range is highly increased, along with a loss of contrast unless sliders are adjusted to counter this effect.
So by that definition exposure fusion is HDR. Ick.
You are in the minority with your opinion of what constitutes an HDR image. I will second everything that Dan wrote. As photographers, we were blending images with layer masks long before the automated process that was then called HDR came along. You can try and insist all you want but this has been discussed here in many other threads, so you can call an apple an orange all you want, yes they are both fruits, but they are not the same thing. (and yes, there are obvious holes in that logic, but hopefully you or the other few will understand the intent).
For this forum's purposes, the following definitions will apply.
HDR: the automated process where by the user can adjust some sliders to adjust aspects of the automated process with 2 or more images. Typically the dynamic range is highly increased, along with a loss of contrast unless sliders are adjusted to counter this effect.
Manual Blending: The manual process of the photographer taking 2 or more images and manually creating layers. These layers are combined by the photographer using layer masks where the photographer decides where the blends take place in the image. This is not HDR, though the dynamic range can increase, it is done specifically by the photographers choice.
I believe in most cases because in general the HDR look is so horrible, that some people want to combine the automated HDR procress and the manual layer mask blending together into 1 term so that they can justify their use of the automated HDR programs.
So for our purposes here, that will be our forums definition of what HDR is, and what HDR isn't. So let's get back to the OP's question...
gdanmitchell wrote:
I often wonder why, in the end, it would matter whether a photograph is the result of two exposures or one.
In the end it doesn't; it doesn't matter if the image was created by magic, if its a nice image. But along the way we like to discuss techniques for achieving the end result (or work our special magic) so it's useful to have common understandings and categories for techniques have particular effects, or affect the image in similar ways. The use of multiple exposures defines a particular family of techniques that are inherently different from ones not using multiple exposures.
It can't be simply that "manipulation" is involved, since essentially no photograph is unmanipulated. It can't be that data from a time frame larger than that of a single exposures is involved, since we often use individual exposures that cover more time. It can't be that a separate processing is done on the dark and light areas since that is what the GND does. It can't be that it looks phony, since well done blending of two exposures tends to show fewer processing artifacts than a single one down with the GND.
No, I don't think reference to HDR is necessarily pejorative (although it has become that way, through habitual association of the term with a technique that is often used for unattractive results), or that it is necessarily superior. But methods using more than one exposure are qualitatively different from other ones, so it's useful to group and understand them on that basis. It is also useful to compare and contrast the different methods used to select the areas (or pixels) from the source images for inclusion in the combined result, and what follow-up processing is needed in each case to make the combination work.
But anyhow, it doesn't matter in the long run how we categorize them, so long as we are clear in our communication about them.
chez wrote:
Come on Jim, tell us how you really feel about HDR.
Ha ha, you know me Harry, I am not shy. BTW, the bold underlined part was not intentional, I probably shouldn't try typing so much using the iPhone. But I stand by what I say, until real recent 99% of all HDR shots posted here could be called out as such because the HDR process leaves telltale signs. Only recently have we seen more HDR photos have been done so that we can't tell, but often even with that the HDR image is then manually blended into another shot where it was used to help recover blown highlights in the sky for example. So to imply that ALL HDR shots look horrible was an over statement ony part, but in general, the vast majority of HDR shots posted here are still looking very weak...
Interesting points. Like you, I'm ultimately interested in the quality of the photographic image and less interested, at least on the aesthetic level, in trying to assign value to the image based on what techniques were or were not used. (I'm not totally immune to that, and in some cases it can be relevant.)
I'm not so sure what you imply when you write that "methods using more than one exposure are qualitatively different from other ones." I'm thinking that you did not mean that as a contradiction to what you wrote in your first paragraph about the value of the image being largely independent of the methods used to create it.
And, by the way, my backwards point in asking the question was essentially to lead to the conclusion that the techniques are simply techniques and that, for the most part, they are not what determine the value of a photograph.
Dan
dsjtecserv wrote:
In the end it doesn't; it doesn't matter if the image was created by magic, if its a nice image. But along the way we like to discuss techniques for achieving the end result (or work our special magic) so it's useful to have common understandings and categories for techniques have particular effects, or affect the image in similar ways. The use of multiple exposures defines a particular family of techniques that are inherently different from ones not using multiple exposures.
No, I don't think reference to HDR is necessarily pejorative (although it has become that way, through habitual association of the term with a technique that is often used for unattractive results), or that it is necessarily superior. But methods using more than one exposure are qualitatively different from other ones, so it's useful to group and understand them on that basis. It is also useful to compare and contrast the different methods used to select the areas (or pixels) from the source images for inclusion in the combined result, and what follow-up processing is needed in each case to make the combination work.
But anyhow, it doesn't matter in the long run how we categorize them, so long as we are clear in our communication about them.
JimFox wrote:
So to imply that ALL HDR shots look horrible was an over statement ony part, but in general, the vast majority of HDR shots posted here are still looking very weak...
Sort of like saying, "All HDR photographs look terrible because all of the photos that look like they used HDR look terrible to me," when other photos do use HDR, but in a way that is subtle and not "terrible." :-)
(I only discovered this when a friend who is a well-known and highly regarded landscape photographer and whose work is largely thought to look very "natural" explained to me how he used HDR - the literal kind - to deal with a particular issue in some of his prints, and did so in a way such that I never would have guessed that HDR was used.)
!!