DPChallenge: A Digital Photography Contest You are not logged in. (log in or register
 

DPChallenge Forums >> Tips, Tricks, and Q&A >> HDR technique versus dodge, burn, curves, levels..
Pages:  
Showing posts 1 - 25 of 27, (reverse)
AuthorThread
02/18/2007 05:58:25 PM · #1
I need a reality check here...

It seems to me that algorithms applied by CS2's Merge to HDR and Photomatix software can be accomplished through a variety of overlays, burning, dodge, curves, levels, image masking, etc.

Is this a correct assumption? It seems as though I tend to get more of what I wanted by manually combining things rather than having software do it.
02/18/2007 06:09:32 PM · #2
I agree, and personally i much prefer the manual method, and think it gives a much, ,uch better result.
I really don't like the HDR 'effect', i think it looks very fake and clearly processed.
02/18/2007 06:15:46 PM · #3
There have been a few HDR photos, actually, I saw a link in another forum post, that had some great images, and looked very real.
02/18/2007 06:32:36 PM · #4
That's a good point actually, maybe i'm being a bit short sighted and just haven't seen any good examples yet.
02/18/2007 06:33:42 PM · #5
....

Message edited by author 2007-02-18 22:51:21.
02/18/2007 07:05:35 PM · #6
How was that processed?
Off topic: I saw only 1 DQ, what happened that you couldn't post?
02/18/2007 07:16:51 PM · #7
I guess I should restate my question:

When would one use Photoshop Merge to HDR, or Photomatix for HDR processing over overlaying multiple exposures and process by curves, levels, masking etc. Is it simply a convenience? What would the advantages of HDR software be?

Paul
02/18/2007 07:25:21 PM · #8
HDR, like any other technique, can be used well or used poorly. HDR and pseudo-HDR became "popular" a short while back, and whenever that happens, you get all sorts of stuff coming out, good and bad.

IMHO, in general good use of HDR would produce an image that is beautiful in an off itself, not because of the HDR. In other words, an image where HDR was put to good use wouldn't scream "Hey, I'm HDR, I look weird!"

As for the advantages, I'm sure others here will be better able to answer that question. There are situations where HDR treatment works wonders, but I haven't been able to pinpoint which situations (it's guesswork for me at this stage to know beforehand if it will work or not). It seems that often you can achieve similar results using other methods, and to me it's a matter of taking each image and trying to see what the best post-processing is to show what I want to show with the image.

The convenience of running [some] images through Photomatix is that I quickly get another point to view so to say.

Here are two examples of HDR processing (not true HDR, but Photomatix pseudo-HDR) that (at least to me) do not "scream" HDR:



and yet Photomatix worked well for processing these two.

You might want to take a look at Neil Shapiro's portfolio, ( nshapiro), he's been doing some great stuff with HDR.

Message edited by author 2007-02-18 19:26:54.
02/18/2007 07:35:13 PM · #9
Here's a shot I took over the weekend that I don't think "screams" HDR either:



But to answer your question, I don't think there's anything magic about using HDR software. But it's an awful lot easier to load the three raw images into photomatix and twiddle a few knobs than it is to do it by hand IMO (it's not something I have any experience doing so maybe I'm wrong...).

splidge

edit: and yes, that image needs some livening up somehow but I'm not sure how to go about doing it :(.

Message edited by author 2007-02-18 19:37:10.
02/18/2007 07:59:38 PM · #10
Ok, Thanks. I guess like anything there are different ways to go about getting to the same thing, and it really depends on the image. All part of a learning curve I guess. But, I suppose the most important thing is not to assume that HDR software is the CORRECT choice.
02/18/2007 08:51:30 PM · #11
The primary purpose of HDR(High Dynamic Range) is to properly capture a scene when the dynamic range (number of f/stops) between the brightest an darkest areas is to great for a camera to capture in just one frame.

In this case you take multiple exposures of the same scene over a range of f/stops so that the captures of the brightest and darkest areas have properly exposed detail. HDR is used to combine all the exposures together to combine detail from the brightest areas with detail from the darkest area to get create one properly exposed capture with detail across the whole dynamic range.

Message edited by author 2007-02-18 20:51:55.
02/18/2007 09:03:21 PM · #12
Right, I understand that. My question is on the methodology to get there. Must one utilize HDR software or will standard, "manual" techniques work and obtain an equivalent result?

Originally posted by stdavidson:

The primary purpose of HDR(High Dynamic Range) is to properly capture a scene when the dynamic range (number of f/stops) between the brightest an darkest areas is to great for a camera to capture in just one frame.

In this case you take multiple exposures of the same scene over a range of f/stops so that the captures of the brightest and darkest areas have properly exposed detail. HDR is used to combine all the exposures together to combine detail from the brightest areas with detail from the darkest area to get create one properly exposed capture with detail across the whole dynamic range.
02/18/2007 09:09:49 PM · #13
I do "psuedo" HDR all the time ... exposing a scene with the intent of blending a double or triple RAW conversion of the same file.

Here's one that had an exposure range going from 1/60sec in the shadows to 1/600sec for the sky. I exposed at a point where as few highlights were blown as possible while still leaving as much detail in the shadows as possible (1/160sec). I then converted the RAW image three times in ACR: once for shadows, once for highlights and once with blend for midtones. I merged the images manually in PS CS2 using layer masks and blending techniques I've learned over the years.



The majority of simiar landscape shots in the accompanying gallery were processed the same way, though some could be done with only "double processing" (2 RAW conversions of the same file). Here's another ...



IMO ... HDR processing through CS2/3 or Photomax (et al) simply allows for convenience and still requires a lot of forsight to do right. Most "true" HDR conversions look like crap, IMO, and many aren't even high-dynamic-range scenes so they just look over processed.

Message edited by author 2007-02-18 21:14:27.
02/18/2007 09:29:49 PM · #14
Thanks Rob.

Your description is exactly what I was referring to. The photo was of a house lit by the sun and a snow covered landscape in shadow.

I took a single RAW and processed it a few times and then brought it into CS2 Merge to HDR. The result was not very good. I then took the image, which I had already exposed for the maximum levels (i.e. histogram was right-centered). I then created a copy of it as a layer and masked appropriately. The result was exactly what I was looking for.

Paul
02/18/2007 09:36:12 PM · #15
Originally posted by PGerst:

Right, I understand that. My question is on the methodology to get there. Must one utilize HDR software or will standard, "manual" techniques work and obtain an equivalent result?


To answer your original question:

If you are working from a single exposure, and either highlights are blown or the shadows are "empty", no amount of processing of that single exposure, by whatever means, can recover the full tonal range of the scene.

If you are working with RAW and a single exposure, then it is possible to make "different" exposures of the same scene in the RAW processor; one for the correct rendition of the highlights, one for the correct rendition of the shadows, and one for correct rendition of the midtones, assuming that the original exposure actually has not emptied the shadows or utterly blown the highlights. This is because the sensor is capable of recording more dynamic range than the screen can "see". If the dynamic range is too extreme, even the sensor cannot cover the whole gamut in a single exposure.

But let's assume you actually do have an image where the highlights are a stop too bright (blown) and the shadows a stop too dark (blocked): now you can actually create these 3 "different" exposures in RAW and, in some way, combine them. You can do this by hand, by layering 2 or 3 images in Photoshop and then creating layer masks on each layer and "painting out" whichever portions of each layer are over or under exposed. Or you can use some sort of "merge to HDR" function (Photomatix is one, Photoshop CS and higher has another) to accomplish the same thing in a more-or-less automated manner. But you DO have control over how these softwares work to produce the finished result, and that is "tone mapping" in Photomatix or "Shadow/Highlight" in Photoshop; so it's not like you're at the mercy of the machine.

Now forgetting the various "merge" possibilities for now, let's assume you want to work on a single exposure with "traditional" PS tools such as curves, levels, dodge/burn and so forth. The first thing you'd do is look at your histogram, and use curves, contrast, or levels to reduce the contrast of the image so neither the shadows nor the highlights are out of gamut. But when you do that you get a very muddy looking image; it lacks local-area contrast. You can try to get around this with all sorts of customized, local enhancements (dodging and burning being two of them) but it's incredibly nit-picky.

Basically the key to a natural-looking image is to retain local-area contrast while flattening the extremes of tonal range. This is what HDR software is designed to do, and it does it very well. The problem is that many (if not most) people who are using this software are exaggerating local area contrast while reducing luminosity range to such an extreme that the shadows are unnaturally glowing and, in many cases, the highlights are unrealistically muted down; this is the sort of work that people are looking at and saying "I don't like HDR imaging, it doesn't look 'right' to me."

HDR software is just another tool; in exactly the same way that you can use hue/saturation either to make your colors look more "natural" or to amp them up so they are totally cartoonish, so you can use HDR software in pursuit of a natural look or to create a more stylized, cartoonish look. It all depends on what you're after. With a properly done, "realistic" HDR merge, there's nothing that pops out of the image and says "Look at me! I'm an HDR image!" So people give the process a bum rap because they judge it solely on the cartoons, which are easily recognizable and not to everyone's (or even most peoples') taste.

As Rob O's images show, this kind of stuff can be done without benefit of specialized programs. The main difference between his results and Photomatix results is that the hand-made versions are much more difficult to work with when it comes to increasing local area contrast. And, of course, working from 3 (or more) actual exposures on the scene is better yet, when you can do it. These can be merged by hand as well. Luminous Landscape has a good tutorial on this. The technique predates HDR software, and indeed predates digital imaging altogether; film photographers have been merging different negatives into single prints for a long time.

So the software, like just about everything else associated with Photoshop, represents a tool to "automate" or "structure" processing techniques that people have been accomplishing by hand for a long time.

Hope this helps.

Robt.

Message edited by author 2007-02-18 22:46:43.
02/18/2007 09:48:39 PM · #16
Thanks. This certainly answers my question. It would appear to me that in my cases, where I have found a manual approach to work better, I may not have had enough dynamic range exposed in the photos I had.
02/18/2007 10:17:16 PM · #17
Great response Robert.

I think the key to success with any of these techniques is shooting with them in mind -- i.e., when I shoot in difficult exposure situations I dial in camera settings based on how I plan to deal with exposure (or whatever0 in the digital darkroom.

Taking a shot in HDR conditions without forsight to multi-processing in RAW then simply hoping you can "save" the image through RAW conversion will leave you wondering why it didn't work; if shadow detail is lost or highlights clipped ... you've got no chance of making the image work.

Same is true for proper use of HDR tools like that in CS2/3 or Photomax: You have to properly bracket and capture multiple shots with the HDR process in mind.

While I try to get the best possible image from the camera through proper exposure and whatnot, I *always* keep the digital darkroom in mind and consider it as much a tool in my bag as any lens or flash or whatever.
02/18/2007 10:22:12 PM · #18
Originally posted by PGerst:

Thanks. This certainly answers my question. It would appear to me that in my cases, where I have found a manual approach to work better, I may not have had enough dynamic range exposed in the photos I had.


Or better yet, the dynamic range of the scenes you shot was not so great as to require use of HDR techniques. After all, scenes have whatever range they have, and your sensor can capture whatever range it can capture. So that's not good or bad, it just is. Speaking of conventional work (not going for exaggerated effects), if you don't need HDR processing then don't use it. It introduces its own set of problems, which outweigh any gains you might make in streamlining the processing. Especially, poorly-controlled HDR processing can get noisy as hell. You can see that in some of my earlier attempts. I wouldn't do them that way again.

R.

Message edited by author 2007-02-18 22:22:42.
02/18/2007 10:25:22 PM · #19
Great. Thanks for the information. I now know that I have been on the right track all along. Now...its time to play and practice!

Its been hard for me to get out recently with taking care of our 9 month old, but this summer we are camping on the cape. I hope to try some stuff out then.
02/18/2007 10:46:20 PM · #20
Originally posted by PGerst:

Great. Thanks for the information. I now know that I have been on the right track all along. Now...its time to play and practice!

Its been hard for me to get out recently with taking care of our 9 month old, but this summer we are camping on the cape. I hope to try some stuff out then.


Well, set up camp near Harwich Port and I'll cook for ya and you can pick my brain on HDRI :-)

R.
02/18/2007 10:59:50 PM · #21
Notice that the "Merge to HDR" option in PS is under the "Automate" tab.

;)
02/18/2007 11:15:52 PM · #22
Sounds like a plan! We'll be up near P-Town and may be heading to MV for one of the days, so we'll be passing near, especially if we stop in Chatham like we usually do.

Originally posted by Bear_Music:



Well, set up camp near Harwich Port and I'll cook for ya and you can pick my brain on HDRI :-)

R.


Message edited by author 2007-02-18 23:17:18.
02/18/2007 11:16:03 PM · #23
Yes....and?

Originally posted by Greetmir:

Notice that the "Merge to HDR" option in PS is under the "Automate" tab.

;)
02/18/2007 11:19:59 PM · #24
Originally posted by PGerst:

Yes....and?

Originally posted by Greetmir:

Notice that the "Merge to HDR" option in PS is under the "Automate" tab.

;)


He's pointing out that it's essentially an action, if you go by location: in other words, it's an automation of a lot of things that you could otherwise do (very laboriously) by hand, which is basically what I was saying.

As far as P-town and MV, can't get much further of a spread than that, on-Cape :-) Be a hell of a long day to drive P-town to Hyannis, hop the ferry, do MV, ferry back, and retrace steps to MV... But I guess you're young :-)

Anyway, I got a spare room if that helps :-)

R.
02/18/2007 11:36:53 PM · #25
And ... ?

... I was just asserting that you were right in your orginal assumption (below) IMO. It is an automated task saving you time but limiting some of your control.

Originally posted by PGerst:

I need a reality check here...

It seems to me that algorithms applied by CS2's Merge to HDR and Photomatix software can be accomplished through a variety of overlays, burning, dodge, curves, levels, image masking, etc.

Is this a correct assumption? It seems as though I tend to get more of what I wanted by manually combining things rather than having software do it.

Pages:  
Current Server Time: 08/30/2025 05:22:59 PM

Please log in or register to post to the forums.


Home - Challenges - Community - League - Photos - Cameras - Lenses - Learn - Help - Terms of Use - Privacy - Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/30/2025 05:22:59 PM EDT.