Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: Robert Boire on July 29, 2010, 01:00:26 pm
-
Hi,
I have just started experimenting with HDR and am somewhat disappointed with the results so far.
I have taken a number of high contrast, landscape scenes in various locations. The composition of these scenes are basically simlar a bright background (top) half (sky, snowy mountains, glaciers ) and a subdued (not dark) foreground (bottom) half (forests, lakes etc). I have bracketed the exposures by 1 stop. In general the middle ("correct") exposure has good contrast, but of course has an underexposed foreground. The + 1 stop has about the right foreground, overexposed background, the -1 stop is generally too dark, though the background is arguably more dramatic. (My meter seems to indicate the range is more likely +/- 2 stops, but this tends to make the defects of the individual exposures even worse).
I am generating the HDR's using the native photoshop tools. The results show approximately correct exposures, though the foreground tends to be washed out. Above all most (not all) of the HDRs are flat with the contrast largely lost. I have tried adjusting exposure and gamma and the various conversion options when converting to 16 bit, but for the most parts the result are disappointing. Except in the most extreme cases, I get better results (but not great results) manipulating the middle exposure.
Any advice out there? Am I doing something wrong?
PS:
I do not understand how the HDR algorithm combines the shots or decides which part of each shot it keeps, but this seems to be part of the problem. Are there HDR applications that allow more flexibility over which part of a shot are included in the composite?I have had some better results with digital GND filtering techniques, but this is extremely labor intensive and requires exacting precision. Are there "automated" tools out there for digital GND?
Thanks
-
Hi,
I have just started experimenting with HDR and am somewhat disappointed with the results so far.
I have taken a number of high contrast, landscape scenes in various locations. The composition of these scenes are basically simlar a bright background (top) half (sky, snowy mountains, glaciers ) and a subdued (not dark) foreground (bottom) half (forests, lakes etc). I have bracketed the exposures by 1 stop. In general the middle ("correct") exposure has good contrast, but of course has an underexposed foreground. The + 1 stop has about the right foreground, overexposed background, the -1 stop is generally too dark, though the background is arguably more dramatic. (My meter seems to indicate the range is more likely +/- 2 stops, but this tends to make the defects of the individual exposures even worse).
I am generating the HDR's using the native photoshop tools. The results show approximately correct exposures, though the foreground tends to be washed out. Above all most (not all) of the HDRs are flat with the contrast largely lost. I have tried adjusting exposure and gamma and the various conversion options when converting to 16 bit, but for the most parts the result are disappointing. Except in the most extreme cases, I get better results (but not great results) manipulating the middle exposure.
Any advice out there? Am I doing something wrong?
PS:
I do not understand how the HDR algorithm combines the shots or decides which part of each shot it keeps, but this seems to be part of the problem. Are there HDR applications that allow more flexibility over which part of a shot are included in the composite?I have had some better results with digital GND filtering techniques, but this is extremely labor intensive and requires exacting precision. Are there "automated" tools out there for digital GND?
Thanks
Sample images would help.. a lot.
I find many people when first starting try to apply HDR techniques to scenes not well suited to HDR, or if they do they get the exposure offsets way off. So it helps to have a peek.
-
It depends on what you're talking about with respect to the term 'HDR'. There's a very good article on this site for example about HDR that has nothing to do with the (often) overused, overprocessed, grungy look that most people associated with HDR: "HDR Plea (http://www.luminous-landscape.com/essays/hdr-plea.shtml)". Vincent Versace has taken to calling the type of image that Alexandre writes about 'eXDR' images for 'extended dynamic range'.
Anyway, Terry White has some good tutorials on using Lightroom & Photoshop to do the 'HDR look' thing - available on the Adobe TV (http://tv.adobe.com/watch/creative-suite-podcast-photographers/lightroom-3-to-photoshop-cs5-hdr-pro-workflow/) site. Photomatix (http://www.hdrsoft.com/) is probably the most used HDR software out there, but others are welcome to disagree!
Mike.
-
You should also stop thinking in terms of what "part" of each photo is used in the result. That's not the way it works. If you shot 5 bracketed shots, the HDR software looks at the same pixel in each of those five shots, takes those five values and applies a formula specific to the HDR software to come with a value for that pixel. That value is expressed using an extended representation (usually one 32-bit floating point number per channel).
--Milt--
-
It depends on what you're talking about with respect to the term 'HDR'. There's a very good article on this site for example about HDR that has nothing to do with the (often) overused, overprocessed, grungy look that most people associated with HDR: "HDR Plea (http://www.luminous-landscape.com/essays/hdr-plea.shtml)". Vincent Versace has taken to calling the type of image that Alexandre writes about 'eXDR' images for 'extended dynamic range'.
Anyway, Terry White has some good tutorials on using Lightroom & Photoshop to do the 'HDR look' thing - available on the Adobe TV (http://tv.adobe.com/watch/creative-suite-podcast-photographers/lightroom-3-to-photoshop-cs5-hdr-pro-workflow/) site. Photomatix (http://www.hdrsoft.com/) is probably the most used HDR software out there, but others are welcome to disagree!
Mike.
Thinking about what Milt said.. Photomatix also allows blending.. and for some shots blending might be preferable.
I've had a bit of fun this weekend with the Sony NEX-5 in the "auto-hdr" mode.. (I'll be shooting it in jpeg until LR3/ACR/C1pro supports it in raw) and its amazing how well it does. At 7fps, especially being able to use the articulated LCD as a waist level finder.. it adds another dimension to a pocket portable. I spend half my time with the NEX-5 deciding if I want a HDR, expanded DR, a HD movie, sweep pano, bracketed exposures, or just a regular old image.. Once I figure out what I want to do with it.. then its easy from there..
-
Welcome to the wonderful world of HDR. It can be a frustrating experience in the early stages.
What you're experiencing is common. Without seeing sample images it's difficult to give any concrete ideas of how to go about addressing it.
What version of Photoshop are you using? The tools up to CS4 weren't great. CS5 is much better with the addition of HDR Pro. There are still ways to get good results in prior versions and you may want to look at what's called 'soft tonemapping'.
Toss some samples up for folks to have a look at. Both input images and untonemapped merged images (drop to 16 bit but do no tonemapping).
-
I probably should have mentioned that the process of my previous post is step 1 of a 2 step process. Once you get the high-dynamic-range representation, you then have to map it down to the limited dynamic range of our conventional image file formats (and our eyes, and our monitors, and our printing mediums). This down-mapping is where there is a lot of subjectivity and there are a lot of choices to be made. In Photomatix there are two modes for doing all this. In mode 1 you explicitly do the 1st step ("Generate HDR Image") and the 2nd step ("Tone Mapping"). In mode 2 ("Exposure Fusion") they hide the fact that it is a two step process, doing it all behind the scenes. Mode 2 is simpler, less powerful, and occasionally better. (I think this is what Steve calls "blending".) Mode 1 is complex, powerful, and, if done properly, usually better. In addition, mode 1 is probably the easiest to misuse image manipulation software on the planet (it is the origin of that garish "HDR-effect" you see so often).
BTW, "dropping to 16 bit" is of course a form of tone mapping -- its just one where you don't get any say in what choices are made.
--Milt--
-
I probably should have mentioned that the process of my previous post is step 1 of a 2 step process. Once you get the high-dynamic-range representation, you then have to map it down to the limited dynamic range of our conventional image file formats (and our eyes, and our monitors, and our printing mediums). This down-mapping is where there is a lot of subjectivity and there are a lot of choices to be made. In Photomatix there are two modes for doing all this. In mode 1 you explicitly do the 1st step ("Generate HDR Image") and the 2nd step ("Tone Mapping"). In mode 2 ("Exposure Fusion") they hide the fact that it is a two step process, doing it all behind the scenes. Mode 2 is simpler, less powerful, and occasionally better. (I think this is what Steve calls "blending".) Mode 1 is complex, powerful, and, if done properly, usually better. In addition, mode 1 is probably the easiest to misuse image manipulation software on the planet (it is the origin of that garish "HDR-effect" you see so often).
PM's Exposure Fusion isn't a 32 bit process. It's an exposure blending algorithm similar to TuFuse or EnFuse.
BTW, "dropping to 16 bit" is of course a form of tone mapping -- its just one where you don't get any say in what choices are made.
--Milt--
Fair enough. Drop to 16 bit but do no 'additional' tonemapping.
-
You should also stop thinking in terms of what "part" of each photo is used in the result. That's not the way it works. If you shot 5 bracketed shots, the HDR software looks at the same pixel in each of those five shots, takes those five values and applies a formula specific to the HDR software to come with a value for that pixel. That value is expressed using an extended representation (usually one 32-bit floating point number per channel).
--Milt--
I guess I understand that to some extent, which is maybe part of my frustration. The question is has does the software come up with the value (some sort of averaging?) for the pixel. It seems to me that there are parts in any of the bracketed shots that clearly should not be considered since they will be vastly over or under exposed.
For those that asked I am using CS4.
This down-mapping is where there is a lot of subjectivity and there are a lot of choices to be made.
I understand that a lot of the art is in the down conversion, I just do not seem to be able to get to where I want to go. However I am admittedly not an expert on curves and level.
Here are some samples. The first is the HDR converted to 8 bits. The result is disappointing, but not horrendous. I have some "worse" examples that I will post ...when I find them.
Thanks for the input
-
Robert,
Since I got this started, I will try to elaborate. However, its clear to me that some of the other posters here probably in fact know more about HDR than I do, so I hope there will be other contributions.
I don't know for a fact just what formula any of the HDR programs uses to compute the HDR value for a pixel. I've not seen any of them published. I do know the formula differs from program to program. For example, PhotoAcute claims that their formula is better than others and that you should use their program for "Step 1". I can imagine some basic things about any such formula however. First, you would have to decide on an exposure weighting to apply to each shot. This can be done by looking at metadata, or it can also be done by looking the variance across the shots between the values for a single pixel and taking some sort of average of that variance. For example, Photomatix used to use metadata, but now they seem to use the latter. Then, when you have an exposure weighting for each shot, a nice simple and obvious formula would be to ignore all blown or null pixel values and just take a weighted average of the remaining values for each pixel. I suspect in real life things aren't quite so simple, and that the improvements to this simple approach are proprietary to each HDR program. One obvious complication is what to do with partially blown or partially null pixels, i.e. blown or null in some channels but not all.
Notice that there is a natural concept of "coverage" here. You can say your bracketed shots have covered a scene if your shortest exposure shot has no blown pixels and if your longest exposure shot has no null pixels. Thus there is at least one valid value for each pixel. An observation made by some HDR practitioners is that even if you have covered a scene, you get a better result by doing so with more shots rather than fewer. For example, even if -2, 0, +2 covers the scene, -2, -1, 0, +1, +2 will give you a better looking result. What is probably going on here is a lowering of noise because of the averaging of more values.
Don't get frustrated over HDR if you are using CS4. Nobody seems to think it is a serious tool.
--Milt--
-
Mode 2 is simpler, less powerful, and occasionally better. (I think this is what Steve calls "blending".)
--Milt--
I'd have sworn Photomatix called it blending.. so I brought up my version and it did.. but when I updated they now call it "fusion" which probably reflects the reality that the process while is essentially blending.. is indeed more complicated than that..
In any case.. I just gave "fusion" a quick go at it (sample below).. not bad, seems improved over blending.. but HDR gives you a great deal more control..
(http://www.bangkokimages.com/Portals/0/_MG_0784_5_6-Edit.jpg)
-
Hi,
I have just started experimenting with HDR and am somewhat disappointed with the results so far.
Any advice out there? Am I doing something wrong?
PS:
Thanks
HDR requires lots of experimentation to get the look you want. The available software tools are all very idiosyncratic, with their own quirks and foibles. Photomatix has lots of sliders and apparent image controls, but IMHO the output always ends up looking very artificial and processed, almost regardless of the settings you choose. I've been especially disappointed with its handling of green foliage in the darker (usually foreground) parts of images, as it seems to make everything look like a plastic potted plant. With a lot of Photoshop massaging I can generally salvage something usable, but the aggravation is often not worth it.
FDR tools is worth a try. It permits you to choose which parts of your source capture are incorporated into the 32 bit image, and it seems to do a better job with natural color. Still looks a bit bizarre, though.
After just two days playing with it, to be honest, I'm much happier with Photoshop CS5's HDR pro feature. The results look a lot more realistic and natural to my eye using the default settings out of the box, and the adjustments are more intuitive. So I'll have to cough up the cash for the upgrade, darn it.
Finally, you may not need to use HDR at all. If the division between light and dark portions of the image are obvious and easily masked, a manual blend of exposures will work just fine. That way you're using all your favorite Photoshop tools for dealing with tonal range, rather than leaving it to some automated software "black box".
-
HDR requires lots of experimentation to get the look you want. The available software tools are all very idiosyncratic, with their own quirks and foibles. Photomatix has lots of sliders and apparent image controls, but IMHO the output always ends up looking very artificial and processed, almost regardless of the settings you choose. I've been especially disappointed with its handling of green foliage in the darker (usually foreground) parts of images, as it seems to make everything look like a plastic potted plant. With a lot of Photoshop massaging I can generally salvage something usable, but the aggravation is often not worth it.
FDR tools is worth a try. It permits you to choose which parts of your source capture are incorporated into the 32 bit image, and it seems to do a better job with natural color. Still looks a bit bizarre, though.
After just two days playing with it, to be honest, I'm much happier with Photoshop CS5's HDR pro feature. The results look a lot more realistic and natural to my eye using the default settings out of the box, and the adjustments are more intuitive. So I'll have to cough up the cash for the upgrade, darn it.
Finally, you may not need to use HDR at all. If the division between light and dark portions of the image are obvious and easily masked, a manual blend of exposures will work just fine. That way you're using all your favorite Photoshop tools for dealing with tonal range, rather than leaving it to some automated software "black box".
Excellent tips from Geoff, as always, and agree with everything he said, especially the last paragraph about doing your own blending.
I've also had good success with Tufuse Pro, which automates the blending without many of the complications true HDR brings. It doesn't seem to be supported anymore, though.
There is a beta program I tested earlier this week called Oloneo (http://www.oloneo.com/). Their disturbingly slick web page has some good sample videos. Just like many other HDR tools the UI can be overwhelming, but it seems to be geared towards realism rather than the cartoony HDR look which thankfully is giving away to the next fad.
-
I have Photomatix. It's pretty good, but I found I got better results when using 3 bracketed shots instead of 5. Using 5 images gave me muddy shadows. Not saying that's the way to do it, but it's what worked for me with Photomatix. One other item I liked about Photomatix is that it takes RAW images and converts them at the same time (as you make HDR).
-
Sadly, you're working with your hands tied behind your back a bit with CS4. Even the tools available in CS4 Extended aren't nearly as good as what Adobe came up with in CS5 HDR Pro.
I'll have a preview of Oloneo up on my blog tomorrow. It's only in beta right now so not generally available anyway.
HDR Expose is another good one to have a look at if you're looking for a more 'natural' result. It's the successor to the very good HDR PhotoStudio. The backend processing has been updated quite a lot. I think it's available for $100 right now. Maybe that sale ended on Jul 31, I'm not sure.
Photomatix can give both natural and the hyper-real results but getting the natural result can take a little more work.
HDR author Jack Howard did a two part video series for the Adorama Learning Center on soft tonemapping in CS4. Since that's what you're using, it might be of some use to you.
Part 1 (http://www.adorama.com/alc/blogarticle/Advanced-HDR-Tone-mapping-in-Adobe-Photoshop-CS4-Part-1). Part 2 (http://www.adorama.com/alc/blogarticle/Advanced-HDR-Tone-Mapping-in-Adobe-Photoshop-CS4-Part-2).
-
Finally, you may not need to use HDR at all. If the division between light and dark portions of the image are obvious and easily masked, a manual blend of exposures will work just fine. That way you're using all your favorite Photoshop tools for dealing with tonal range, rather than leaving it to some automated software "black box".
Well you may be right. I guess I was hoping there was a way of having the same control and results as a manual blend with a lot less effort. Perhaps wishful thinking on my part. I will have to try some of the evaluation HRD applications and see what happens.
Does anybody have comments on the examples I posted - from a HDR point of view.
Thanks all for the info.
-
Sorry, I missed the examples earlier.
What is it you're looking for? The tonemapped result isn't that bad. You might want to do some additional post-tonemap tweaking but what are you finding about it that's so offputting?
In looking at the source images, one thing I'd suggest is to bracket wider. You've still got some dark shadows in your 'overexposed' shot and some too bright areas in your 'underexposed' shot. The biggest key to good quality, natural results from HDR tonemapping is capturing enough brightness range. This is the biggest problem most people run into and the biggest reason why they have problems getting a result they like.
The other thing to keep in mind is that the tonemapped image isn't necessarily the ending point. It may be just the starting point. Think of the tonemapped image as a new image that contains all of the range of brightness that existed in the scene when you shot it. It may be a bit flat and lacking visual interest, but that's OK. You now have the opportunity to create your contrast and create the visual interest as you saw it or as you want it. The job isn't usually done with the tonemapping. In many cases, the job is only beginning when the tonemapping is done. Thinking of that tonemapped image as a new 'RAW' file might get you thinking about further editing and optimisation as you would with an actual RAW file out of the camera.
-
I toyed a bit with your images (hope you do not mind); just for comparison, "enfuse.jpg" is what enfuse yields using the default options, and "final.jpg" is what I get after some tone-mapping with pfstools.
-
I have just started looking at this process. I am not convinced that I will be doing a lot of it though. This link should be very helpful to you.
http://photo.net/learn/digital-photography...-photoshop-cs5/ (http://photo.net/learn/digital-photography-workflow/advanced-photoshop-tutorials/hdr-in-photoshop-cs5/)
In particular this passage sums it all up?
Quote
If you truly want to end up with realistic looking images—and are willing to invest the time this takes—it is best to take full control of the HDR blending process using Photoshop layers, blending modes, the Gradient Tool, and the Paint Brush Tool. This Hand HDR process is explained in Creating HDR Images by Hand and Creating HDR Images by Hand. Usually, Hand HDR means avoiding automated HDR tools such as Merge to HDR Pro—although automated HDR can be used to create a background layer, or to handle difficult areas like a forest with complex lights and darks.
Unquote
In essence HDR is only part of the process and not the start and the end of it as some examples on the web would have you believing.
-
In essence HDR is only part of the process and not the start and the end of it as some examples on the web would have you believing.
Exactly!
Step 1 is the collection of the bracketed exposures.
Step 2 is the merging of those exposures into one HDR image. This includes (amongst others) ghost removal for subject motion.
Step 3 is tonemapping. This i a separate operation, also essental for LDR images IMHO.
Robert, from your 3 small JPEG exposure brackets, I merged and tonemapped the following example:
[attachment=23526:C0004_HDR_3_.jpg]
I used SNS-HDR Pro (http://www.sns-hdr.com/), a relatively new tonemapping application from a very talented Polish programmer. The program IMO still needs work in several areas, but it's tonemapping capabilities are the best I've seen sofar (and I've seen many).
Cheers,
Bart
P.S. at the time of posting the SNS-HDR website seems to be down, he is probably updating with a new version.
-
I have just started looking at this process. I am not convinced that I will be doing a lot of it though. This link should be very helpful to you.
http://photo.net/learn/digital-photography...-photoshop-cs5/ (http://photo.net/learn/digital-photography-workflow/advanced-photoshop-tutorials/hdr-in-photoshop-cs5/)
In particular this passage sums it all up?
Quote
If you truly want to end up with realistic looking images—and are willing to invest the time this takes—it is best to take full control of the HDR blending process using Photoshop layers, blending modes, the Gradient Tool, and the Paint Brush Tool. This Hand HDR process is explained in Creating HDR Images by Hand and Creating HDR Images by Hand. Usually, Hand HDR means avoiding automated HDR tools such as Merge to HDR Pro—although automated HDR can be used to create a background layer, or to handle difficult areas like a forest with complex lights and darks.
Unquote
In essence HDR is only part of the process and not the start and the end of it as some examples on the web would have you believing.
Very true. Using pmatix or other tools as a one-stop-shop is bound to lead to disappointment... it's like using auto levels in photoshop and calling it a day. If I'm merging images together, I'll set up the tone mapping to be very flat initially - the histogram is a simple curve in most cases - then use adjustment layers and your standard pshop tools to push and pull it to how I want. This is no different than non-HDR images, except I'm taking advantage of HDR's ability to see past the camera's limitations.
-
I'm sorry but the idea that you must use manual blending to create a realistic looking expanded drange image is completely, utterly, totally false and misleading.
1. Creating an expanded drange image manually is NOT creating an HDR image. HDR has a specific meaning.
2. Photomatix is a very good piece of software (as are several others on the market) and anyone who says it's not possible to get a natural looking image out of PM doesn't know how to use the software. Are there programs out there that are better at the more natural look? Yes. Is it impossible to do with PM? No.
3. As I noted previously (and as others have stated), the tonemapped LDR image is, in most cases, not the end point but rather a starting point. If, again as I noted earlier, that tonemapped LDR image is considered as a new 'RAW' file with adjustments done from there, very good results can be achieved.
-
A lot of misinformation going around about HDR. But Bob Fisher has got it right.
First thing is to think of yourself as collecting data, which needs to be good data, data that you know how to work with.
Shoot 14-bit RAW, and capture to 16-bit TIFs using the best capture software available using the "neutral" or "linear" gamma. You want the most linear data you can get. Bracket +-2EV steps. Sometimes it is better to delay white balancing until later, but if you are shooting daylight, go ahead and white balance before capture. Do not use the built-in RAW converters in Photomatix.
In making an HDR file, you are creating a robust representation for your scene data that will withstand tremendous amounts of adjustment. All of the "shadow" data here is oversampled and quantized into fractional values using floating point. This is as robust as we can get. In general, Photomatix uses the metadata to determine relative exposure values and perform basically a "normalize and average" function, while discarding blown pixels and shadow data below a certain (option selectable) level.
Try Photomatix for making HDR and tonemapping. As Bob says, do not expect to get the results you are looking for out of Photomatix. It is just the starting point.
In doing tonemapping, play with the "strength" control for more or less natural look -- whatever your artistic vision calls for. Again, do not strive to get a final image here. Your aim is to get as clean a representation as you can get, a 16-bit TIF similar to vanilla capture from a single RAW. Take that image into photoshop and sculpt it to your taste with curves, and local contrast enhancements. This 16-bit TIF will withstand a lot of processing without falling apart. With a little practice, you will find the right compromise and produce the output with Photomatix that is most amenable to your style of post-processing.
-
I used SNS-HDR Pro (http://www.sns-hdr.com/), a relatively new tonemapping application from a very talented Polish programmer. The program IMO still needs work in several areas, but it's tonemapping capabilities are the best I've seen sofar (and I've seen many).
In this post (http://luminous-landscape.com/forum/index.php?s=&showtopic=41884&view=findpost&p=354436) of this thread (http://luminous-landscape.com/forum/index.php?showtopic=41884) I did a quick test of SNS-HDR vs TuFuse tone mapping capabilities:
(http://img263.imageshack.us/img263/7202/mapeos.jpg)
The conclusion was that both performed quite OK, but SNS-HDR produced a too finished image for my taste, while TuFuse produced a more neutral image ready for extra processing. IMO I prefer the TuFuse version.
But something I have noticed about HDR images, is that many programs produce fairly good results when the output is observed alone. But when compared to a good manually tone mapped image (curves, layers and/or mask layers, paintbrush,...), none of the automated results can resist the comparision. Unreal lightings (only noticeable when comparing face to face), halos, etc... are very common on automated software outputs. Images obtained manually take more time but are usually better IMO.
Regards
-
Sadly, you're working with your hands tied behind your back a bit with CS4. Even the tools available in CS4 Extended aren't nearly as good as what Adobe came up with in CS5 HDR Pro.
If I am intending to upgrade to CS5, is it still worth getting Photomatrix?
HDR tends to kill contrast and modeling by making everything the same shade - which is very easy to program... does CS5 or Photomatrix leave you with editable grey scale masks that you can fine-tune and adjust?
-
If I am intending to upgrade to CS5, is it still worth getting Photomatrix?
HDR tends to kill contrast and modeling by making everything the same shade - which is very easy to program... does CS5 or Photomatrix leave you with editable grey scale masks that you can fine-tune and adjust?
Maybe, maybe not as far as the need for Photomatix and HDR Pro. Depends on what you're looking for. Download the trial and play with it. See how you like it.
You're talking about 'flat' contrast. Yes, tonemapped HDR images can have very flat contrast. It depends on how you apply the tonemapping. In some cases flat contrast is the sacrifice for a more natural, realistic looking result. You can get a very flat result or one that has more contrast. Do either HDR Pro (CS5) or Photomatix leave you with enough information to create your masks (I think that's really the question you're asking). Yes, they both can. In many cases, a post-tonemap Curves or Levels adjustment in PS will bring back much of the desired contrast. I know it's going to sound like I'm shilling but have a look at my blog. I've got reviews of several of the HDR apps on the market with sample images - including HDR Pro and Photomatix. The one thing you can do in HDR Pro that you can't in other applications is what can be called 'soft tonemapping'. That is, using the 'standard' tools available in PS to work on a 32 bit image with all of the data and flexibility 32 bit gives you and ignoring the actual tonemapping functions.
-
Making an HDR image look 'natural' is very hard. Many of the images that people display after using HDR processing look strange and unnatural...though sometimes intriguing. Success means that you can't tell HDR was used.
I think the key point that people miss about the HDR process is that the actual HDR image is worthless for any sort of display. It needs to be mapped (as several folks have said), and that mapping is subjective business of setting the black point, white point and contrast, and necessarily clipping some information (throwing it away) for the image to look good. The alternative is a very muddy low-contrast image that would please no one.
The best results I have ever seen to deal with dynamic range that exceeds the camera's capabilities haven't actually gone through HDR. They have involved manually blending light and dark regions of two images. This image is manually blended.
Larger version available here: http://www.trailpixie.net/general/rocky_point_wat.htm (http://www.trailpixie.net/general/rocky_point_wat.htm)
(http://www.trailpixie.net/Rocky-Point-Falls1-cropped-small.jpg)
I do a lot of woodland photography and making an HDR image work with detailed leaves that may move a fraction of an inch can be very frustrating. For this reason, manual blending may be a better solution. Here is a reasonably successful example of a woodland HDR that used Photomatix (it's also a mosaic).
Larger version available here: http://www.trailpixie.net/general/primeval_forest.htm (http://www.trailpixie.net/general/primeval_forest.htm)
(http://www.trailpixie.net/Otter-Crk-50D-6382-18-Pano_hdr_tonemapped-compression-S.jpg)
In the individual exposures, the ground is washed out in sunbeams or there is NO detail in the tree trunks. I could have preserved more of the dark detail in the lower corners, but it didn't improve the overall composition of the image.
-
Maybe, maybe not as far as the need for Photomatix and HDR Pro. Depends on what you're looking for. Download the trial and play with it. See how you like it.
You're talking about 'flat' contrast. Yes, tonemapped HDR images can have very flat contrast. It depends on how you apply the tonemapping. In some cases flat contrast is the sacrifice for a more natural, realistic looking result. You can get a very flat result or one that has more contrast. Do either HDR Pro (CS5) or Photomatix leave you with enough information to create your masks (I think that's really the question you're asking). Yes, they both can. In many cases, a post-tonemap Curves or Levels adjustment in PS will bring back much of the desired contrast. I know it's going to sound like I'm shilling but have a look at my blog. I've got reviews of several of the HDR apps on the market with sample images - including HDR Pro and Photomatix. The one thing you can do in HDR Pro that you can't in other applications is what can be called 'soft tonemapping'. That is, using the 'standard' tools available in PS to work on a 32 bit image with all of the data and flexibility 32 bit gives you and ignoring the actual tonemapping functions.
Thanks, Bob.
Do these programs work with Hasselblad Phocus raw files?
Phocus is magic to the extent that you seldom need HDR (I often bracket in case I need HDR, and find that Phocus manages fine without it).
-
Don't know about Photomatix. It uses dcraw as its RAW conversion engine so it would depend on whether dcraw handles your Hassy RAW files. That said, Photomatix works better with TIFFs as the input files. It's faster and you get a better WB.
As far as Photoshop, if ACR will handle the Hassy RAWs then HDR Pro should.
I've read that the brightness range of those Hassy RAW files is pretty spectacular.
-
As the guy that started all this, I have to say that many of the responses have lead me to a lot of reading material that I have been plowing my way through. Thanks a lot. I suppose the good news is that "HDR Frustration" is common and that getting realisitic results requires a lot of work.
Many of the responses refer to tonemapping. At the risk of appearing ignorant I have to ask the question - here goes- what is tonemapping? Is this the process that takes place when converting down from a 32bit image to a 16 bit image? If so, how is using the local adaptation curve in photoship different from the typical use of curves in photoshop?
Also, at the risk of waxing philisophical I did not quite get this comment:
1. Creating an expanded drange image manually is NOT creating an HDR image. HDR has a specific meaning.
Don't both try to do the same thing- create an image with a wider range than thje original bracketed shots?
-
Yes Robert, both manual blending (even automated blending via tools like Enfuse, SNS-HDR or Photomatix Exposure Fusion) and HDR have the same end goal in mind. The goal with both is to increase the brightness range beyond what the camera could capture in a single shot. JP Caponigro came up with a great term for encompassing all these methods. He refers to it as XDR - eXtended Dynamic Range.
HDR - High Dynamic Range - or more properly HDRI where the I stands for Imagery does have a specific meaning. HDR images are in a 32 bit floating point space. Manual blending and automated blending methods never leave the 8 or 16 bit world. The benefits of the 32 bit space for editing and pushing brightness values around can be invaluable. It allows for much more precision in adjusting values. It also allows for much more precision in contrast and colour adjustment which is where you get the whacked out hyper-real results from.
Tonemapping is the process of mapping the wide dynamic range of a 32 bit image that's, let's say, 15 stops or more into a smaller range that we can see on our monitors and print on paper. It's in the 4 to 9 stop range roughly speaking. Tonemapping happens in the 32 bit space. Simply dropping an image from 32 bits to 16 (or 8) isn't tonemapping. Most HDR applications do some form of automated tonemapping initially to give you some measure of on screen preview as a starting point. Some will allow you to apply tonemapping operators to 8 or 16 bit images, but this isn't tonemapping in the same way as with a 32 bit image. If you do it in Photoshop, for example, it will upconvert to 32 bit (without the benefit of a 32 bit space, think of it as converting an 8 bit image to 16 bit) apply the tonemapping then drop back down to 16 bit. Conversely, I can open an HDR image in PS go to Image>Mode>16 bit, choose Exposure & Gamma from the tonemapping dropdown menu, click OK and my image will be converted to 16 bit with no tonemapping applied and the image will still be beyond what I can use on my monitor or in a print.
-
I wonder why tone mapping should necessarily require the use of any 32-bit floating point format. Isn't tone mapping about mapping tones? you can map tones on 16-bit integer files, and you can fit a huge DR scene onto a 16-bit integer image prior to tone mapping.
So, who/where invented/declared that tone mapping is a 32-bit floating point matter? I don't understand this definition. ???
Regards
-
I wonder why tone mapping should necessarily require the use of any 32-bit floating point format. Isn't tone mapping about mapping tones? you can map tones on 16-bit integer files, and you can fit a huge DR scene onto a 16-bit integer image prior to tone mapping.
So, who/where invented/declared that tone mapping is a 32-bit floating point matter? I don't understand this definition. ???
I think it's just a convention in HDR circles to refer to 32-bit to lower-bit conversion as tonemapping - but you're right, tonemapping doesn't require 32 bits (or 16 or 8).
Not to derail the discussion, but I never really understood the "need" for 32-bit images. We can't reproduce them at the moment, and even if/when they can be reproduced at 32 bits, human eye can't possibly distinguish nearly that many colors.
-
I wonder why tone mapping should necessarily require the use of any 32-bit floating point format. Isn't tone mapping about mapping tones? you can map tones on 16-bit integer files, and you can fit a huge DR scene onto a 16-bit integer image prior to tone mapping.
That is correct. Tonemapping just , well, (re)maps tones from one brightness level to another. It is not specifically reserved for 32-bit/channel originals.
So, who/where invented/declared that tone mapping is a 32-bit floating point matter? I don't understand this definition. ???
Don't know either, it probably seems that way because it is often used in combination with HDRI, which renders as very low contrast images if the whole range is squeezed in without some serious tonemapping. To me, applying a curves correction is also (global) tonemapping, as is a low amount high radius USM, but that is local/adaptive.
Cheers,
Bart
-
I do a lot of woodland photography and making an HDR image work with detailed leaves that may move a fraction of an inch can be very frustrating. For this reason, manual blending may be a better solution.
May I ask how you deal with leaf motion with manual blending? Do you simply choose the image with the best overall exposure of the leaves, and then mask them out completely in the other image layers?
-
May I ask how you deal with leaf motion with manual blending? Do you simply choose the image with the best overall exposure of the leaves, and then mask them out completely in the other image layers?
I've had good success with waving flags with low-light photography by shooting a separate frame with high-enough ISO to freeze the flag. I then paste the flag over the final blended image in post.
Similar technique might be too labor-intensive for landscapes with lots of moving detail, but might be worth testing.
-
Tonemapping is the process of mapping tones from a wider range into a range that's usable. Images captured in a single shot in 8, 10, 12, 14 or 16 bit are already within that usable range. Some of the tones may be blown out, some may be blocked up, but that's what can happen with a single shot and there's not a lot that can be done to bring those back in a single shot. But OK, I'll grant that in an absolute technical sense tonemapping doesn't have to refer to 32 bit images. I think it'd be the rare instance where the term was used referring to LDR imagery.
As far as why the need for 32 bit because it can't be used, the same argument could be (and often has been) made about wider gamut colour profiles like ProPhoto. Even a couple years ago these couldn't be reproduced. Now, newer printers can reproduce pretty much all of the ProPhoto space. Newer monitors can display wider colour gamuts than was possible in the past as well. Why? Because technology changes so why not be ready for it.
-
Now, newer printers can reproduce pretty much all of the ProPhoto space.
No, not even close really. It's true today's inkjets can exceed Adobe RGB in certain areas (but not total volume); but that's a far cry from reproducing all of ProPhoto RGB.
-
As far as why the need for 32 bit because it can't be used, the same argument could be (and often has been) made about wider gamut colour profiles like ProPhoto. Even a couple years ago these couldn't be reproduced. Now, newer printers can reproduce pretty much all of the ProPhoto space. Newer monitors can display wider colour gamuts than was possible in the past as well. Why? Because technology changes so why not be ready for it.
This is a common counter-argument, but I haven't seen any evidence suggesting that the human eye can distinguish anywhere near 32 bit colors, so there's really no point even if they could be reproduced.
And the other common counter-argument is about 32 bits withstanding editing better than 16 bit. This is used also in discussions of 8 bit vs 16 bit, but examples of 8 bits not being enough are rare. Therefore 16 bits should be enough for 99.99999% of editing. Note I'm not talking about stress-testing by testing color gradients, but editing real photos.
Finally, since almost no digital camera has true 16 bit color output (MFDBs included), 32 bit color is necessarily computed. While this may arguably yield accurate results when blending multiple exposures, their real-world application and usefulness is questionable.
32 bit might allow theoretically "better" color gradation, but it is overkill. I haven't seen a decent reason to produce anything in that bit depth given the extreme difficulty working with the files.
-
And the other common counter-argument is about 32 bits withstanding editing better than 16 bit. This is used also in discussions of 8 bit vs 16 bit, but examples of 8 bits not being enough are rare. Therefore 16 bits should be enough for 99.99999% of editing. Note I'm not talking about stress-testing by testing color gradients, but editing real photos.
But that is exactly what happens when tonemapping huge dynamic range images. They are manipulated to the extreme (in small ranges of brightness) to creata a perceptually convincing image. Most of the image tones are only subjected to small contrast enhancements, but it's those selective areas that need all the bits they can have to avoid micro posterization. Typical problem areas are bright sky areas and deep shadow areas.
Cheers,
Bart
-
But that is exactly what happens when tonemapping huge dynamic range images. They are manipulated to the extreme (in small ranges of brightness) to creata a perceptually convincing image. Most of the image tones are only subjected to small contrast enhancements, but it's those selective areas that need all the bits they can have to avoid micro posterization. Typical problem areas are bright sky areas and deep shadow areas.
I'm talking about the final result; to rephrase there's no compelling reason to produce images for reproduction in 32 bit space. 32 bit imaging might be required in some workflows as an interim step to get to a useful 16 bit image. For example, I understand all HDR image packages use 32 bit operations. But one can produce "perceptually convincing" images from outrageously high dynamic range subject matter (http://www.harrijahkola.com/panoramas_Hong_Kong_Cityscape_At_Night.html) by manual blending, while staying securely in 16-bit space throughout the entire workflow.
-
But that is exactly what happens when tonemapping huge dynamic range images. They are manipulated to the extreme (in small ranges of brightness) to creata a perceptually convincing image. Most of the image tones are only subjected to small contrast enhancements, but it's those selective areas that need all the bits they can have to avoid micro posterization. Typical problem areas are bright sky areas and deep shadow areas.
I agree, but 16-bit integer can also encode a huge DR, just need to make use of gamma expansion (on the left linear levels per stop in a linear encoding, right gamma 2.2):
(http://www.guillermoluijk.com/article/superhdr/tabla.gif)
We are talking about 20 stops of DR sufficiently well defined with 16-bit integer, and we could even improve this by using a higher gamma curve. This 16-bit integer TIFF file (http://www.guillermoluijk.com/download/superhdr.tif) encodes a scene of >16 stops of real DR, and shadows can be lifted by 15 stops with no posterization.
So the point of using 32-bit floating is to make easier and/or more efficient from a software point of view to make calculations with linear values. But for the purpose of just encoding a lot of DR, there is no technical need for 32 bit.
Regards
-
So the point of using 32-bit floating is to make easier and/or more efficient from a software point of view to make calculations with linear values. But for the purpose of just encoding a lot of DR, there is no technical need for 32 bit.
Hi Guillermo,
Yes, I agree that for a final result, 16-b/ch is plenty, even if there will be some colormanagement involved for output (e.g. printer profile conversion). I don't see any issues in that stage of the workflow.
But when the images are going to be a basis for further work, they offer sometimes a bare minimum quantization precision. Blending of colors in Photoshop layers is e.g. best done with a linear gamma setting in the P.S. Color settings, and linearizing pre-gamma adjusted integer values and re-introducing gamma for the final output risks introducing posterization or luminosity shifts (http://"http://www.openphotographyforums.com/forums/showthread.php?p=100651#post100651") in smooth gradients or sharp edges. Nothing to get too exited about, as long as one is aware of what the results can be, and how to avoid nasty surprises.
Image math is customarily done in linear gamma colorspace (e.g. Lightroom apparently does it for some operations by default), and Photoshop offers the choice for some of the blending operations. The HDR renderers work in linear gamma and so do good video/graphics cards when mapping textures (http://"http://msdn.microsoft.com/en-us/library/bb173460(VS.85).aspx").
Cheers,
Bart
-
No, not even close really. It's true today's inkjets can exceed Adobe RGB in certain areas (but not total volume); but that's a far cry from reproducing all of ProPhoto RGB.
Not according to Epson with their newer x900 models. I went to an Epson Print Academy seminar shortly after the x900 models came out. Schewe, Rodney, Caponigro et al were the speakers and I can't remember who it was showed gamut overlays of the x900 inkset on Exhibition Fibre and ProPhoto and the ProPhoto gamut was nearly all encompassed.
But putting that aside, the gist of the argument I was making was that technology is improving. So why not be ready for it? Sure, if I've got my original RAW files I can always go back and reprocess images to take advantage of technologicial improvements but why do that extra work when simple steps can be taken today to be ready for it down the road.
Bart, what setup do you have in PS that allows you to work in a linear gamma environment? Your links aren't working either. Both have an extra 'http' at the front. When I remove the extra 'http' from the luminosity shifts link, I'm eventually taken to a wikipedia article on hypertext transfer protocol.
Feppe, how long did you spend at the computer creating that image?
-
But putting that aside, the gist of the argument I was making was that technology is improving. So why not be ready for it? Sure, if I've got my original RAW files I can always go back and reprocess images to take advantage of technologicial improvements but why do that extra work when simple steps can be taken today to be ready for it down the road.
While I agree when it comes to gamut, I disagree when it comes to bit depth. I'm positive I can see more colors than today's printers can produce, perhaps all the way up to Prophoto, therefore I use Prophoto to future proof my workflow. But I'm also positive I can't see the gradation of colors in a 32 bit image - it's probably much closer to 16 bits than 32 bits, perhaps even lower than 16 bits.
Therefore 32 bit imaging is overkill, and since the files are a pain to work with due to the need for tonemapping, I stick with 16 bit imaging as it's just a future proof as Prophoto.
-
And that's fine. I don't know that I can see all the gradation in a 32 bit image either. I don't know how much gradation the human visual system can perceive. It's going to be different for every person. I know trained printers who claim they can see unsmooth gradation in 16 bit prints. I have no way to verify that so have to take them at their word.
I really don't want to get into an 8 vs 16 vs 32 bit urinating contest. It's almost as productive as Canon vs Nikon, ETTR vs non-ETTR, UniWB vs regular WB. :o ::) :P ;D
To suggest HDR files are a pain to work with simply because of the need to tonemap them into a visible range seems a bit of an odd argument. Perhaps you've got your manual blending down to a smooth science and can crank off large, manually blended panos in a few minutes but I don't think most people could so I'm not sure the amount of work or hassle involved in the two methods is necessarily the issue. It comes down, in that case, to a comfort factor. And that's fine too.
-
Not according to Epson with their newer x900 models. I went to an Epson Print Academy seminar shortly after the x900 models came out. Schewe, Rodney, Caponigro et al were the speakers and I can't remember who it was showed gamut overlays of the x900 inkset on Exhibition Fibre and ProPhoto and the ProPhoto gamut was nearly all encompassed.
I suspect your memory is a little fuzzy and they were actually comparing to Adobe RGB (which still has more total gamut volume, but doesn't fully contain the gamut of today's inkjets in certain areas). No current inkjet, including the Epson x900's, comes even close to encompassing ProPhoto; and I don't believe Schewe et al would have made such a claim.
But putting that aside, the gist of the argument I was making was that technology is improving. So why not be ready for it? Sure, if I've got my original RAW files I can always go back and reprocess images to take advantage of technological improvements but why do that extra work when simple steps can be taken today to be ready for it down the road.
I don't really disagree. I use 16-bit ProPhoto RGB and print through Canon's 16-bit plugin. I'm all for making the most of what technology can do. But when incorrect statements go unchallenged they have the potential to become "internet facts".
-
While I agree when it comes to gamut, I disagree when it comes to bit depth. I'm positive I can see more colors than today's printers can produce, perhaps all the way up to Prophoto, therefore I use Prophoto to future proof my workflow. But I'm also positive I can't see the gradation of colors in a 32 bit image - it's probably much closer to 16 bits than 32 bits, perhaps even lower than 16 bits.
Therefore 32 bit imaging is overkill, and since the files are a pain to work with due to the need for tonemapping, I stick with 16 bit imaging as it's just a future proof as Prophoto.
It's not about color gradations. The 32-bit formats are meant to encompass huge dynamic range in linear encoding, as Guillermo said. If I'm not mistaken, the common HDR formats have their roots in CGI and 3d rendering, and the 32-bit floating point format was chosen for good reasons.
Photographers (or I guess Photomatix) kind of co-opted these formats for our uses, they weren't originally designed with high-contrast photography in mind.
One could argue that 32-bit floating point formats are superior. The fact that the files are more trouble to work with is the fault of our software, more than the file formats. As computers get faster, the reasons for staying with integer math diminish (the biggest reason it's here for the foreseeable future is legacy code more than anything else).
-
To suggest HDR files are a pain to work with simply because of the need to tonemap them into a visible range seems a bit of an odd argument. Perhaps you've got your manual blending down to a smooth science and can crank off large, manually blended panos in a few minutes but I don't think most people could so I'm not sure the amount of work or hassle involved in the two methods is necessarily the issue. It comes down, in that case, to a comfort factor. And that's fine too.
I think it's more about what we're used to, and should have qualified my statement that blending works better for me. I've tried "proper" HDR techniques and programs, and none of them produce results I like - I feel constrained by the global adjustments, and just can't get my head around all of the rather abstract sliders. Granted, I haven't done much experimenting, as using layers and blending is much easier for me. For example, I use PTGUI's exposure fusion as an extra layer, as it does produce good results in some part of the image, while flattening others too much.
Agree fully on the pointlessness of the almost-religious discussion of 32 bit vs 16 bit :)
-
It may be, Jeff. Unlike many, I don't profess to be perfect.
-
May I ask how you deal with leaf motion with manual blending? Do you simply choose the image with the best overall exposure of the leaves, and then mask them out completely in the other image layers?
In this crop you can clearly see that branches (not just leaves) moved between shots... the darker branches are from the darker "sky" shot.
This is about a 1 Mpx crop from the background of a 50Mpx picture.
I think I created a "select colour" mask from the sky in the lighter (not sky) picture... and merge ignored the white (sky) where the branches were in a different position... What I was doing was using the sky from one picture (of a minibus, previously posted, I think) and merged it into another. You can see some purple fringing, but it was fit for purpose.
When I got the latest version of phocus, I could recover the highlight so well I did not need to HDR.
-
In this crop you can clearly see that branches (not just leaves) moved between shots... the darker branches are from the darker "sky" shot.
This is about a 1 Mpx crop from the background of a 50Mpx picture.
I think I created a "select colour" mask from the sky in the lighter (not sky) picture... and merge ignored the white (sky) where the branches were in a different position... What I was doing was using the sky from one picture (of a minibus, previously posted, I think) and merged it into another. You can see some purple fringing, but it was fit for purpose.
Thanks, Dick. The mask based on select color is probably a good choice when the motion is against the sky.
-
I am glad to see all this interest in HDR. I've been working with it for a couple of years now and continue to learn new stuff. Getting back to the original post: It is a process and it starts from the ground up, good tripod, lins and camera. Expand your brackets to 2 stops instead of one. Shoot in RAW and process them in the conversion software of your choice (Photomatix should not be used as a raw converter) you should maintain your exposure differences while correcting for white balance (I select all three images but only correct the average exposure image, applying the same adjustments to all the images). Convert your files to .tiff or .psd. Now you can combine these images with the software of your choice (I use Photomatix). After you've finished creating your new HDR file put some polish on it using Photoshop or Lightroom (I use both). It's a time consuming process and the results can be surprising, some good, some bad. It takes time to understand all the interrelated adjustments, just keep trying, you'll get it!
Guillermo: I have always been interested in your Zero Noise, but could never understand the process. Any changes to your User Interface?
Gramps
This is my present stage of development:
-
Anybody see the PR for Nik's new "HDR Efex Pro" software? It wasn't clear to me from the press release whether it's just a psuedo-HDR contrast/detail enhancer in the same vein as Topaz Detail, or if it can actually tonemap 32-bit HDR files.
-
As far as I can tell, Jeff it's a plugin that contains a full application that opens inside Photoshop. It'll work on single or multiple image merges. What I can't figure out, like you, from anything I've seen or read is whether it will work on exisiting 32 bit images. I've not seen anything on whether it's a true HDR app, creating 32 bit image files or whether it's an exposure blending app that never leaves the 8 or 16 bit space. Either way, at $160 it's more expensive than other options on the market, much more expensive than some.
-
This is my present stage of development:
Is this how you wanted it to turn out?
I would want more contrast - I thought Photomatrix gave you control.
-
I thought Photomatrix gave you control.
It does, yes. Depends on the result that's desired. But PM isn't the last step in the process generally, either. Many people tonemap their 32 bit images then bring the resulting LDR images back into their image editor of choice for additional work. They get a good working version through the HDR tonemapping process (i.e., treat it like an enhanced RAW file), then do final work in the low bit depth world as usual.
-
Thank You Mark, that's exactly what I do. Photomatix is just one stop along the way. As I noted in my original reply. I may be pretty whacked out, but I adjust in PM by using the histogram. Rule #1: No clipping. The PM Processed image then moves to PS for Noise reduction and a simple curves adjustment (and object removal, if necessary) then on to Lightroom for minor tweaks and possible cropping.
Regarding the observed lack of contrast; my objective is to produce what I call soft detail, the art of defining the contrast via color tone, not so much luminosity (light to dark). Think: straight curves.
Here's another example from the same shoot. The location is Savannah, Georgia.
Gramps
-
As a follow up to my earlier comments regarding Nik's new offering, I sent an email to their support group this morning. They've confirmed that it's a full HDR application hosted inside PS. It will create 32 bit files on merge and it can tonemap existing 32 bit files.
-
As a follow up to my earlier comments regarding Nik's new offering, I sent an email to their support group this morning. They've confirmed that it's a full HDR application hosted inside PS. It will create 32 bit files on merge and it can tonemap existing 32 bit files.
Good to know. I agree it's on the expensive side (Nik plugins always are). The thing that has me intrigued about it is using the U-Point technology for local adjustments, as well as their claims about halo suppression.