Farm Sky - Comparing HDR Software

4 HDR software compared.

We had some nice skies recently so was curious how HDR app would treat them. I am not a fan of black clouds as one often sees in HDR skies - these were all processed as "photographic" or normal - no dramatic effects.
I know the range and subject matter is not the best for HDR processing but as I said.....curious.
They were all processed from 3 RAW files 2 stops + and - with no tweaking in the tone mapping other than to ensure a more natural look. The same amount of local contrast and sharpening was added in PP (using Photoshop) to make it a fair compare.
Thought I would share results.

1. Photoshop CS5

I proposed to check the SNS-HDR Pro.

Pebal wrote in post #10508193
I proposed to check the SNS-HDR Pro.

Thanks Pebal....see below - HDR and PP processed the same as above.

5. SNS-HDR Pro DEMO

Interesting and revealing comparison. Thanks for taking the time to prepare and present it. Would also be interested in seeing HDR Photo Studio.

CannedHeat wrote in post #10510189
Interesting and revealing comparison. Thanks for taking the time to prepare and present it. Would also be interested in seeing HDR Photo Studio.

Thanks CH - am beginning to think I should have used a different scene with a broader range of luminance. I started out wanting to focus on skies and clouds but now see there seems to be a big difference in the field/greens.
Anyway, here you go:

6. HDR Photo Studio

Thanks for the comparisons. At the current state of pp for all the images, clouds and land, I think I still like the overall look of photomatix. I realize they could all use additional improvements in pp, but the photomatix render seems to need the least...in my opinion and based on the style I like anyway.

And as anything else in the world of art, it's all personal taste! LOL.

I would put the photomatix not at the bottom, but near it. It made the sky overly dramatic and also did what photomatix does on almost everything: made the greens and yellows glow although in this scene less so.

Spike, thanks for adding HDR PhotoStuido. They market themselves as the answer to realistic HDR, but in this image I think they did a poor job - one of the worst. The clouds are dull, lifeless, lacking contrast and the ground was poorly brought up.

There seems to be a tradeoff in every one of them, but overall (and if what you are after is a realistic scene) I think the SNS HDR did the best job.

Also - consider that most users could get completely different results using the same image and the same apps. This kind of comparison is sort of arbitrary without some common intent or relationship among the settings of thee various applications to somehow normalize the results. Also, if a user does not know how to use a particular application, the results may not be so great. For example, the HDR PhotoStudio image is lifeless and blue, but the application has several ways to remedy this, including a 32 bit WB tool - and, CannedHeat, I know you are no fan of HDR PhotoStudio for the installation problems you had, but in your statement

it is somewhat important to note that "they" (the creators of HDR PhotoStudio) didn't do anything with this image. This tonemap is a function of the OP's use of the app. Nothing against the poster, it is just something that is pretty obvious - like most of the other apps out there, HDR PhotoStudio has controls for local and global contrast, exposure, in addition to spot color correction and 32 White Balance and sharpening. The CS5 image is a mess as well, with a color cast and muddy contrast - this, again, is something that can be addressed with the multitude of tonemap settings in the new HDRPro part of CS5.

I would guess that the OP is most familiar with using Photomatix.

If all of the apps had an "Auto" function and that was all that was done to generate the images, then I could see that at least be a somewhat reasonable comparison to start with.

Kirk

kirkt wrote in post #10511089
Also - consider that most users could get completely different results using the same image and the same apps. This kind of comparison is sort of arbitrary without some common intent or relationship among the settings of thee various applications to somehow normalize the results. Also, if a user does not know how to use a particular application, the results may not be so great. For example, the HDR PhotoStudio image is lifeless and blue, but the application has several ways to remedy this, including a 32 bit WB tool - and, CannedHeat, I know you are no fan of HDR PhotoStudio for the installation problems you had, but in your statement

it is somewhat important to note that "they" (the creators of HDR PhotoStudio) didn't do anything with this image. This tonemap is a function of the OP's use of the app. Nothing against the poster, it is just something that is pretty obvious - like most of the other apps out there, HDR PhotoStudio has controls for local and global contrast, exposure, in addition to spot color correction and 32 White Balance and sharpening. The CS5 image is a mess as well, with a color cast and muddy contrast - this, again, is something that can be addressed with the multitude of tonemap settings in the new HDRPro part of CS5.

I would guess that the OP is most familiar with using Photomatix.

If all of the apps had an "Auto" function and that was all that was done to generate the images, then I could see that at least be a somewhat reasonable comparison to start with.

Kirk

I agree with you to some extent.

But first, to clear up a point. My installation problems with HDR photostudio has no bearing on my evaluation of the functionality of the product. I used to have horrendous problems with NIK plugins (memory / crashing PS) but if you check my post history you will find I have consistently praised them and in fact, I would call NIK SilverEfex the single greatest plugin ever created.

I have thoroughly tested HDR PhotoStuido. First, for the initial trial period, then when they had the panorama contest and offered an extended trial period, with that.

At best, I would evaluate it as average. And I fully know the controls and tested it with a very wide range of input subjects. On many images it flat out failed miserably. AFAIK, of the bunch above (dedicated hdr apps), it is the most expensive so considering that with the average performance, I would judge it to be a poor buy.

You are correct about testing conditions. But then again, anecdotal observations can be noteworthy too - if the tester is equally unfamiliar with many different apps then their treatment by him are relevant.

On the other hand, I politely disagree with your comment about testing an Auto setting. An auto setting is only as good as the programmer who creates it. A HDR application that might be evaluated as the best when used in manual mode might very well be judged the worst if they had an Auto algorithm that sucked.

The only fair way to evaluate these apps would be utilizing standardized testing with control files. Unfortunately, that is beyond the scope of us average folks in this forum. The only thing we have left to depend upon is each other and our personal experience. It, too, will vary but averaged together may indicate a trend.

navydoc wrote in post #10510882
Thanks for the comparisons. At the current state of pp for all the images, clouds and land, I think I still like the overall look of photomatix. I realize they could all use additional improvements in pp, but the photomatix render seems to need the least...in my opinion and based on the style I like anyway.

I think I'm with you. the photomatix sky was the most appealing and if it didn't look like THE sky it looked like I wished it would.

This is great, and I really appreciate your sharing this effort. I am only familiar with Photomatix so it's a huge benefit to see other software going head to head. However, I have to agree with kirkt and question the results of this effort. In Photomatix I have found that a wide variaty of results are possible depending on the photo and the look you're going for. Since I"m not familiar with the others I can only assume this is true of them as well. So with these varying possibilities how could you provide a fair comparison?

Not saying you have an agenda, but if you wanted to skew the results to show any software better than an other it would be very easy to do so. In fact, since I'm sure each program works differently and your knowledge of each must vary to some degree it would seem that a fair comparison could not be had. No offense meant to you in any way. Just seems that this one scenareo isn't a very fair test. Still, I'm going to go back a study the results because I've been looking for another solution to processing HDR's. Thanks again, Jon

J-Blake wrote in post #10512481
This is great, and I really appreciate your sharing this effort. I am only familiar with Photomatix so it's a huge benefit to see other software going head to head. However, I have to agree with kirkt and question the results of this effort. In Photomatix I have found that a wide variaty of results are possible depending on the photo and the look you're going for. Since I"m not familiar with the others I can only assume this is true of them as well. So with these varying possibilities how could you provide a fair comparison?

Not saying you have an agenda, but if you wanted to skew the results to show any software better than an other it would be very easy to do so. In fact, since I'm sure each program works differently and your knowledge of each must vary to some degree it would seem that a fair comparison could not be had. No offense meant to you in any way. Just seems that this one scenareo isn't a very fair test. Still, I'm going to go back a study the results because I've been looking for another solution to processing HDR's. Thanks again, Jon

Question: what was it that you observed in his test images that would lead you to believe he did not produce the best image possible for each respective application? Do you have access to the files? Did you process them yourself and achieve different results?

It would seem to me that somebody that would automatically assume he did not, without giving him the benefit of the doubt, might have more of an "agenda" than the OP himself.

I have no agenda, nor did I observe anything in his images which would indicate that he did. I thought I was quite clear on this. I have no access to his files, wasn't in the room watching, and didn't process them myself. I made no assumptions on his agenda except to give him the benefit of the doubt. I'm merely trying to point out that due to the variety of software I'm familiar with and assuming the others are similar it would be very difficult to provide a fair test. Just because I raise question to methodology of this effort doesn't mean there is any ill will or accusation towards the OP.

The real problem is defining what is the best image! There's no such thing. What I like, you might not. The question is what will do the job that you want done the best. Which software will do what you want it to do. It may not be the same thing I want it to do, so everyone's opinion of which one is better may be different.

J-Blake wrote in post #10512590
I have no agenda, nor did I observe anything in his images which would indicate that he did. I thought I was quite clear on this.

"I have to agree with kirkt and question the results of this effort."

"So with these varying possibilities how could you provide a fair comparison?"

"Not saying you have an agenda, but if you wanted to skew the results to show any software better than an other it would be very easy to do so. In fact, since I'm sure each program works differently and your knowledge of each must vary to some degree it would seem that a fair comparison could not be had. "

Fair enough. The point I was trying to make is that the exact same doubt I cast upon your post (an "agenda") was the same doubt you cast upon the fairness of his tests (an "agenda") in your quotes above.

The bottom line is that I have no reason at all to believe that you have an agenda (an affiliation with, or a preference of) any specific application, and neither do you have any reason at all to believe the op did anything other than the best try he could to be fair with his tests.

Both scenarios deserve the benefit of the doubt.