Req: Better HDR exposure merge

[bkh]

Maybe I didn't explain the difference well: When you create a HDR, you re-combine the images according to their different exposures – so if you use a part of an input image which was overexposed by 2 steps, then you underexpose that part by 2 steps before it's added to the final image. In that way, you create an HDR image which has the same tones as the scene.

I don't think so, at least in PL's case, because the resulting image doesn't retain the same tones -- that's why one has to boost both the saturation and the contrast of the final image.

You are probably right as far as PL is concerned – apparently it doesn't let you separate the creation of a HDR image and tonemapping, even if contrast=0.

Luminance HDR, however, does allow one to alter the EV value of the incoming images to affect its final result, but I'm not sure it's algorithm compensates for EV as you've described rather than just combining the images using the specified overall EV settings.

I'm quite sure that Luminance HDR produces a true intermediate HDR. Imo, It's got the most scientifically rigorous approach (but this doesn't necessarily guarantee the best results).

HDRMerge uses a somewhat different approach and aggressively chooses the highest non-clipped values for the final image on a pixel-by-pixel basis. It can do that because it's working directly with the RAW data and creating a RAW (DNG) final result -- no de-mosaicking is involved.

I never tried HDRMerge, but if it follows the strategy you describe, it must take the EVs of the input images into account, otherwise results would be very strange (if you photograph a black-to-white gradient with different degrees of overexposure, then you'd get tone jumps or even coloured stripes where the individual images start clipping). enfuse resolves this by often averaging between neighbouring images, but one should be able to exhibit a similar behaviour when using the "hard mask" option or very low values for sigma.

Cheers

Burkhard.

[photoken]

I never tried HDRMerge, but if it follows the strategy you describe, it must take the EVs of the input images into account, otherwise results would be very strange.

Here's its description of its approach:
https://github.com/jcelaya/hdrmerge#so- ... e-hdrmerge

[greenmorpher]

At any rate, I don't see much (if any) difference between the results that can be obtained with PL and with Enfuse.

I have had this kind of discussion before with you, perhaps, and with others on this forum about various features of PL.

Theoretically, I can get similar results with PhotoLine HDR as I can get with Bracketeer using Enfuse.

In fact, I have never had a good result with PhotoLine HDR until now. That was because I didn't know you should begin by going to gamma which defaults at 1 and set it to 0. Then you talk about a whole lot of other stuff with adjustment layers and different modes.

Forget it!!!

With Bracketeer, I dumped my seven images in, clicked, and had an excellent image pop up in PL for a little lightening and darkening if I felt like it. In fact, getting the lightness right was actually as simple as deleting the first pic in the series, then clicking again, and getting pretty much exactly the image I wanted. Just a matter of saving it.

As I said, theoretically I can get that in PL with a lot of messing about. In Bracketeer I can get what I want with one or two clicks.

That's what it is all about.

Herbert in the enfuse thread reported similar results using the three platform donation ware: http://software.bergmark.com/enfuseGUI/Main.html

Here is a point -- has the HDR in PL been improved since it was introduced? If not, it is about four years old, more than a lifetime in app terms and in urgent need of rejuvenation.

Cheers, geoff

[photoken]

I hesitate to make this request because hand painting the masks of the incoming images is extraordinarily difficult to do without creating beaucoup problems, but....

In addition to providing the alignment points, the HDR creation window could also include editing of the masks of the incoming images. Selecting an incoming image would show its automatically generated mask, and the masking tools provided would be:

• Include in mask.
• Exclude from mask.
• Brush editor (same as the Active Brush options available for the Painting tool -- the brush softness is very important).

There are two somewhat rare cases where this would be useful:

1. Overriding the use of a specific area of a specific incoming image.
2. Removing moving objects such as birds flying across the field of view, or pedestrians, or cars.

[Martin Huber]

I don't think so, at least in PL's case, because the resulting image doesn't retain the same tones -- that's why one has to boost both the saturation and the contrast of the final image.

You can see what PhotoLine is doing by pressing Shift while leaving the dialog with OK.
Basically it is mixing two or 3 images by creating blending masks and adjusting the brightness of the images, so that the brightness of the images matches.

Martin

[photoken]

You can see what PhotoLine is doing by pressing Shift while leaving the dialog with OK.

Holy smokes! That's one heluva option!

It's going to take me a while to digest that....

[Herbert123]

I don't think so, at least in PL's case, because the resulting image doesn't retain the same tones -- that's why one has to boost both the saturation and the contrast of the final image.

You can see what PhotoLine is doing by pressing Shift while leaving the dialog with OK.
Basically it is mixing two or 3 images by creating blending masks and adjusting the brightness of the images, so that the brightness of the images matches.

Martin

Whaaat?

That's pretty cool - are there any other operations in Photoline with those kind of hidden options?

[photoken]

The PL documentation states that two or three images are used for the HDR exposure merge. Is that the maximum number of images that are used?

I've been testing with 5 images. Are all 5 used?

[photoken]

In the HDR exposure merge window, it would help if there were an icon to set the view percentage to "View all" -- finding the alignment markers requires zooming out to the full image view.

[Martin Huber]

The PL documentation states that two or three images are used for the HDR exposure merge. Is that the maximum number of images that are used?

I've been testing with 5 images. Are all 5 used?

No, just three.

In the HDR exposure merge window, it would help if there were an icon to set the view percentage to "View all" -- finding the alignment markers requires zooming out to the full image view.

I have a "Full View" icon at the bottom right of the dialog. Is that missing in your dialog?

Martin

[photoken]

I have a "Full View" icon at the bottom right of the dialog. Is that missing in your dialog?

Ah, yes, the icon is there. I was confused because it's labeled "100%" and I didn't want to view the image at 100% (actual size):

HDR icon.png

[bkh]

Ah, yes, the icon is there. I was confused because it's labeled "100%" and I didn't want to view the image at 100% (actual size):

Actually, "100%" is another icon left of the "full view" icon, not a label. A while ago, I also got confused by this – maybe one should add a frame to the "100%" icon as well …

Cheers

Burkhard.

[photoken]

Actually, "100%" is another icon left of the "full view" icon, not a label. A while ago, I also got confused by this – maybe one should add a frame to the "100%" icon as well …

Now I see -- right you are. One of these years I'll start paying attention to the tooltips....