Local Tone Mapping tool

[CMurd]

In the Wish List I talked about how the Local Tone Mapping tool doesn't work properly. Specifically, the settings 0.1 through 0.9 have no effect. You have to set the tool to 1.0 in order to see any difference. I would like to be able to use smaller values, so the tool needs to be fixed so that the settings 0.1 through 0.9 actually do something.

However, the reason for this new post is that I have found a new flaw in the tool.

As I said in another thread, I sell beads for a living. Since presentation isn't important, I just put the beads on a white plate and photograph them that way. So tonight I filled a white plate with bright orange beads and I photographed them. When I used Local Tone Mapping on the picture, it improved the appearance of the beads in the middle, but the beads on the outer edge became de-saturated and look brownish. Curiously, this only happens when the photo is at full size (20 million pixels). When I apply Local Tone Mapping after reducing the picture, the effect doesn't happen. Nonetheless, this is clearly a bug, and it needs to be fixed.

Below is the picture (substantially reduced) before and after applying Local Tone Mapping.



Corel should check to make sure that there are no other tools which don't work properly on very large files.

By the way, why is the tool called Local Tone Mapping? What's the logic of that name? It used to be called "Clarify".

(Note: I've reported this bug to Corel, but I think it should stay in this thread as a reminder to them that this tool needs fixing. Once my 30 days of free support are up, I won't be able to contact them at all, and bugs will go unreported! Amazing.)

[photodrawken]

the beads on the outer edge became de-saturated and look brownish.

I don't know which of your photos is the "before" and which is the "after", but in both photos the beads at the outer edge are noticeably less saturated than the beads in the middle, even when viewing those reduced size photos....

[CMurd]

The two pictures don't look the same to MY eyes! If you can't see the difference between the pictures, then you need to get a new monitor PRONTO.

The outer edge of beads in the top picture (before LTM) may look slightly lighter than the beads further in, but that's only because -- being at the edge of the grouping -- more light is passing through them (the beads are semi-translucent). But in the bottom picture, the outer 2 or 3 beads all around the grouping is desaturated and brownish looking.

[photodrawken]

The two pictures don't look the same to MY eyes! If you can't see the difference between the pictures, then you need to get a new monitor PRONTO.

Before I spend the money on a new monitor, I'll wait for you to carefully read what I posted.

I stand by what I wrote: in both images, the outer beads have less saturated colours than the inner ones. So, no matter which one of the images was your "original", that lack of saturation will explain the results you got after applying a tone mapping adjustment.

[CMurd]

I stand by what I wrote: in both images, the outer beads have less saturated colours than the inner ones. So, no matter which one of the images was your "original", that lack of saturation will explain the results you got after applying a tone mapping adjustment.

But I'm not seeing what you're seeing. In the top photo, all the beads look about the same to me. In the bottom photo, the beads around the outside look brownish to me -- and not just the outermost bead, but the outermost two or three beads.

I can clearly see what I'm talking about. If you can't, then there's a problem at your end.

[photodrawken]

In the top photo, all the beads look about the same to me.

They're not the same at all -- neither in terms of saturation, nor in terms of lightness, nor in terms of contrast.

You've seen that to be true when you used the tone mapping adjustment. If you use other adjustments, those will also confirm this.

For example, here's a quick test I did using PhotoLine:

The outer beads are very difficult to get to match the saturation, lightness and contrast of the inner beads. Here are the layers I wound up needing:

The settings for the Hue Editor and Curves had to be extreme, to say the least:

Even so, there is still a noticeable difference between the outer and inner beads....

[CMurd]

I already addressed this issue. In the top picture, the very outermost bead looks lighter (which means less colorful) than the other beads because the beads are semi-translucent and more light is passing through the outer bead. That's the way the beads look in actuality.

What I'm seeing in the 2nd picture is a brownish tint in the outer two or three beads around the perimeter; it's not limited to just one bead at the edge. If you can't see it, then there's something wrong at your end. Now, this discussion is over.

[photodrawken]

I already addressed this issue. In the top picture, the very outermost bead looks lighter (which means less colorful) than the other beads because the beads are semi-translucent and more light is passing through the outer bead. That's the way the beads look in actuality.

What I'm seeing in the 2nd picture is a brownish tint in the outer two or three beads around the perimeter; it's not limited to just one bead at the edge. If you can't see it, then there's something wrong at your end. Now, this discussion is over.

OK, so you admit (again) that the outer beads, being translucent, have different saturation, lightness, and contrast because of the white light passing through them. That difference is most obvious in the beads at the perimeter because they have the most white light passing through them, and that difference continues to an ever-lessening extent in the beads inward from the perimeter because of the ever-lessening white light passing through them.

Although it's easily seen, if you can't see that in your original image, you should have realized what's going on by the result of your adjustment. So, two different editing programs using a total of three different adjustments all show that the original image is defective.

The bottom line is this: God knows that PSP has beaucoup problems and limitations, but this specific issue with this specific image is not one of them.

[CMurd]

The original image isn't defective. The issue is what's happening in the second photo. The Local Tone Mapping tool is putting an unnatural discoloration on the beads around the perimeter, and the discoloration is two or three beads deep. I'm sorry you can't see it; but since you can't, there's no point in discussing it. You either need a new monitor, or you need to see an eye doctor.

[LeviFiction]

CMurd, I don't think photodrawken can't see the discoloration. He seems to be suggesting that the discoloration is a natural result of the local tone mapping and not a bug in the local tone mapping. Saying that there is a discoloration already along those beads, so it makes sense that there would be a discoloration on them after the local tone mapping. So, it's not a lack of ability to see the issue, it's just a belief that the issue is a non-issue for PSP and only an issue for your image.

In fact I saw something similar when using Photoshop's HDR Toning (set to Local). PSP is attempting to to tone the bright pixels down improving their contrast and brightening the darker pixels. The end result is what you see. I too do not think this is a bug. But an unfortunate result of exactly what the function is attempting to do. I could of course be wrong, just my opinion.

[bobthegoat2001]

The original image isn't defective. The issue is what's happening in the second photo. The Local Tone Mapping tool is putting an unnatural discoloration on the beads around the perimeter, and the discoloration is two or three beads deep. I'm sorry you can't see it; but since you can't, there's no point in discussing it. You either need a new monitor, or you need to see an eye doctor.

I'm all about finding bugs in PSP and getting them fixed. That being said, I agree with both LeviFiction and photodrawken. On my monitor, I can actually see desaturation in *both* photos, just by looking at them (though it's more pronounced in the second photo with Local Tone Mapping). I don't believe it's a bug in PSP. None of the photos are defective.

After saying all this, I'm hoping we can finally get back on topic about new features we'd like to see in PSP x8.

[CMurd]

CMurd, I don't think photodrawken can't see the discoloration. He seems to be suggesting that the discoloration is a natural result of the local tone mapping and not a bug in the local tone mapping. Saying that there is a discoloration already along those beads, so it makes sense that there would be a discoloration on them after the local tone mapping. So, it's not a lack of ability to see the issue, it's just a belief that the issue is a non-issue for PSP and only an issue for your image.

In fact I saw something similar when using Photoshop's HDR Toning (set to Local). PSP is attempting to to tone the bright pixels down improving their contrast and brightening the darker pixels. The end result is what you see. I too do not think this is a bug. But an unfortunate result of exactly what the function is attempting to do. I could of course be wrong, just my opinion.

No, what I'm seeing looks completely unnatural to me. Here is another picture in which I applied LTM at a level of 25 instead of 2, which gives exaggerated results. Is the ring of brown beads -- 2-4 deep -- visible to everyone? If this is the way the tool is supposed to work, then they need to redesign it. Both the Clarity and LTM tools are similar, but the Clarity tool doesn't darken the picture. The LTM tool was at one time called the Clarify tool (in PSP 10). But darkening the outer ring of beads like that has nothing to do with clarifying the image. The disparity in appearance between the outer beads and the inner beads is much greater than any disparity you can see in the original picture.

In the original picture, you have a slight lightening of the beads around the edge because more light is getting to them. Why should Local Tone Mapping result in an image in which the outer beads are darker instead of lighter if the purpose of the tool is to "clarify" the image? If the tool were working properly -- i.e., if it were just "clarifying" the image -- the outer beads would remain slightly lighter than the inner beads, just as they are in the original picture.

Now, if this browning effect happened only at high settings like 25, I wouldn't object to it. But I can see the effect when applying even the lowest setting of 1. The purpose of the tool is to clarify the image, not to give us a strange-looking negative of the original. In short, the tool isn't living up to its stated purpose of clarifying the image by adding contrast; it is making light areas dark, which is not its purpose -- and in so doing, it is ruining the image, which makes the tool useless.

[hartpaul]

[ In the original picture, you have a slight lightening of the beads around the edge because more light is getting to them.

Which is why I suggested tripod and proper setup. Use a light tent and get your lighting even in the first place. If it is important enough for you, take more care with your shooting of the original photograph. Too many take a photo with undue care and then expect the editing program to correct the shooting errors.

It seems one of your original statements is untrue -

Since presentation isn't important, I just put the beads on a white plate and photograph them that way...

as well from looking at a couple of pictures on your site you only seem to show a smaller cropped area from around the middle.

BTW have you tried processing in PS to see if the same error bug exists there?

[CMurd]

Using a tripod wouldn't have made any difference as far as that picture is concerned. A tripod is really only good at preventing hand shake.

As for the proper lighting setup, I am using two bright 5000k bulbs on either side of the plate. The plate appears to be evenly illuminated to me. No matter what my lighting setup, the beads at the edge of a grouping will always get more light and appear slightly lighter than the beads in the center of a grouping -- and that is doubly true if the beads are semi-translucent, as these beads are.

I don't know why you say that my statement that presentation isn't important is untrue. I am selling raw materials to designers who make beaded jewelry. Their photographs require an attractive presentation, but my photos just need to show the beads accurately. I use a white plate because I can simply wipe the plate clean. It saves me from having to remove lint from felt (or whatever other material I might use).

I only show a crop area from the middle of the plate because it serves no purpose to show the entire picture. A crop that shows 80 beads is enough to give the customer an idea of what the beads look like. Believe it or not, many of my online suppliers put just ONE bead in their photos.

The problem with these orange beads is that the uniform color, and their semi-translucent quality, make them look like an amorphous mass. An additional problem is that Bayer-type cameras don't photograph reds and oranges very well. I decided to use the Local Tone Mapping tool to make them look a little more distinct. I had already tried the Clarity tool, and the results didn't look natural.

[hartpaul]

A tripod is really only good at preventing hand shake.

As for the proper lighting setup, I am using two bright 5000k bulbs on either side of the plate. The plate appears to be evenly illuminated to me.

Wow!! A tripod allows you to keep the same point of view, the same angle. It allows you to use a long exposure and move a single light around for more even exposure. It allows you to have a semi-permanent setup for consistency from one shooting to the next. If you are hand holding did you know that reflection off your skin / clothing, can influence the result - whereas a tripod would allow you to use the timer to expose and avoid that possibility.

If you are using those strong lights then of course the lighting will be uneven as the different distance from the centre to light and edge to light will produce different results. Try moving the lights further away so the Inverse Square Law has less effect.

The plate appears to be evenly illuminated to me ... yes the eye is s funny thing. Incident Lightmeter??
My reference to presentation referred to the whole picture and not whether it was on a plate or felt. It seems that you are worrying about parts of the picture that you are cropping off anyway .
Does Photoshop give the same result?