2019 P52 - 1st Qtr - Sandy(Skitterbug)

[Manfred M]

Ted - I think you are not looking at it in the same way I am.

I (and many other photographers) have found that in cases where the camera indicates blown highlights via either the "blinkies" or via the histogram displayed by the camera, the raw data downloaded from still has at least 1 stop of "headroom" where the data is not blown out.

Here is an image I shot in JPEG and raw; both are using the AdobeRGB colour space.


If I import the JPEG image into Adobe Camera Raw, the red areas indicate clipped highlights





When I do the same thing with the raw file, I get this:




The JPEG has a lot more highlight clipping...

[xpatUSA]

Ted - I think you are not looking at it in the same way I am.

I (and many other photographers) have found that in cases where the camera indicates blown highlights via either the "blinkies" or via the histogram displayed by the camera, the raw data downloaded from still has at least 1 stop of "headroom" where the data is not blown out.

Here is an image I shot in JPEG and raw; both are using the AdobeRGB colour space.


If I import the JPEG image into Adobe Camera Raw, the red areas indicate clipped highlights





When I do the same thing with the raw file, I get this:




The JPEG has a lot more highlight clipping...

Perhaps my test was not the same as yours because a) I don't have a Nikon and b) I don't use ACR.

Still, from what you said, a Panasonic should be no different. So a JPEG with clipped highlights should be much more clipped than it's raw opened, converted and rendered in any converter - if I understand you correctly - and more clipped by "at least one stop".

So, using your example method instead:

OOC JPEG opened in my converter/editor with the blinkies (black) set to 253 (default):



RAW rendered in my converter/editor with the blinkies (colored black) set to 253 (default):



I see no sign of "at least a stop" of headroom at all, even after following your method but with my Panasonic.

Perhaps the effect you describe applies only to certain certain cameras and certain converters?

Just for completeness, here's the raw histogram for the above:



Totally blown, no headroom at all. Perhaps your interpretation of "headroom" is different.

[Manfred M]

Ted - My definition of clipping is showing a value of 255 in any individual channel. Same applies at the other end of the scale, a 0 value in any channel represents a crushed shadow detail.

The JPEG image I showed are very much in line with what I saw on the back of my camera when I took the shot. The second image is the histogram from the raw data, with all adjustments turned off. That tells me that I am not picking up any lens correction, contrast enhancements and sharpening built into the JPEG data. I'm also using 16-bit, rather than 8-bit files. I'm using Photoshop with Adobe Camera Raw to evaluate these images.

If I look at the back of my camera when shooting and I see the first image I shot, when I go down one stop or so, I will see something quite similar to the second shot I posted.

This empirical method seems to work well and is used by a lot of very competent photographers. I can only suggest that there is something you are missing in your analysis.

[xpatUSA]

Ted - My definition of clipping is showing a value of 255 in any individual channel. Same applies at the other end of the scale, a 0 value in any channel represents a crushed shadow detail.

The JPEG image I showed are very much in line with what I saw on the back of my camera when I took the shot. The second image is the histogram from the raw data, with all adjustments turned off. That tells me that I am not picking up any lens correction, contrast enhancements and sharpening built into the JPEG data. I'm also using 16-bit, rather than 8-bit files. I'm using Photoshop with Adobe Camera Raw to evaluate these images.

I seem to have failed to do anything in the same way as your good self, Manfred! Short of going out and buying a Nixon so I can chimp my stuff, I'll just bow out of this discussion.

If I look at the back of my camera when shooting and I see the first image I shot, when I go down one stop or so, I will see something quite similar to the second shot I posted.

This empirical method seems to work well and is used by a lot of very competent photographers. I can only suggest that there is something you are missing in your analysis.

I thought the analysis was simple enough but oh well ...

[skitterbug]

I've been holding onto this entry a bit because I didn't want to derail the discussion about light and clipping. Interesting view points and interesting information.

But here is Week 5 - I did make it out for snow pics. I haven't narrowed it to what I want to work with, if any. But I had the opportunity to catch this beautiful sunset (to me anyway). Hopefully I brought out the tones a bit. I'm wondering if I should have lightened the image a bit more in PP? I didn't because I wanted the sky to keep it's blaze....... Comments are welcome.

Settings used: ISO125 - f/4.5 - 1/100
Before Editing:



After Edit - Sunset during a winter evening

[Cantab]

Sandy, it was good to see another picture appear in your P52 thread. The photo gives a good impression of a cold evening. (Arctic air has reached us out here on the west coast but our low last night was 22 degrees Fahrenheit— not what you and others have dealt with back east.)

If this photo was mine, I’d be curious what it would look like with some very selective brightening. One possibility might be to brighten only the red/orange hues; that can be done in Lightroom. It might not change the image for the better but it would be free to try. I’m looking forward to reading the comments of others.

[Geoff F]

I would be wary of doing any more, Sandy. The red/orange area at the centre point is approaching maximum level now. You could try slightly brightening the edges and bottom but I fear it may begin to look unnatural.

At the moment, you have brightened the foreground while darkening the top sky area which is looking good to me.

[xpatUSA]

I've been holding onto this entry a bit because I didn't want to derail the discussion about light and clipping. Interesting view points and interesting information.

But here is Week 5 - I did make it out for snow pics. I haven't narrowed it to what I want to work with, if any. But I had the opportunity to catch this beautiful sunset (to me anyway). Hopefully I brought out the tones a bit. I'm wondering if I should have lightened the image a bit more in PP? I didn't because I wanted the sky to keep it's blaze....... Comments are welcome.

Lovely shots, Sandy!

For cloud enhancement, there's a useful technique called 'wavelet processing'. With an image like yours, one can play with things like contrast based on the size of objects in the image. With your clouds, playing with the 8px size slider had a marked effect, plus a little added at 2px and 16px.

I liked the purple cast in the sky above the clouds, so I gave it some more.

And finally, I personally like trees in the middle distance to be sharp, artistically incorrect of course. A bit of what's called "de-convolution" fixed that.

Pardon me messing with your stuff:

[Cantab]

Lovely shots, Sandy!

For cloud enhancement, there's a useful technique called 'wavelet processing'. With an image like yours, one can play with things like contrast based on the size of objects in the image. With your clouds, playing with the 8px size slider had a marked effect, plus a little added at 2px and 16px.

I liked the purple cast in the sky above the clouds, so I gave it some more.

And finally, I personally like trees in the middle distance to be sharp, artistically incorrect of course. A bit of what's called "de-convolution" fixed that.

Ted, I'd not heard before of wavelet processing. In the context of photo editing, googling did not help me much. Is this something that's doable with ordinary photo editing software or is restricted to editing by retired engineers?

I've heard of deconvolution techniques for sharpening (electron?) microscope images but have the vaguest of ideas that the term is broad enough to cover "ordinary" photo sharpening algorithms. A number of years ago, I once went so far as downloading a deconvolution program in the hope of cleaning up a blurry photo of an eagle with a duck but had no success. Did you use ordinary sharpening or something more sophisticated?

[Stagecoach]

And finally, I personally like trees in the middle distance to be sharp, artistically incorrect of course. A bit of what's called "de-convolution" fixed that.

I'm not sure that adding artefacts and halos around the trees is a good idea

[joebranko]

I like the PP you did. Particularly raising the shadows and cropping the foreground . Good job!

[Manfred M]

Ted, I'd not heard before of wavelet processing. In the context of photo editing, googling did not help me much. Is this something that's doable with ordinary photo editing software or is restricted to editing by retired engineers?

I've heard of deconvolution techniques for sharpening (electron?) microscope images but have the vaguest of ideas that the term is broad enough to cover "ordinary" photo sharpening algorithms. A number of years ago, I once went so far as downloading a deconvolution program in the hope of cleaning up a blurry photo of an eagle with a duck but had no success. Did you use ordinary sharpening or something more sophisticated?

I believe ACR / Lightroom and Photoshop use these techniques, but give them names that are more useful to photographers. I suspect that "Clarity" is probably using deconvolution algorithms as the functionality primarily introduced local contast a.k.a. microcontrast.

[xpatUSA]

Ted, I'd not heard before of wavelet processing. In the context of photo editing, googling did not help me much. Is this something that's doable with ordinary photo editing software or is restricted to editing by retired engineers?

I've heard of deconvolution techniques for sharpening (electron?) microscope images but have the vaguest of ideas that the term is broad enough to cover "ordinary" photo sharpening algorithms. A number of years ago, I once went so far as downloading a deconvolution program in the hope of cleaning up a blurry photo of an eagle with a duck but had no success. Did you use ordinary sharpening or something more sophisticated?

Hello Bruce, thanks for asking.

For quick wavelet contrast processing, I used RawTherapee's Contrast By Detail Levels:

http://rawpedia.rawtherapee.com/Cont..._Detail_Levels

RT also has Richard-Lucy de-convolution sharpening as well as the more common unsharp masking:

http://rawpedia.rawtherapee.com/Shar..._Deconvolution

That's what I used on the trees, albeit a bit OTT - as has already been seized upon by Grahame in this thread.

In addition to Contrast By Detail Levels, RawTherapee has an entire tab devoted to wavelet processing, not just contrast:

http://rawpedia.rawtherapee.com/Wavelets

Probably too much for most of us

Not being an Adobe product, RawTherapee appears to be alien to many here - a pity because it has some seriously good functions above and beyond simple Adobe stuff such as "Clarity".

[skitterbug]

Wow, thanks for all the comments!

Bruce - I did try brightening the image a bit but didn't like my results. I kept ending up with washed out sky rather than making it pop. And you are right, it doesn't cost a dime to play! I do a lot of that!

Geoff - Glad you like it and yes, when I did try shifting the colors a bit, it ended up looking contrived rather than a real evening sunset.

Ted - you are definitely welcome to work with any photo I put in the P52 project. I'm here to learn and to get opinions from others! I have RawTherapee software but I've not done much with it. I mostly use Lightroom (desktop) and Affinity Photo. I'll check out your links.

Manfred - Thanks for the additional info. I do have "Clarity" as a tool in Lightroom and in Affinity Photo.

Joe - Thanks for the praise! Makes me smile! It was a beautiful evening!

[Stagecoach]

I did try brightening the image a bit but didn't like my results. I kept ending up with washed out sky rather than making it pop.

Sandy, if you want a very easy way of getting images such as this to 'Pop', whilst being able to protect and preserve highlights near the limits as in this sky, download the NIK collection and use the 'Pro Contrast' filter in the 'Colour Efex' module.

[george013]

Ted - I think you are not looking at it in the same way I am.

I (and many other photographers) have found that in cases where the camera indicates blown highlights via either the "blinkies" or via the histogram displayed by the camera, the raw data downloaded from still has at least 1 stop of "headroom" where the data is not blown out.

Here is an image I shot in JPEG and raw; both are using the AdobeRGB colour space.


If I import the JPEG image into Adobe Camera Raw, the red areas indicate clipped highlights





When I do the same thing with the raw file, I get this:




The JPEG has a lot more highlight clipping...

A late reaction. But you're using 2 different converters. The in camera converter and acr. You can't compare the jpg made out of the camera's converter with the result of a conversion with another converter.

George

[Manfred M]

A late reaction. But you're using 2 different converters. The in camera converter and acr. You can't compare the jpg made out of the camera's converter with the result of a conversion with another converter.

George

George - if I was merely trying to compare what two different raw convertors do, then I would agree with you, but that is NOT what I have been writing about. This is all about a comparison to what one sees on the camera's display and how that compares to what I see in Adobe Camera Raw.

The first image effectively what I saw on the camera's display. The whole background was blown out. When I open the image in Camera Raw, I the second image is what I see when I open the image. That tells me (and many other photographers I have discussed this with) is that just because we see "blinkies" on the back of the camera, it doesn't mean that the image is actually clipped. What the second image shows is approximately what I see on the camera display when I stop down one full stop.

This suggests that seeing a bit of blinking on the camera display should not automatically force the user to stop down to eliminate the "blinkies" as the raw data still has "headroom" in the highlights. My testing with my camera suggests I have someone over one stop of "headroom" when I see "blinkies" on my display. Every camera is going to be a bit different, so testing is worthwhile. Just don't panic if you see a few "blinkies" if you shoot raw, as this seems to be a function of the way the camera has been set up, rather than reflecting blown out highlights.

In the two images I posted, the "blinkies" were so extreme that even the raw file contained might contain unrecoverable data.

If I push this same image in my raw convertor by pulling back the highlights and the whites, I can bring the whole image back to a point where there is no highlight data loss. Both these sliders have been pulled almost all the way to the left and this is quite an extreme example of this. This is dependent on the dynamic range of the camera and with the ~ 14 EV that my camera has, I can (just barely) recover the highlight details. Some cameras will have slightly better performance, but most will not have this much dynamic range to work with.

[skitterbug]

Sandy, if you want a very easy way of getting images such as this to 'Pop', whilst being able to protect and preserve highlights near the limits as in this sky, download the NIK collection and use the 'Pro Contrast' filter in the 'Colour Efex' module.

Hi Grahame - Thanks for the suggestion! I'll check NIK 'Colour Efex' module out. I have a version of it on my computer. Just haven't used it very much.

[skitterbug]

I think maybe I'll be caught back up with Week 6 posting. I didn't like the results of my snow pics but I did like the one I'm going to display.

I didn't do a lot of post processing other than some tweaks and a crop.
The settings I used for the capture: f9.0 - 1/160 - ISO125. I used the 100-400mm lens, manual mode and I bracketed 3 exposures. I liked the results of the exposure using EC +2/3.

Comments definitely welcome!

Before edits:



After Edits:
Frozen Bird Food!

[xpatUSA]

Hi John - To start my reply, I'm chuckling because of the applications I have regarding photo processing:

1. Aurora HDR
2. Luminar
3. Raw Therapee
4. Gimp
5. Apple Photo
6. Adobe Elements 6 that I never did figure out very well
7. Lightroom 6 that I use to import photos for the start of my viewing process and some digital processing
8. Digital Photo Professional 4 that I also use to view photos to see where my focus point actually ended up
9. NIK free apps
10. Affinity Photo
11. Snapheal

Whew.... I didn't realize how many I've tried over time!

Quite a list, Sandy!

Of those, I too have RawTherapee - now up to V5.5
The GIMP, 16-bit V 2.9 by Elle Stone - a scion of color management
Elements 6 on a disk somewhere - never use it these days.

The GIMP ** is useful for it's color de-composition function - for example splitting into 3 layers showing hue, saturation and value (brightness). Not the easiest of User interfaces.

RawTherapee has more of everything! Again, not easy.

I also have FastStone Viewer which has a really good comparator - perfect for weeding out my many bad shots ...

** I say "the GIMP" because that is how it's commonly said, being short for:

"the GNU Image Manipulation Program"

https://www.gimp.org/

Looks like V2.10 is out ... tempting!