Skip to main content
Adobe Photoshop Family

46 Messages

 • 

562 Points

Sat, Mar 17, 2012 5:53 PM

4

Lightroom: Iimprovement suggestions on Clarity and Print Module

A couple of nits to pick about Lightroom 4, addressing these will create a much stronger product:

1. The clarity slider now makes substantial structural and tonal changes, especially visible on smooth surfaces like skin. This makes is much less useful and usable than the earlier versions. Yes, it prevents halos, but the result is generally unacceptable for me anywhere beyond 15-20 depending on the image.

2. The print module still does not have a good way of placing text with precise control below the image on the template. I have a workaround, but it is, well, a workaround. The watermark feature has a great set of tools, why not enable them for the identity plate placement on the page? The watermark feature allows for placing the text outside the image with negative offsets but then it is not visible! Why have this option and then hide the text?

3. I would love to have multiple line text in the print module with a simple graphic line without doing all that in Photoshop and bringing it as a graphic identity plate. See example of workaround and one output from LR at:

http://www.keptlight.com/2012/01/prin...

See what I would like to get direct from Lightroom, top left image at:
http://www.keptlight.com/shop/orchid-...

Responses

46 Messages

 • 

562 Points

9 years ago

I fully understood your description from your first post Sam. The unpredictability of the adjustment is a major problem for me as well. On some images it darkens the plain blue sky on others it lightens it. It is very, very difficult to understand the logic behind it. I am not a happy camper, I think everyone here knows it!

513 Messages

 • 

11.1K Points

9 years ago

Regarding the unpredictability:
The "wonders" of image adaptive technology it seems.

Every time software tries to be smarter than a human in creative matters, it fails one way or the other.

I haven't tried this (I guess LR 4.2 may be stable/useful enough for me to give it a go) yet, but unpredictability/inconsistency problems should also appear with other basic panel controls in the PV 2012 version.

A workaround for panorama stitchers is to apply the processing to the stitched image, rather than the individual images.

66 Messages

 • 

1.2K Points

9 years ago

Apart from the unpredictability, is anyone actually pleased with the cumulative effect? After all many commentators regard the new clarity/shadow tool combination to be great as a form of controlled HDR. Maybe it is the name that misleads? We are used to clarity in LR3 and really the new one isn't the same at all so ought to have a new name. Maybe also offer another tool actually based on the old clarity?

Also, in my opinion, I'd never try to apply clarity or shadows to images automatically. That seems to be asking for trouble as has been found. Even sharpness needs individual tweaks. That's my view anyway.

4.5K Messages

 • 

76.3K Points

New clarity is awesome on some photos. And I think it complements the rest of PV2012 nicely. I have my issues with it, but mostly I wish I could control where it's applied, without so much painting. Some nice auto-masking support would help, like a more sophisticated version of U-points...

Don't get me wrong - I miss old clarity too sometimes, when new clarity is doing "too much" (in some ways, and/or "not enough" in others...), but as I've said before, PV2012 naturally has independently controllable "clarity" (intra-region contrast/detail) in shadows, midtones, and highlights, so old clarity is not needed so much for that, like it was in Lr3.

Although it took some getting used to, I probably am applying new clarity to over half my photos now, but typical global value ranges between 3 & 9, which I consider fairly light. 10-20 I consider fairly strong, and 20-30 I consider very strong. I've only gone above 30 on a small handful of photos. And, it is not uncommon for me to spot reduce clarity here and there using the brush, or add an extra touch...

Regarding shadows: if I apply strong +shadows along with strong -blacks, it's possible to have "too much" intra-shadow contrast. I regularly shift to using more +blacks and reduce shadows slider until an appropriate amount of shadow detail has been accomplished. The point curve can be used to set black point if need be, and/or the tone curve can be used to further optimize shape of shadow tones if need be. I consider this skill to be essential for defining striking vivid shadows when desired or more subtle natural shadows if desired. (note: blacks extends further past the midline than shadows, so other compensations are necessary if similar overall tone and color are to be maintained. You can see the exact formula I am presently using for this in the PV2012 tone section of the cookmarks page - photo adjustment links. Look for Blacks/Shadows).

PS - I think most of any HDR-ish look a photo may have comes from improper adjustment of basic controls, and also from having more tonal detail than you are used to. The smaller part is due to the localized toning magic, e.g. new clarity (assuming it has been applied in moderation).

46 Messages

 • 

562 Points

9 years ago

Here is a simple test of the differences in the clarity behavior. Save the attached image, which is a simple gray scale ramp in two directions. Import it to LR 4, make a virtual copy and change the process to 2010 for that copy. Now, you have two identical images, one in 2012 engine and the other in 2010. Look at their histograms which should look like a lot of spikes

Watch what happens to the tonality of the squares as you increase the clarity in 2012, lighter squares will become darker and darker ones will get lighter. Observe the histogram as it bunches the tones in the middle and eventually creates a continuous histogram.

Now try the same thing in 2010. You will see minimal changes in the tonality of the squares and only where there are edges that can be made "clearer". You will also see the histogram generally unchanging in its distribution of the spikes but will form tiny heaps as the tonality around the edges change.

I bet if you try a different distribution of gray patches you will likely get a different behavior. I will leave that for you to develop and try.

For me, I have created a preset in LR4 which saves only the process version which I have set to 2010. I also marked that preset to apply on import. Now, any imported image will have the 2010 engine applied to it instead of 2012. I must admit, the 2012 engine handles overexposed images remarkably well. For those situations I will switch to 2012. Otherwise, I will mostly work in 2010 engine. I would rather know how a tool will behave than on some occasions working very well.

Cemal

4.5K Messages

 • 

76.3K Points

Thanks Cemal. Yep, the behavior of the 2 clarities is different alright.

513 Messages

 • 

11.1K Points

9 years ago

Good stuff, Cemal.

Who was asking for a change in the clarity behaviour again?

4.5K Messages

 • 

76.3K Points

9 years ago

I confess, I do get aggravated at having to correct for the "over-clarifying" effect of new clarity so much. Even at a value of 9 (one of my pet favorites), I often have to paint away the new brand of "haloing" - over-stratified areas designed to enhance local contrast... This comes up frequently in landscape shots where one mountain peak is followed by another - it can look very unnatural with even a moderate dose of clarity applied (imagine the squares in Cemal's example above were the mountain peaks). Dark backgrounds sometimes get "over-clarified" too but that's generally easier to correct, when it's not along a subject edge. etc...

Summary:
------------
In areas of a photo where there is no "right light" (meaning one really has no sense of how illuminated the various parts should be), new clarity is truly a fantastic improvement over old clarity. But, in other cases, there is a "what's wrong with this picture" thing due to the tonal redistribution of new clarity.

I wish the Lightroom clarity was designed with user-controllability in mind. Don't get me wrong - I like to paint, but it's very time consuming, and if not done with precision, results in "painting artifacts" instead, which can also look unnatural...

46 Messages

 • 

562 Points

9 years ago

My final wrap up and Lightroom 43 (four-three) solution ;-)

http://www.keptlight.com/2012/04/ligh...

Cemal

4.5K Messages

 • 

76.3K Points

Let us know how the "dual process version" workflow pans out, and which process version you end up using the most - OK?

427 Messages

 • 

7.7K Points

9 years ago

I made an animated gif using Cemal's image: Left PV2010 with clarity 0...100, right PV2012 with clarity 0...50 (using half the values of PV2010 as suggested):



You have to click on the image to see the animation (the preview is not animated).

From what I see, the new clarity actually *decreases* the global contrast, which really seems to be some sort of tone mapping...

By the way: Did you notice that for _negative_ values, there is NO difference between the old and the new clarity??

46 Messages

 • 

562 Points

9 years ago

Very nice demo, thanks. It makes the point very clearly (pun intended!) I will report how my Lightroom 43 experience works out. So far, no regrets, I seem to be getting the best parts of each for my needs.

Regarding the negative values of clarity, I do not have much experience in that since I rarely use it. I also have my biases about skin softening using any kind of blur carelessly (suprise!) I have several posts about this matter as well..

Cemal

4.5K Messages

 • 

76.3K Points

I used to double-edit in Lightroom sometimes in the interest of applying dual sharpening parameters or overlayed vignettes. Haven't done that for a while but it's worth considering for clarity now. i.e. Edit first pass using PV2012, and apply no clarity, or minimal clarity, then export a tif or jpeg back to the catalog and finish with PV2010 clarity. Note: you can still go back and edit with PV2012 and have that reflected in the final PV2010 result, just by re-exporting.

46 Messages

 • 

562 Points

Another idea for workaround, thanks Rob. This Lightroom 43 idea will kind of work, until Adobe sees the light and corrects Lightroom!

Cemal

46 Messages

 • 

562 Points

LRusser24, I forgot to say something about the global vs local contrast. In tone mapping a trade off is made, sometimes incorrectly, to favor local contrast over global contrast. That gives the image a flatter but crunchier look which a lot of HDR aficionados seem to like. The old clarity favored only local contrast. The new one seems to reduce global contrast and slightly increase the local contrast which helps to minimize the halos and makes the resulting image perceived "clearer", read as "slightly sharper". This also explains why the perception of sharpness quickly gives way to muddy tones, because the amount tonal shift overpowers the small gains in local contrast.

Cemal

4.5K Messages

 • 

76.3K Points

I'd be interested to know more about what other problems (besides clarity) you are having getting photos to look the way you want in PV2012. If it's the subtle differences in PV algorithms, then there is no solution, except to use PV2010. But my experience is that the largest part is often due to improperly adjusted basic sliders - not all of it, but >50% - eh?

46 Messages

 • 

562 Points

I will post a sample photo and make the RAW file available if you like to experiment. Stay tuned

4.5K Messages

 • 

76.3K Points

9 years ago

I liked the idea so well of being able to double-edit PV2012 photos using PV2010 that I automated it. The logic for this has been added to the Dev-Correct/2012 plugin:

http://www.robcole.com/Rob/ProductsAn...

Just invoke 'Edit with PV2010' to edit a PV2012 photo using PV2010 to take advantage of old clarity (or Lr3 fill-light). Note: Subsequent PV2012 edits will continue to be reflected in the PV2010 edited version.

46 Messages

 • 

562 Points

Thanks for the plugin Rob, I will give it a whirl. I'm off to a trip and may be inactive for a few days after tomorrow, but will resume reading comments and replying as appropriate.

66 Messages

 • 

1.2K Points

9 years ago



My image of a pretty cloud illustrates some of the problems. Incidentaly the sky looks far better on my PC than here but it serves its purpose.

The boundary between the cloud and the sky shows a halo effect on the right for example. I often use negative values on a gradiant over a sky to soften it a little in my landscapes.

4.5K Messages

 • 

76.3K Points

Looks like you have a pretty heavy dose of clarity in this photo. Still, I agree that all the hype of clarity halos being mostly gone is not entirely justified - now we have a new breed of haloing...

Note: be careful of clarity in local adjustments when converting PV2010 photos to PV2012 - Lightroom halfens global clarity but leaves clarity in local adjustments at full strength.

66 Messages

 • 

1.2K Points

The excess negative clarity is deliberate to give a glow to the clouds as part of an experiment. Saves using a third party plugin.

This type of haloing is the devil to remove locally though. Applying clarity locally or in a gradient however is a fantastic new tool. I love it. So what if you need to spend time on the images. You're worth it! :-)

ps I'm not a pro so time doesn't matter!

946 Messages

 • 

13.8K Points

Negative clarity is the same between PV2010 and PV2012.

46 Messages

 • 

562 Points

9 years ago

OK, here are some images to compare. You will see three JPEG exports, two from LR4 using 2010 and 2012 processes and just to make sure things are up and up, there is also an export of the same image direct from LR3. The best way to look at the differences is to load them to an image viewing app, even Lightroom (any version will do) and go back and forth between the 2010 and 2012 process images. You will see that the 2010 has a more open, airy feeling with better detail and contrast on the fish and the same on the net on the left, even in the shadow areas. The 2012 is the best I could do in LR4 yet you will notice the dull fish and less open net on the left. If you like I will e-mail DNG files which I believe will carry the process information so that you can see all the settings. Let me know what you think. Mind you, I use a calibrated monitor and I expect that you make the comparisons on a calibrated display as well to minimize monitor differences. If you want me to send the DNG files to you, drop me a note at ac{dot}ekin{at}keptlight{dot}com and I will e-mail them to you. I ask that after you experiment with them, you delete files.

These are the difficulties I am having, lack of finesse and subtlety while maintaining highlight contrast and detail.

Cemal

Direct from Lightroom 3


Export from LR 4 using 2012 process


Export from LR4 using 2010 process

66 Messages

 • 

1.2K Points

9 years ago

I've downloaded these and on my calibrated screen, can see what you mean.

However, it is impossible to know whether this really is the best you could do, if you see my point. It looks to me as if a tweak of the shadow slider in LR4 would have helped for example.

I actually used IrfanView with the files consecutive in a folder so I could instantly switch from one to the other back and forth very fast to highlight the differences, which to me are subtle. Without a comparison, either LR3 or LR4 versions would look good.

46 Messages

 • 

562 Points

9 years ago

In the basic panel, I have the shadows up to +28, whites +64, blacks -24, highlights -100. There is -14 clarity and +33 vibrance. Plus a little more fixing in the tone curve. What I miss are the sparkle on the fish, especially the group at the top, and the contrast and detail in the net on the left. The blue on the tray took a while to bring into line. But, you see my point I hope. I can get better results in 2010 with less fuss than I can get from 2012 with a lot of fuss. Yes, the difference is subtle but that makes all the difference in the final image, at least to me. A new and "improved" version of the product should make obtaining better results with less fuss possible, at least better results with the same effort. Alas! Lr4 does not deliver that to me.

Cemal

4.5K Messages

 • 

76.3K Points

I have been working Cemal's fish image image (he sent me the raw). My evaluation:

1. One can not get exactly the same result in PV2012 as PV2010, and in some ways, the PV2010 result seems better.

2. Trying to get exactly the same result is what lead Cemal to a set of adjustments that are not optimal for the best result in PV2012.

3. One of the main problems adjusting this image is that the range of tones adjusted by the highlights slider is much larger than the range of tones adjusted by the shadows slider.

re 3:
-----

slider behavior is not what one would expect: it exhibits some of the "role reversal" phenomenon I wrote about in another thread somewhere:

highlights slider does not recover all highlights, even at -100, and they are just barely clipped to begin with, or so it appears from the histogram.

whites slider does recover all highlights, and is more effective at moving the upper-most tones leftward than the highlights slider is. In some photos, the opposite is true - Neither Cemal nor I knows how to predict this in advance.

highlight slider extends across the midtones to the shadows.
shadow slider does not extend into the midtones.

I was able to adjust this photo with PV2012 fairly readily by:

1. Knowing to what extent the final result would differ from PV2012 due to inherent process differences. Note: To some extent, these are different on purpose, and are often good, but not always, and user can not control them. I'm talking about the handling of the darker shadow tones and lighter highlight tones, differences in fill-light algorithm, auto-highlight/shadow recovery, and different effect of clarity...

2. Knowing to check highlights/whites behavior and shadows/blacks behavior and use whichever works for what I want to do.

Final result in PV2012 is very pleasing to me - as good or better than the PV2010 result. The reasons PV2010 result may seem better, at least at first:

1. It has a way of illuminating parts that are hard to "reach" with PV2012 (in both shadows & highlight regions).

2. PV2010 created intra-highlight stratification in a way that was nice, and impossible to reproduce using PV2012, due to the way the highlight slider was moving such a large block of tones, and whites slider not able to separate highlights without blowing out the whites.

2. We are more used to the results of PV2010.

3. Optimal adjustment using PV2012 is very tricky.

Luckily, this is a photo that Lr4 clarity works well on, although at higher values, some spot reduction would be optimal to squelch some of the "over-clarification" in some areas.

Conclusion:
---------------

Another case where both results had their pros & cons, and both were awesome, but the best PV2012 result was trickier to get.

Final Thoughts:
--------------------

My opinion of PV2012 has not changed much since I posted my "Lr4 beta #1 - Final Results" in the beta forum:

* Potential for improved quality results.
* Basic controls are "squirrelly".
* PV2012 is not as controllable as I'd like.

What has changed is:

* My proficiency
* Development of tricks for those times when I want results that are different than the results PV2012 wants to give me.
* Some things were improved before Lr4.0 final was released.

Rob

4.5K Messages

 • 

76.3K Points

9 years ago

I have been working Cemal's fish image (he sent me the raw). My evaluation:

1. One can not get exactly the same result in PV2012 as PV2010, and in some ways, the PV2010 result seems better.

2. Trying to get exactly the same result (together with some idiosyncrasies of PV2012 that will be discussed below) is what lead Cemal to a set of adjustments that are not optimal for the best result in PV2012.

3. One of the main problems adjusting this image is that the range of tones adjusted by the highlights slider is much larger than the range of tones adjusted by the shadows slider.

re 3:
-----

slider behavior is not what one would expect: it exhibits some of the "role reversal" phenomenon I wrote about in another thread somewhere:

highlights slider does not recover all highlights, even at -100, and they are just barely clipped to begin with, or so it appears from the histogram.

whites slider does recover all highlights, and is more effective at moving the upper-most tones leftward than the highlights slider is. In some photos, the opposite is true - Neither Cemal nor I knows how to predict this in advance.

highlight slider extends across the midtones to the shadows.
shadow slider does not extend into the midtones.

I was able to adjust this photo with PV2012 fairly readily by:

1. Knowing to what extent the final result would differ from PV2012 due to inherent process differences. Note: To some extent, these are different on purpose, and are often good, but not always, and user can not control them. I'm talking about the handling of the darker shadow tones and lighter highlight tones, differences in fill-light algorithm, auto-highlight/shadow recovery, and different effect of clarity...

2. Knowing to check highlights/whites behavior and shadows/blacks behavior and use whichever works for what I want to do.

Final result in PV2012 is very pleasing to me - as good or better than the PV2010 result. The reasons PV2010 result may seem better, at least at first:

1. It has a way of illuminating parts that are hard to "reach" with PV2012 (in both shadows & highlight regions).

2. PV2010 created intra-highlight contrast in a way that was nice (separated, but smoothly), and impossible to reproduce using PV2012, due to the way the highlight slider was moving such a large block of tones, and whites slider not able to separate highlights without blowing out the whites. (as well as the inherent differences in PV2012 process version which can result in unsmoothness when there are strong bright highlight tones - this was improved in Lr4 final, but still there can be abrupt tonal transitions in brightest highlights).

2. We are more used to the results of PV2010.

3. Optimal adjustment using PV2012 is very tricky.

Luckily, this is a photo that Lr4 clarity works well on, although at higher values, some spot reduction would be optimal to squelch some of the "over-clarification" in some areas.

Conclusion:
---------------

Another case where both results had their pros & cons, and both were awesome, but the best PV2012 result was trickier to get.

Final Thoughts:
--------------------

My opinion of PV2012 has not changed much since I posted my "Lr4 beta #1 - Final Results" in the beta forum:

* Potential for improved quality results.
* Basic controls are "squirrelly".
* PV2012 is not as controllable as I'd like.

What has changed is:

* My proficiency
* Development of tricks for those times when I want results that are different than the results PV2012 wants to give me.
* Some things were improved before Lr4.0 final was released.

View final images (note: the differences are a lot more subtle in the web view than they are in Lightroom...): http://robcole.com/Rob/Personal/Pictu...

Rob

46 Messages

 • 

562 Points

Rob, thank you for putting so much time into this problem and specifically to this photograph. I have looked at all the versions you sent to me. Yes, with extra work in 2012 you have come close to my original vision. However, the adjustments you have introduced and their directions are anything but intuitive. Your summary above is very thorough and spot on as far as the 2010 and how it compares to 2012. Your knowledge of 2012 seems significant, but not a transferable adjustment slider on Lr, too bad!! Tools should not force their ways onto us, but do what we ask from them very well. Here we have a situation where even knowing what to ask from the tool has turned into a mystery.

Greetings from Istanbul.

4.5K Messages

 • 

76.3K Points

Cemal - Yer welcome, from SF.

Give yourself some time to adapt to PV2012.

Settings will begin to make more sense after a while and your proficiency will improve.

See more comments below.

Rob