How to fill shadows
Sometimes, simply adjusting shadows after getting exposure & contrast right is all that's required (see link above for a more Lr4-esque editing paradigm).
And, (thinking dangerously from a more Lr3-esque mindset), sometimes, simply doing
+shadows +exposure -highlights
does about the same thing as Lr3-fill. But, sometimes not...
Mastery of "fill" in Lr4 requires mastery of the blacks slider (in conjunction with the others) and knowing when and how to supplement with the tone curve. Shadows, exposure, (contrast), and highlight sliders alone do not suffice. Even whites slider plays into it too, sometimes critically.
1. What PV2012 thinks are highlights, midtones, and shadows (and blacks and whites) is not necessarily the same as what you think they are. And, it's not necessarily the same from image to image (part of the image-adaptive thang). It can be tricky because adjusting one zone always affects the adjacent zones. So adjusting shadows will always affect blacks and adjusting blacks will always affect shadows. But shadows aren't blacks and blacks aren't shadows.
2. Shadows slider may have a bit of a "blind spot" in the deeper shadows that may be "seen" (affected) better by the blacks slider.
3. Blacks & whites sliders work very differently than shadows and highlights sliders (all of which work very differently than the exposure slider, despite their sibling-like relationships in the histogram zone graph). The former (blacks & whites) have a very smooth effect from one end of the histogram to the other. Blacks affect blacks most, shadows next, midtones next, highlights next, and whites the least. Likewise, whites slider affects whites most, highlights next, midtones next, shadows next, and blacks the least. This is by design. In very different fashion, the design of the shadows slider is to affect shadows most, but have minimal effect on all other tones - required compensations only. Likewise, highlights slider is designed to affect highlights most, and have minimal effect on all other tones - required compensations only...
4. Sometimes the blacks slider has more effect on upper shadows than the shadows slider does (possibly because they're being defined as midtones by PV2012, and not shadows, and may be affected even more by the exposure slider). This is a result of the very different way blacks and shadows sliders work and the different ways PV2012 can define tone zones.
I like black. As a result, I learned -blacks very quickly. I steared away from +blacks because it would generally unseat my coveted blacks. Although +shadows would often pull up the blacks/deep-shadows enough, I would often have too much shadow stratification, or not enough deep fill. Early on, I learned an effective compensation trick (+whites -exposure), which compresses shadows from the right end leftward, and leads to some awesome highlight detail, but sometimes one does not want that highlight effect and really a certain kind of shadow fill was the primary goal. Recently, I learned that if one wants more deep shadow fill they need to use +blacks whether it unseats the blacks or not, then reseat the blacks using the tone curve, if need be. It's really that simple - sorry for all the buildup... But this was a critical missing piece from my puzzle for a long time.
I recently processed a photo where PV2012 had defined such a huge chunk of tones as highlights (and/or midtones, including "real" highlights and half of what I thought of as shadows too), and such a tiny block of tones as shadows, that adjusting either highlights or shadows in either direction did nothing but make the photo worse. All fill and highlight control had to be done using blacks & whites sliders.
Many photos fall somewhere in between "+shadows +exposure -highlights works perfectly - no blacks/whites required", and "+blacks -whites works perfectly, highlights & shadows sliders must stay 0". For those photos, blacks and shadow sliders must be balanced appropriately for that image...
Moral of the story: Never assume that what works on one photo will work on the next. You gotta learn to think like PV2012 does!
2nd moral of the story: You may need to adjust the blacks using the tone curve after achieving the proper shape to the fill by balancing blacks & shadow sliders (and exposure...).
From an Lr4-esque point of view, the bottom end is set using blacks, shadows, and tone curve, and mids are set using exposure.
From an Lr3-esque point of view (not recommended in Lr4), fill is defined by blacks, shadows, exposure, and tone curve.
If you can't get good fill in Lr4, it's because:
A. you haven't learned the basic PV2012 editing paradigm yet, or
B. you haven't fully unleashed the blacks slider yet! (simple tone curve compensation may be required, especially when using +blacks).
Think outside the box!!
* Nothing wrong with adjusting contrast to achieve whatever editing goals you have for your photo (quite the contrary), including fill quality, but it should not be necessary to drop contrast, just to achieve enough fill (but of course if ya gotta ya gotta...).
* New(est) Lr4.0 clarity has a strong propensity to blacken the bottom-most tones. If you are having trouble filling the deepest tones, consider lowering clarity.
* Lr3-fill has a quality of enhancing local contrast within deep shadows that you can not always achieve with Lr4. Often you can get close or better, but sometimes you just won't be able to get that look you like from Lr3-fill. Only recourse if you must have it is to switch to PV2010, or use local enhancements in PV2012.
* Camera profiles can also influence shadow tonability, and v4 profiles have very different black tone characteristics.
How to adjust highlight tone
Highlight toning is vastly improved in Lr4, by way of:
1. Automatic highlight recovery and clipping protection (roll-off...)
2. The whites slider.
3. Highlights slider.
re #1: You don't have to do anything to take advantage of this, but it allows you to do things you couldn't do before, namely up the whites, exposure, (contrast), and/or highlights without excessive clipping, and without having an ugly transition between clipped and un-clipped whites, and without losing good tone and detail in the highlights.
re #2: Arguably, the most important slider in PV2012 is the whites slider. Definitely the most important slider when it comes to optimizing highlight toning in PV2012.
re #3: Since the highlights slider is the main slider for controlling highlight "recovery" and setting level of highlights, most people discover it pretty quickly. But using it in conjunction with whites (and exposure) opens a whole new set of doors that are absolutely essential to getting the most from PV2012.
re #4: Because one can readily tweak highlights using the locals (without much affecting midtones or shadows), one can push global highlight toning much further when locals are to be used too. Of course one can take advantage of the locals even if not pushing global adjustments too hard ;-)
To make a short story long...
In Lr3, one had to use the highlight recovery to recover highlights, or decrease exposure, and/or keep brightness and contrast low. In Lr4 most highlights are recovered for you automatically. The rest can be recovered by:
Or any combination of the above (even if in conjunction with positive values from some of the others).
-highlights: pulls highlights leftward, doing a combination of recovery and stratification, but will not decrease midtone contrast in the doing.
+highlights: increases highlight level while attempting to minimize compression of highlights/whites. Like everything, protected by automatic roll-off logic.
+exposure: increases all tones smoothly, from low to high, but protects highs from getting too too bright. Some like to think of it as being somewhere between PV2010 exposure and PV2010 brightness. The exposure slider has been designed to complement the other sliders. For example, increasing exposure increases highlights in a way that -highlights can bring back down. And since -highlights does not reduce midtone contrast, one can use +exposure -highlights to brighten the midtones in an image without making it look washed out like increasing brightness in PV2010 would.
-exposure: decreases all tones smoothly, but tries to keep highest tones from getting too dim - this is a big difference between PV2012 and PV2010.
+whites: stretches entire tone curve rightward, smoothly - whites most, blacks least. Very much like a white-biased version of exposure. Very similar to the right-side equivalent of blacks slider. +whites also supports graceful driving into clipping. Use a brush with -highlights to recover highlights driven a little too far.
-whites: recovers clipped highlights and moves them leftward along with all other tones smoothly, all the while and nevertheless attempting to keep highest tones high, after they've been recovered.
PV2012 includes some fancy programming that allows us to do things that were not even close to possible in PV2010. The fact that the positive function of the sliders is not necessarily the exact inverse of the negative functionality is representative of this sophistication. And the slider behavior can change dynamically depending on image and other settings... This can make it tricky, but also makes it much more powerful than one might think given it's modest basic slider set.
In Lr3, fill always did the same thing, and highlight recovery always did the same thing. While predictable, they did not allow one to optimize image quality and tone, nor provide the same level of flexibility as PV2012.
One of the things I don't like about PV2012 is that the highlights, by default have a muted or processed look I sometimes describe as "powdery", as opposed to unadulterated and glowy - more detailed to be sure, but a bit flat.
One of the things I do like about PV2012 is that the whites slider allows one to restore much of the original highlight quality, and more.
To shorten a long story:
I strongly recommend to use +whites to the extent the image allows in order to have maximally shiny and detailed highlights (unless that's not what you want of course). You may have to decrease exposure and/or highlights, and increase shadow level as compensation.
In a nutshell:
+whites is good - use it to it's fullest!
And on the flip side:
Sometimes -highlights just won't cut it. Either it just can't do the job (long story not told) or it produces too much highlight stratification. In either case, that's when -whites steps in. Knowing when to use -whites is as important as remembering to use +whites whenever possible, but is required far less frequently on most images.
To increase intra-highlight detail and/or unflatten highlights: +whites, usually accompanied by some combination of -exposure and/or -highlights, and +shadows as compensatory adjustment.
To recover highlights that would otherwise be too hard to reach, (or in cases of unusual tonal distribution, when -highlights compresses shadows too much), and/or to compress tonal range from the right-end leftward: -whites.
In PV2012 it is generally not necessary to reduce contrast to recover and stratify highlights, although you can if you want ;-)
Advanced bonus formula:
+whites +highlights -exposure +blacks +shadows +contrast (-saturation) to unflatten highlights, but minimize highlight stratification.
+whites: the workhorse - required for unraveling auto highlight recovery which is primarily responsible for the flattening.
+highlights: keeps highlights from being separated as much when +whites is done.
-exposure: keeps top end from being too bright due to the above, but also darkens bottom end.
+blacks +shadows: raises bottem end back up and pushes on highlights re-compressing them from the left.
+contrast: further compresses highlights. Note: due to +blacks +shadows, +contrast won't darken bottom end too much or stratify midtones as much.
(-saturation): keeps +contrast tonal.
OK, enough for now - sorry for putting you through all this.
How to reduce shadow detail
Lr4 (PV2012) recovers and enhances shadow detail. This is most often, but not always a good thing. This post is about how to suppress shadow detail when it is more distracting than pleasing. It relies on some differences in how blacks and shadows sliders work in Lr4, to review:
+shadows will usually unanchor and brighten blacks more than +blacks will (similar to how -highlights will usually drop whites more than -whites will).
Which means -shadows will tend to compress blacks more than -blacks will.
Note: Otherwise, +blacks has a very similar effect in the shadows as +shadows does, but (unlike +shadows) +blacks extends it's effect all the way into the highlights...
The guts of the formula is:
But other things are required to confine and enhance it's effect, namely:
+blacks -shadows -highlights +contrast & some tone curve manipulation.
- +blacks fills the shadows whilst -shadows unfills them, but the combined effect is compression of shadow tones, especially deeper ones. But it raises highlights too.
- -highlights drops the highlights back to where they were, and then some.
- +contrast further compresses the shadows and lifts the highlights back to where they were, plus restores midtone contrast lost by the +blacks adjustment.
- tone curve further enhances and customizes the desired result.
For details, and the exact formulas presently being used for this, please see the PV2012 tone section under "cookmark photo adjustment links" on the cookmarks page:
The problem with attempting to use -blacks to suppress detail in the deeper shadow tones is that it can end up clipping too much and still not really get rid of the detail in the shadows, in fact just the opposite, since it actually increases intra-shadow contrast.
How to get the most natural looking results
In PV2012 one can get a variety of results, all of which may have very similar overall tone and color, but some will look more natural than others.
What accounts for these differences?
PV2012 has some sophisticated programming which allows one to extract intra-region contrast and detail, or have one part be disproportionately bright without causing loss of detail or contrast in adjacent regions... - but sometimes the results don't look organic. Interesting and/or punchy: yes, but natural: not necessarily.
The trick for the most natural looking results is to avoid the things that result in rearranging original tonal relationships.
Moderation in all things..., but especially highlights and shadow sliders - they are the ones that most allow separation and recombination of tones in potentially "un-natural" ways.
Assuming you have already chosen a camera calibration profile which is closest to what you want without much adjustment:
As always (in PV2012 especially), exposure needs to be set right, but then:
If you still need some fill light (after reducing contrast as much as you dare), then favor +blacks over +shadows. If that makes the top end too bright, then bring it down using -exposure, -whites, and/or -contrast - not the highlights slider. Likewise, increase contrast or -blacks if need be (but not so much as to require +shadows and/or -highlights as compensation).
+whites may still be a good thing, but only to the extent that you don't have to use -highlights to compensate (-exposure to compensate if need be).
If you do decide to use -highlights and/or +shadows, just go easy, and use the same (or nearly same) value for -highlights as +shadows, since that combo results in a smoother distribution of tones most like the original photo.
You can use +exposure -highlights -shadows or -exposure +highlights +shadows to brighten or dim the midtones without impacting highlight and shadow levels (or +highlights -shadows to add "contrast" without making it look "so contrasty"), but again - go easy...
Finish with a smooth tone curve, if need be (after basic sliders).
PS - go easy on the clarity.
Note: If you've converted a PV2010 photo to PV2012 and it doesn't look as natural, in addition to checking all the global things mentioned above, also check the locals for clarity, exposure, and/or -sharpening adjustments.
New clarity - how to keep it's effect from being too black, and how to confine it's effect more to the midtones.
Lr3 clarity did not clarify shadows.
Lr4 introduced a new kind of clarity that also affects shadows.
Lr4-beta clarity mostly brightened. Lr4-final clarity mostly darkens. There are advantages to each:
Lr4-beta: Not much of a problem with over-blackness, but then one may have had to drop exposure and bump saturation after using, maybe even take some additional noise reduction actions.
Lr4-final: Results usually look more natural and pleasing, and does not desaturate as much, and exposure need not be downthrottled after using..., but it darkens and sometimes drives the blacks down too much.
So what can you do?
why +blacks along with +clarity, of course, but then that will brighten highlights too, so final formula:
+clarity +blacks -highlights
But that will still clarify shadows a lot, to reduce clarification of shadows and confine more to midtones:
+clarity +blacks -shadows -highlights -contrast
+clarity: clarifies midtones and shadows, plus darkens.
+blacks: brightens (bigger dose than in previous formula) all the way into the highlights.
-shadows: recompresses shadows. along with the extra brightening of highlights, leaves the image over contrasty.
-contrast: normalizes midtone contrast
-highlights: brings highlights down.
You can view and/or use this formula here:
(PV2012 tone section under cookmark photo adjustment links).
How is PV2012 different than PV2010?
PV2012 can be a challenge to learn, especially when coming from PV2010 experience.
PV2012 is, in general, better than PV2010, although some people are not feeling it yet.
1. People are more used to the characteristic look of PV2010.
2. Editing with PV2012 has not yet been mastered.
What are the differences in characteristic look?
* PV2010 fill-light has a personality. Color aberrations are sometimes delightful, and it has a way of exagerating local contrast in the deeper shadows that can not be reproduced in PV2012, although new clarity fills the void to some extent.
* PV2012 recovers clipped/near-clipped highlights and shadows automatically, PV2010 doesn't. Thus deep shadows and highlights tend to have more detail in PV2012. This additional detail is often a blessing, is sometimes distracting, but in any case it's a difference that takes getting used to.
* In general, one can often boost contrast and brightness (in all three major regions) further in PV2012 than PV2010 without image quality degradation. Thus, PV2010 photos are often dimmer than their subsequent PV2012 edited versions end up being. In my experience, it's easy to under-brighten images in PV2010 (to maintain "richness", and highlight detail), and it's easy to over-brighten images in PV2012 - because you can...
* PV2012 tries to maintain intra-shadow and intra-highlight contrast by keeping deepest shadow tones dark and lightest highlights bright as you edit. This is often responsible for gorgeous clear photos with bright detailed highlights, it may not be what you want, and in any case takes some getting used to...
* PV2012 uses a different algorithm for enhancing local contrast, which is usually better, sometimes not, but different and takes some getting used to.
What are the differences in editing with PV2012
The controls are similar, but different, and I shan't try to outline individual differences or usage here - that's covered elsewhere. But, what I would like to say is:
* PV2012 control behavior not only adapts to individual images (can be different from image to image), but the control behavior can change as you edit an individual image.
Read that last bullet again - it's been a major stumbling block for me, and some other folk...
What that means is:
* Highlights slider may not behave the same on the next image as it did on the previous image (for example).
* Highlights slider may not behave the same on this image after increasing exposure as it did before increasing exposure (for example).
This is usually a good sensible valuable thing, and is a primary reason PV2012 has so much power and flexibility with so few controls. But, it can be a problem (or confusing) if you have expectations.
So, my advice to you:
Try to be as flexible with PV2012 as PV2012 is with your images. Once you learn to go with it, instead of fighting it, things really start to flow...
PS - Personally, I would forget the notion of doing basic toning using presets. Learn to be proficient at basic toning using the sliders, and save your presets for color and style...
My hat is off to you or this post, very informative and has gotten my grey matter stirring, dangerous for a Saturday night. Maybe you could give me some guidance?
I shoot multi-row panoramas, sometimes stitching up to 200 images. I shoot in manual exposure mode with manual focussing. This is to ensure I capture the scene as "seen". I have experiemented with PV2012 selecting my prime image, make a couple of adjustments, then synchronize the remaining files. Then export to my Pano directory, then fire up my Pano software (Auto Pano Pro) and it has a dickens of a job stitching and just looks aweful. Variations in the sky tones and bunch of other atrifacts. Whereas using LR3.6 I get a good pano.
So do you have a tip for PV2012 where I can apply the same changes equally across all images in the pano structure or should I stick with the earlier PV?
Cheers and thanks for all the effort you have put in here
I don't have personal experience with Panoramas that use PV2012, but from what I know, there is definitely no way to apply a common set of adjustments to different components of a Panorama and expect to have consistent results across the components, which is what you've already experienced, if I got what you said in that last post.
Certainly you'll have problems if you've applied highlights, shadows, blacks, or whites, since their effect can vary depending on the image. And certainly clarity is problematic, since the tone of the same area of sky (or anything else) can be different depending on surrounding tones...
I dont know if its possible to get a uniform result using only exposure and contrast. I suspect even that may be problematic too. Perhaps somebody with more knowlege will chime in, or maybe you could try it and let us know.
I do think it may be possible to get a uniform result if all adjustments are '0', although I'm not even sure about that part - is there some baseline laplacian magic applied even at zeros? - dunno really. My knowlege comes mostly from experience adjusting ordinary photos.
I'd be curious to hear if you try either of the above mentioned scenarios how that works out.
It's entirely possible that you'll have to stick with PV2010 for the components. Of course you could still use PV2012 for the final result after stitching.
Sorry I wasn't more help.
Your reply was very helpful, and I think I have made a decision on my work flow; and will only use LR for cataloging my final panoramas and not the content they are comprised of.
Panos are constructed from overlapping (around 20%) images and the software builds control points based on recognizing overlapping components of the scene as well as tonal values, a good example here is sky which can be a challenge in it's own right. So to work effectively, the tonal values and other stuff, need to be consistent from frame to frame (that is why I shoot in manual mode).
So looks like I will just load my RAW images into APG as is, have that do the rendering and export as a 16 bit TIFF then import into LR for cataloging and maybe printing.
Sounds good. One advantage of having APG handle the raw input is you don't have to save a tif for every component image. As long as APG can make a good enough rendering of the components (?) If not, you can always use Lightroom for the components after editing with PV2010. Also, Photoshop makes good panos, no?
Hate to say it, but Photoshop makes just OK panos, but really hit and miss when it comes to large panos. Crashes a lot is not very efficent with system resources compared to APG. But then APG is dedicated software and has the ability to manually align images if needed. But has a nice feature in that you can export layers as a PSD file which combined with an Alpha Channel can make corrections like someone walking through the scene and showing up in multiple frames.
Thanks for your insight, now heading off for a day's shoot, tripod and pano head over my shoulder freshly charged batteries and off to do what I enjoy most.
How to edit a photo quickly, starting with Auto-tone
Auto-tone is notorious for creating results that don't look good at first. But, it often does some good stuff that you can take advantage of, if you start with it, but don't end with it.
After clicking auto-tone you can get the most bang for your buck, by making adjustments to these sliders, listed in order of most bang for buck first:
So, here's a procedure that works on a lot of photos:
1. Click Auto-tone
2. Coarse adjust exposure - if photo is obviously too dark, +exposure, if obviously too light -exposure. Don't try to make it perfect, yet, but this needs to be done most of the time, but not always - sometimes a lot, sometimes a little...)
3. Adjust blacks slider a little (or a lot), if need be, usually leftward. The objective of moving leftward is to remove the milky/washed-out look of the shadows, if it exists, but without over-clipping, or to clip a bunch, if it's one of those photos.
4. Fine tune exposure. Many photos may look good enough at this point, depending on your standards...
5. If photo too contrasty (e.g. looks heavy/over-saturated, and darks are very dark, highlights are bright but lack detail...) then -contrast. If photo is too flat looking, then +contrast. (consider a small adjustment to exposure and/or blacks again if contrast was altered substantially in this step).
6. If highlights are a bit bright, -highlights, although consider -exposure first. (or, if too dim, +highlights, although consider +exposure first).
7. If shadows are a bit dim, +shadows, although consider +exposure first. (or, if too bright, -shadows, although consider -exposure (and/or +contrast) first).
8. Try moving whites a little to the right, and left, then leave it how you like it best.
Repeat steps as necessary until you get it how you want it...
Auto-tone trys to distribute tones to fill the histogram, with minimal clipping. Sometimes in so-doing it:
* goofs exposure (or maybe you just want a lower-key (or higher-key) photo...).
* insufficiently seats the blacks
* makes for an under (or overly) contrasty image
* makes highlights and/or shadows non-optimally adjusted (especially when other compensatory adjustments are made by you after auto-toning), - ditto whites.
Although frustrating when exposure is much worse than before auto-toning, more often than not it gives a better starting point, after correcting gross exposure snafu that is, and blacks...
This may not work for all photos, but for most photos: yes, eh?
To adjust basics sliders in the order presented, or not to - that is the question. And why...
Disclaimer: I don't fully understand all the nitty gritty details about this, but I do understand a lot, and have experience with top-down order and other orders.
From what I glean, there are two motivations for top-down:
* There are some slider behavioral dependencies.
* Sliders do not necessarily do what one might think.
Examples of things NOT to do:
* Adjust whites slider first, as if it where a normal tone-curve/levels tool, right-hand slider - it's not.
* Adjust blacks before exposure: It's behavior is not like a normal tone-curve/levels tool, left-hand slider.
* Try to fine tune highlights and/or shadows before exposure and contrast are set (but in my opinion, coarse tuning highlights and shadows before contrast is set is AOK, ditto for blacks & whites).
The single most important slider dependencies:
* The behavior of contrast, highlights, & shadows is dependent on exposure.
A big part of the value of doing things top down is that it helps to avoid the things one shouldn't do. But, it has it's limitations...
First step however is *always* to coarse tune exposure (or click auto-tone and then coarse tune exposure).
*if* black point is pretty good from the get go, then doing contrast before shadows and finishing with blacks works out just fine. But that is not always the case, and when it's not, adhering to the top-down regimen will lead to suboptimal toning. It's better in this case to do blacks after coarse tuning of exposure.
Likewise, if white point will need a strong adjustment, it makes no sense to try and fine-tune highlights first, or contrast, or even to set exposure in stone first. After coarsely setting the exposure, you may realize you are going to want a relatively large +whites value. At that point, it may be better to do it, so you can balance the exposure setting you'll be having, then go back and set the contrast and/or highlights.
And sometimes one may need large -whites and/or +blacks values. If that will be the case, it's better to set those before contrast and highlights/shadows.
PV2012 rocks! (once you learn how to use it).