Susan S. wrote:

 

That's fine Rob - I don't think you misrepresented me. Maybe overstated my competence (but that isn't going to upset me any!)

R

Snow business.

PS - If white balance and/or split toning doesn't get rid of that cast, try the RGB channel curves, eh?

Indeed, heavy on the +whites was the ticket :-)

R

Eric, thanks for the info.

PS - No combinations of exposure / contrast that looked even half-way decent would have allowed me to accomplish my toning objectives for this photo (snow business), which were:

- Eliminate all blacks, shadows, and midtones.

- Ram the whites up against the wall, and shape the highlights exactly the way I wanted.

Save whites till last for fine tuning ya say? - Not me, but you can do whatever you want...

Try to do as much as possible using exposure & contrast, followed by highlights then shadows if necessary? That works sometimes...

Rob

Hi Hamish,

I took that picture while on vacation with some people who get very annoyed when I get all camera-buff on 'em.

Anyway, it was taken @0EV (exposure bias) and as you say could have easily been 2/3EV+. Also, I normally shoot with Active D-Lighting ON, which tries to help by lowering exposure even further - if thinking on my toes, I'd turn it OFF when shooting snow pics.

(also, that picture was taken shortly after sunrise, and I'm not a morning person . also, it was taken in the Eastern Sierra Madres (Mountains in California) in January (near zero temperature...) - keeping my hands warm while fiddling with the camera was a huge issue...  I have some more excuses too if you want... (PM me).

hamish niven wrote:

 

I see you are siting an example where you can't get the results you want from LR4...

Not at all!!! - I'm totally stoked about how that picture turned out! - Way better than what I was able to do in Lr3!! The fact that the image integrity held together so well despite the extreme processing is a testament to Lr4 improvments! Far more control over the nuances of highlight tone in Lr4 (via basic controls) than Lr3. And I think the auto highlight recovery also helped in this case (I'm not sure at the moment - maybe whites were jacked up so high there wasn't much auto-highlight recovery left (Eric would know) and maybe that's why they weren't flat, hmmm...). Please do not read anything in. My purpose in posting it was to point out that it benefited from the +white -exposure treatment, AND that this doesn't work for that photo:

- Adjust exposure so it's nicely exposed (or exposed for the midtones...).

- If necessary, adjust contrast so overall contrast is nice.

- If necessary, adjust highlights to taste.

- If necessary, adjust shadows to taste.

- Usually not necessary, but fine tune whites if warranted...

- Ditto for blacks.

Summary:

------------

Despite being underexposed to begin with, the final settings (which I am very, very pleased with) had reduced exposure. I would have never been able to come up with optimal settings for this photo by "following the rules".

Thanks - I realize asking Lightroom to make up for my shooting deficiencies is asking a lot .

PS - I understand why the controls stop where they do (at 100) - after that the image falls apart - no problem - could have pushed a little further using the tone curve (or an enhanced camera profile).. 100 was enough for this photo.

Yep, there have been a few "sub-plots" since this thread started....

2nd to Last thing - I arrived at the settings for this image very quickly - I knew what I wanted, I knew the secret, and I had the results optimized within a few minutes (no need to balance midtone and shadow considerations made this a very easy photo to work on - could hardly go wrong...).

Last thing:

hamish niven wrote:

 

I may be breaking your question...

There is no question - if you up exposure and down whites the highlights will be more compressed - it's a statement.

Rob

hamish niven wrote:

 

grand stuff Rob

As I said I've not pushed LR4 to extremes, but that is good to know that its holding quality with that much processing.

 

As for "following the rules" - that is a reason I left Englnd to come to South Africa, less rules, so I'm with you there.

I aim to make a photo look nice, real nice verging on the flipping awesome, if that means having a tone curve that would frighten Lee Jay, or as you've implemented, apparent crazy high settings and the results look good, you like it, friends and if professional, clients like it, then you've achieved, go pat yourself on the back and pour that whisky at end of day.

 

Rob Cole wrote:

 

There is no question - if you up exposure and down whites the highlights will be more compressed - it's a statement.

 

Rob

 

Mainstream commercial photography is ultimately an art, making things look good or better than they are, (like that amazing new product called fotoshop from adobé) start with a close approximation and manipulate, and LR is doing to do that much better.

For me, the technical details made by the brains incredible department make the software do 'kin amazing things, I dont need to know how knees and black curves and depressed highlights compacted with reverse engineered whimwhoms work. I don't care if my tone curves look less like a hysteresis curve and more like the curves of the back of the Lochness Monster, but as long as my edits work and the result is super.

 

Initially it did not matter if at 1st I did not get the results from slider 4 and 3 combined, I could fall back and try 5 and 3 with a little vibrance or tweak a curve, so many iterations and combinations of curves, sliders and numbers all interact with each other to wriggle that histogram and make it dance.

After some 600 images being run through LR4 Beta, its already not taking long to look at a pic and know,  - as you said with the snow photo, what I want to achieve from my image, and the processing is rapidly becoming intuitive, like playing the piano with new keys and different harmonics, its just a learning curve, and one that is a little like a giraffes neck -  steep, but not spotty.

 

 

As for annoying non camera buff people, I'm sure they have annoying non photographer habits.

 

I still get - "You take a long time to take a photo"

followed by

"Those are very good photos"

 

 

 

go figure

 

Message was edited by: hamish niven - typos

Well said .

R

Love the chiaroscuro drama! (you're from the north pole you say?).

How to maximize Love for Lightroom 4(b1), by Rob Cole

1. Adjust whites along with exposure.

     a. optimal ratio depends not only on how well highlights will like being spread out, but also midtone and shadow considerations (+whites -exposure expands highlights and left-shifts mids and lows...).

     b. set total "exposure" with that pair (initially) with the following adjustment of highlights in mind.

        ii. If you can see you're going to need to jack the highlights down, then factor in an estimated boost of total "exposure".

2. Set contrast lower than you think you'll want.

    a. highlights/shadows and the rest are easier to set when contrast is not too high, and will also affect contrast.

    b. finely tuned contrast is best done using tone curve anyway.

3. Adjust highlights / shadows & blacks...

5. Finish contrast adjustment using the tone curve.

    a. my tone curve adjustments are often more targeted at the bottom end, since Lr4 has more latitude in adjusting highlights via the basics, and the "help" afforded by auto-shadow recovery ranges from perfect to not-so-much.

    b. fine tune contrast to compensate for inability to perfect "type 1" toning issues (ref: previous page) using basics.

6. Add a touch of clarity maybe.

7. Final touches using locals.

    a. negative clarity for all the places where new clarity overbrightens or just doesn't look nice.

    b. increase contrast, etc... on edges to simulate the effects of old clarity, in the cases where the excessive edge darkening halo thing was actually a plus.

    c. intricate toning adjustments to brighten / de-brighten places where the local adaptive behavior has not been ideal (type 2 toning issues, ref: previous page).

    d. if you pushed a little hard with the +whites -exposure thing, spot reduce overboard highlights.

I'm still working on just how to integrate bottom end basics into all this - story still unfolding...

legend: you meaning me, and anybody else who may benefit...

Of course, my rules are made for breaking (or ignoring)...

Rob

Well, whites are similar to blacks, and highlights are similar to shadows (and contrast seems pretty symmetrical), BUT there is no equivalent to exposure on the dark side, which is why I normally use a tone curve to primp the left tones.

R

Jeff Schewe wrote:

 

Yes, you must deal with a reduction of overall levels when tweaking shadows, but there is a symmetry there if you recognize it...don't punt automatically and resort to curves (if you can avoid it).

I'll have to think about that - not sure what you mean just yet...

PS - I sing Lr4's virtues too...

So you're talking about shadows being the flip side of highlights? (symmetry-wise). - I've had that concept now for a long time... (I was referring to it in post above).

Eric,

MadManChan2000 wrote:

 

...entirely, if you go far enough.

I'd love to know more about that.

MadManChan2000 wrote:

 

What it doesn't counteract is the color recovery component.  We found in our testing that it was not photographically useful to have unpredictable color shifts due to highlight clipping & white balance.

A good thing, no doubt.

Whites is an awesome feature, and I've definitely been having big fun playing with high-key highlights and clipping, and also +whites -exposure for low-key but more separated highlights (also -whites +exposure for highlight compression). Way more flexibility than just an on/off switch for auto-highlight recovery (although that would have been nice too for some photos...).

Thanks - Lr4b is awesome, despite my personal challenges with it...

Rob

Hi Rose,

Sorry I wasn't more clear (pun not originally intended) - yes: definitely for brushing only - not gradients.

wdmn wrote:

 

Since a blur has been attained in the  past and to date using -clarity and -sharpness combo is it that the -sharpness can be eliminated and get a similar blur effect?

The effect of -sharpness hasn't changed (that I can tell so far) - it's lightroom's blur tool (making a distinction here between true blur and blur-like declarification).

To review (my apology if you already know all this):

---------------------------------------------------------------

+clarity makes dark tones darker and light tones lighter, and new clarity is bias more toward lightening, compared to old clarity.

-clarity does the opposite of +clarity.

+sharpness works kinda like clarity too except only on fine detail / edges.

-sharpness is not at all like the opposite of +sharpness.

-1 to -50 just masks global sharpener, and -51 to -100 is a true blur - it's a "double use" of the sharpening slider, one could even say "triple" depending on how one counts.

----------------------------------------------------------------

Anyway, I've mostly been using -clarity solo (or sometimes in conjunction with the others) for spot reducing overbrightness in the bokeh, which is already "blurred", but I have also used it a few times to smooth tonal gradients in subject matter. In the latter case you definitely need to be more careful to stay away from edges lest they become "declarified" too.

wdmn wrote:

 

In your post further back with the review and 'final touch using locales' there is the suggestion to use -clarity to offset the brightness created by clarity- I assume you are not getting a blur in this situation.

Local -clarity is added to global +clarity for net clarity effect. Thus if you have global at +10 and local at -10 (full flow and density) there will be 0 net clarity applied (no de-clarification).

However, if you have 0 global clarity and negative local clarity, then Lightroom will try to declarify, whatever that means - I'm not sure exactly, but if you are willing to bear with me... -

Clarity can be thought of as having 2 components:

1. "low-radius" detail component (more akin to sharpening).

2. "high-radius" big-area component (more akin to global contrast, but in a local adaptive / smart way...).

The "type-1" declarification is indeed akin to a "blur".

The "type-2" declarification is what I have been taking advantage of for debrightening the bokeh.

I have also "evened-out" tonal distribution in some subject surface areas using -clarity, in which case ya get the whole monty (both types). If you want to "re-clarify" the details, +sharpen and/or -nr.

The reason type-2 -clarity is so effective for debrightening is that it won't disturb tonal areas where there isn't much tonal diversity in the first place, but when applied to bright areas it pretty much understands you want them less bright (whether in a shadow a mid or a high), and if they're next to dark areas, it will lighten them just a smidge. Since the goal is generally to remove distracting over-brightness and strong tonal variation in the background - it's effect can be almost exactly what the doctor ordered. (but not everybody likes the same doctors ;-}

To compare, if debrightening using:

- negative exposure - it debrightens everything, so you have to be really careful.

- negative highlights - has no effect, unless it's a highlight, and then it affects them all the same (-clarity doesn't).

- negative shadows - has no effect, unless it's a shadow, and them affects them all the same (-clarity doesn't).

- negative contrast - this was what I used to use primarily for debrightening lights and lightening shadows, but it has no effect on midtones, and lacks the smarts of -clarity which can target just the right thing.

Clear as mud (pun intended)?

Rob

wdmn wrote:

 

Well, the mud is being diluted and requires a filtration system at the moment <grin>

Its a lot of left-brained stuff for most right-brained photographers to fathom. Not sayin' you're either, just sayin'...

wdmn wrote:

 

I can see why the previously used -100 clarity and -85 sharpness even when used with a low level flow is MUCH stronger. I had got so accustomed to that procedure (brush or gradient) previously.

It wasn't that I figured out that there was something radically different about clarity and its combinations I just did a 'knee-jerk' response and lowered the amounts in my favourite local brush. If your translation ends up being true then there will be some resaves of favourite brush presets.

Yes, many presets will need to be redone (both brush presets and reglar).

Another gotcha to watch out for - locals that included brightness will not translate exactly - this was also responsible for a few wrong conclusions in the early days (and still is sometimes).

wdmn wrote:

 

Who has time to investigate the book module, video etc?....I have those modules turned off because there is enough to absorb rewiring the brain for the develop changes.

Lightroom has a book module? (just joking)

wdmn wrote:

 

Have you noticed that if you go back and change an initial setting that it is VERY slow to refresh. Its also slow to display if testing out a setting and in that case the slider is continued to be moved because I think nothing has happened.

I have noticed some stuff like this occasionally, although I've not scrutinized it. - thanks.

Rob

Tip o' th' day:

==========

Intro:

------

Lr4b1's handling likes to maintain upper most tones as much as possible to preserve intra-highlight contrast and detail.

Probably more often than not, that's a good thing - one of the reason's Lr4b1's highlights can be so bright and still have detailing and color. But recent discovery indicates it can be problematic in the case when one channel extends very far rightward.

For example, if you take a picture of a white flower in sunlight, it could be that the blue channel extends way past the others to the right.

Or a picture of foliage in bright light could have the green channel extending way far rightward. Ditto for red.

This can make it hard to control the tone of the uppermost highlights.

So, the tip is:

----------------

Use the RGB curve to tone down the offending channel and the highlights will be much better behaved. (e.g. a downward bend at the very right end...).

Rob