Skip navigation

Automatic highlight recovery - a double-edged sword.

Feb 5, 2012 2:32 PM

  Latest reply: Rob Cole, Feb 27, 2012 6:07 AM
Replies 1 2 3 Previous Next
  • Currently Being Moderated
    Feb 7, 2012 10:34 PM   in reply to Rob Cole

    You can't do that with the tone/point curve.

     

    eh?

     

    R

     

    hey rob,,,

    editing,,, is not my native language,,, foremost a photographer,,older school of action,,,get it dam! near what is in my head and heart in the camera !! b the moment..:))  that's where my 95% of editing is i guess,, and i hope it always remains...so when i do lr/ps i'm hopefully not trying! to make a picture...enhance! you betcha!... but the last thing i want to do is dilute what comes out of the camera...still bracketing here ...i don't really want lr to whitewash anything...i just really want the really kool!! raw tweaks to the tonality/color/contrast that lr provides...so yea,, wb/curves/color/...now softproofing!!! f! wonderbar!!! that will be where i can see i'll be spending time,,,actually enhancing the prints!! not gawking at a histogram..

     

    den

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 1:03 AM   in reply to hamish niven

    VeloD and Hamish...

     

    I appreciate your participation and am sure you may have valid concerns.

     

    However, I have to call a "foul" for "piling on".

     

    The need for speed of processing and relative presets, valid or not, have not been a part of this thread.  If need be, start a separate thread here or in the requirements forum.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 1:11 AM   in reply to jrsforums

    Jrsforms - c'mon man

    The thread title is Auto highlight receovery - a double edged sword

     

    hamish niven wrote:

     

     

    ...processing of large volumes has to be done quickly. I'm waiting for relative development presets, so that I can Auto an image and then add/remove a little highlight, shadow, contrast etc

    Then we are getting places.

     

     

    I pointed out that the automatic highlight recovery working with adjustment presets would be a very useful and powerful way of using the auto highlight recovery, so most definately the 2 words "relative presets" are not off topic.

     

    so your foul is offside

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 1:14 AM   in reply to hamish niven

    Sorry...I disagree and see both as totally different discussions.  You may have an axe to grind, but don't lump it in here.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 1:36 AM   in reply to jrsforums

    Thanks for letting me know, your opinion is seen

    jrsforums wrote:

     

    Sorry...I disagree and see both as totally different discussions.  You may have an axe to grind, but don't lump it in here.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 2:32 AM   in reply to jrsforums

    You've got to be kidding. Put the flag back in your pocket.

     

    Read through the thread and let's talk about who's piling on who. Rob has been taking all kinds of abuse for trying to work through the development issues in his own way. Adobe may well be following the substance of these threads but they're not doing a particularly good job monitoring the tone. Far too much hostility and abrasiveness from a small group of contributors who've forgotten they're just like the rest of us. Victoria sets a good example others should follow.

     

    For the record I'm quite capable of deciding for myself what's relevant, and I'm sure Hamish is too. thanks.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 2:48 AM   in reply to VeloDramatic

    Let's see. Eric Chan and others have requested evidence form Rob in the form of files and they haven't appeared. We are all busy people who try and help the staff at Adobe to improve Lightroom for us the users. To wade through endless chatter about this and that with nothing concrtete to back it up and especially by someone who has blatantly promoted his own plugins that go against what has been recommended as best practise by Adobe, well is it little wonder that at times people get rather direct in their comments.

    By the way none of the Community Professionals get paid by Adobe, we do what we do because we really enjoy helping the team to make all our lives easier through product improvement.

     
    |
    Mark as:
  • Victoria Bampton
    5,302 posts
    Apr 1, 2008
    Currently Being Moderated
    Feb 8, 2012 3:31 AM   in reply to VeloDramatic

    VeloDramatic wrote:

    Victoria sets a good example others should follow.

     

    M, thank you.  We all try!  I've learned to turn the computer off when this forum is driving me mad...  

     

    As far as Adobe monitoring the tone goes, these are user-to-user forums, and moderation is very limited.  I'm sure it's a tough line to draw, between keeping a friendly atmosphere and being accused of limiting free speech or covering things up.

     

    Geoff makes a very good point.  It's great to be able to talk and work through the development issues, but we're talking about a very visual subject, so without photos, it's of limited value.  In fact, almost the opposite - if the engineering team are having to wade through chatter and figure out what people are trying to say, when they could just look at a picture and see immediately, it's taking away time that they could be spending actually improving the product.

     

    I'll be honest, when threads get this long, without pictures to back up the statements, I'm far more inclined to skim or even skip them.  I doubt I'm the only one.

     

    So in a slightly more light-hearted way, let me add to the pleas - http://www.autos.ca/forum/Smileys/CarTalk/ttiwwop.gif

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 9:36 AM   in reply to Rob Cole

    Rob Cole wrote:

     

    I really don't care if you or anybody else, who's mission seems to be to give me a bad time, knows the answer.

     

    Then why do you work so hard to persuade other people that you know the answer(s)?  If you genuinely do not care what anyone else (other than Eric) says, then stop arguing.

     

    If, on the other hand, you actually do care whether any of the other very knowledgeable people here might be able to help, then listen to what they have to say and do your best to work *with* them -- by, for example, providing some samples to illustrate your point.  It can just be a test shot, you don't have to share your "good stuff"...

     

    If you do neither of these things, then you are just costing time while adding little value for all the people who are here in earnest, especially those who are volunteering their time to help others.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 11:38 AM   in reply to Victoria Bampton

    hey victoria....  yes...back to the picxs....  

     

     

    Car Problems

     

    Image URL (for hotlinking/embedding): http://imgs.xkcd.com/comics/car_problems.png

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 12:41 PM   in reply to Victoria Bampton

    Victoria, nice try to convince Rob to come up with image examples to show his points. However, I am convinced he would not do so. It is not really understandable that he has not even one image, which would be suitable to post here, and which can support his case.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 2:10 PM   in reply to decoyle

    llllaughing my arrrrrse off!!!!!!!!!!!

     

    we can drive our 1dxxxxxxx's to work and shoot picts while driving.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 2:40 PM   in reply to Andrew Rodney

    I always loved that quote from Bruce...

     
    |
    Mark as:
  • TK2142
    407 posts
    Jan 20, 2010
    Currently Being Moderated
    Feb 8, 2012 4:44 PM   in reply to Jao vdL

    Jao vdL wrote:

     

    You SHOULD be able to use the white slider as a final adjustment.

    Are you sure?

     

    Admittedly, I haven't spend a lot of time with LR4 beta yet (I may stick to LR3 depending on whether or not LR 4 final will add better retouching support) but when I played with the "Whites" slider it appeared to stretch the whole histogram. I don't think that behaviour is a bug but intended.

     

    Is it the case that one can look at "Blacks" and "Whites" as moving the end points in a "Levels" tool, while "Exposure" moves them both in tandem (roughly)?

     

    I really wish there were a published explanation as to what the sliders are meant to do. This explanation can optionally appeal to intuition, but AFAIC, there should be one that also explains the effects in terms of low-level image manipulation terms. This would not only help to use the sliders as intended, but also to understand which "Basic control" sliders can be augmented/counteracted/complemented by tone curve controls.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 4:59 PM   in reply to TK2142

    The tone curve controls follow the Basic panel controls (in our processing pipeline).  Hence they are fully complementary to the Basic panel, and are intended to be used (if at all) after the Basic panel has been optimized. 

     

    Whites and Blacks are meant to determine how much of the image is mapped to pure white or pure black, without changing overall tonal relationships.  Similar to Levels in Ps, this means the entire histogram must be affected.  Unlike Levels in Ps, the implementation in Lr 4 Beta is more sophisticated in that it concentrates the effect more at the respective endpoints, again in an effort to preserve tonal relationships.  Furthermore, the implementation in Lr 4 Beta has some "image smarts" to make sure the sliders always have enough range & precision to map a significant portion of the image to pure white/black (if so desired), regardless of how much contrast the image had to begin with.  This is particularly effective for low-contrast images.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 5:13 PM   in reply to TK2142

    TK2142 wrote:

     

    Is it the case that one can look at "Blacks" and "Whites" as moving the end points in a "Levels" tool, while "Exposure" moves them both in tandem (roughly)?

     

    Yes, in concept but it's a lot more complicated...The Exposure in PV 2012 is like a smart brightness slider that doesn't really behave at all like the middle slider in PS levels. Plus Exposure in PV 2012 has a rolled off clipping in plus settings and a degree of he old highlight recover in minus settings. Blacks and Whites are really designed as gentle tweaks on really like the end point of levels.

     

    If you look at an image and evaluate what it needs, most of the time the heavy lifting is done with Exposure and Contrast–and if you don't use Contrast in PV 2012 then you're gonna have problems.

     

    Highlights and Shadows are further adjustments above or below the middle range where Exposure is the most aggressive. Whites and Blacks are really a fine tuning of the previous slider adjustments. If you have the Exposure and Contrast sliders set optimally, you prolly shouldn't be needing to make major moves with Highlights, Shadows or Blacks and Whites.

     

    And all of this is fundimentally different than PV 2003/2010. So, what you think you know needs to get adapted to the new controls. And that takes time and experience...but the results, not withstanding the title of this thread, are superior...

     

    LR 4.0 won't have much added from the public beta...some things may change in an update. Historically, the first dot release has actually gotten some new features primarily because Camera Raw gets locked in so early in the CS development. Thomas Knoll has been able to add things after the CS main release as long as it happens by the end of the quarter in which Photoshop ships. Anything Camera Raw gets in terms of features, LR also gets...so, who knows, maybe LR 4.1 will get some new stuff in it.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 5:14 PM   in reply to MadManChan2000

    Eric beat me to an answer...I guess I was too long winded!

     

    :~)

     
    |
    Mark as:
  • TK2142
    407 posts
    Jan 20, 2010
    Currently Being Moderated
    Feb 8, 2012 7:02 PM   in reply to MadManChan2000

    Eric, first of all, thank you very much for replying to my post. Much appreciated.

    MadManChan2000 wrote:

     

    The tone curve controls follow the Basic panel controls (in our processing pipeline).  Hence they are fully complementary to the Basic panel, and are intended to be used (if at all) after the Basic panel has been optimized. 

    Yes, but as the effects of the basic controls and the tone curve overlap, there is more than one way to skin a cat, so to speak, and my question was geared towards how to best distribute the image manipulation work between basic control and tone curve and whether or not some kind of adjustments were equivalent.

     

    For instance, changing the extreme end points in the tone curve is equivalent to moving the black and white sliders in a "Levels" tool. From what you are saying, I gather none of the basic control sliders can be mapped (at least roughly) to simple operations on the tone curve. I also believe this partly to do image masks being used rather than just being limited to shift contrast (as the tone curve basically forces one to do). However, the masks seem to

    use rather soft gradients, otherwise it wouldn't be necessary to affect the whole histogram if one wanted to just boost the whites, for example.

     
    |
    Mark as:
  • TK2142
    407 posts
    Jan 20, 2010
    Currently Being Moderated
    Feb 8, 2012 7:08 PM   in reply to Jeff Schewe

    Jeff Schewe wrote:

     

    If you look at an image and evaluate what it needs, most of the time the heavy lifting is done with Exposure and Contrast–and if you don't use Contrast in PV 2012 then you're gonna have problems.

     

    I have never used "Contrast" in LR3 for the following reason:

     

    The working point for contrast will usually not be optimal for any particular image I'm working on.

    I hence achieved the affect of a contrast boost by manipulating the tone curve. By doing that, I have a choice where the steep (vertical) part of the "S" curve is. When I use contrast, the steep part of the equivalent "S" curve is always in the middle, correct? I don't think that is optimal.

    If the contrast slider had a sub-slider which I could use to set the working point, I'd probably use it but then the tone curve seems to offer even more flexibility with no disadvantages.

     

    For the reason mentioned above (fixed working point) I would be uncomfortable to touch "Contrast" in LR4.

     

    Is some my thinking wrong?

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 7:20 PM   in reply to TK2142

    TK2142 wrote:

     

    For the reason mentioned above (fixed working point) I would be uncomfortable to touch "Contrast" in LR4.

     

    Is some my thinking wrong?

     

    Well, for PV 2012, yes I think you are wrong...Exposure and Contrast are now image adaptive which means that the image source has an impact on how the adjustments work on the image. Not using any contrast adjustment is leaving a lot of lifting to the curves–which as Eric indicated happens in the pipeline AFTER all the Basic panel adjustments touch the image.

     

    I would suggest using Exposure for the main tone control followed by Contrast to adjust the base level tone curve followed by Highlights to tweak above mid tone and Shadows for below mid tone. Whites and Blacks are the last leg to set clip. Highlights is also where the largest highlight recover comes in (although Exposure and Whites has some I "think"). Shadows has the same sort of functionality Fill Light had but it's been optimized to really reduce the halos Fill Light had.

     

    I've pretty much gotten away from using Parametric Curves now with PV 2012 and only bounce into the Point Curve for editing extreme highlight detail to tease more specific textural detail that Highlights and Whites can't really adjust. Plus, now we have Red, Green and Blue curves to adjust color casts. In many respects, the new Basic panel is almost like Parametric Curves on image adaptive steroids with highlight recovery and fill light thrown in...

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 9:35 PM   in reply to Jeff Schewe

    hey jeff...   so,, one can presume that the curves algorithm remained unchanged from lr3 ?

     

    tnx den

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 8, 2012 10:28 PM   in reply to decoyle

    decoyle wrote:

     

    hey jeff...   so,, one can presume that the curves algorithm remained unchanged from lr3 ?

     

    Well, they added Red, Green and Blue color curves, so no, LR 4 has more functionality. As for the parameteric curves, no, those weren't changed.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 1:30 AM   in reply to Rob Cole

    I love it. This is exactly how it should have always been. Nobody should have to worry about blown highlights or clipped shadows. This is precisely what software was made for. Maybe they could add checkboxes to unclamp highlights and shadows for people who like blown or crushed looks, but personally, I doubt I'd ever unclamp.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 5:40 AM   in reply to Mark Alan Thomas

    Mark, thanks for your feedback.  This is actually what we did with the +Whites direction.  As you push Whites to the right, it will "straighten out" the highlight shoulder to provide the digital "hard clip" and once you go past that point, it will continue to hard-clip more and more.  So, as you say, photographers who deliberately want blown/crushed highlights (e.g., for effect) have a means to get them, without going into curves.

     

    Rob, I disagree.  Our feedback so far from photographers is that they are using curves less now, not more.  (If overall people are using curves more in 2012, then we've failed ...)

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 5:49 AM   in reply to TK2142

    TK, when "boosting whites" it's a judgment call as to exactly which tones get "boosted."  For example, if you leave the shadows & midtones exactly where they are, and only raise the light tones closer to pure white, then there's a side effect: increased midtone-highlight contrast.  In terms of curves, that's like increasing the slope of the upper part of the curve ("stretching out" those tones), and that's what I mean by changing tonal relationships. 

     

    That's a side effect we didn't want, because the point of the first 4 controls (Exposure thru Shadows) was for the photographer to first establish the primary tonal relationships (overall brightness, and balance of highlights to midtones to shadows).  Once those are established, we didn't want Whites/Blacks to mess up that balance.  That's why these controls affect the whole tonal range (rather than a limited tonal range, like Highlights/Shadows).  Their purpose is to set the desired white/black endpoints, but their implementation involves moving the whole tonal range as needed to satisfy the desired constraints.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 6:24 AM   in reply to Rob Cole

    The reason I strongly don't recommend using Whites before Exposure (and the reason the controls are placed the way they are) is that it will shift the white point.  That is, if you've used Whites to already set the amount of clipping as desired, then adjusting Exposure will very likely shift the clipping level ... thereby invalidating what you did with Whites. 

     

    Another photographer recently asked about using the point curve before the Basic panel.  Also a bad idea, for the same reason.  If you've carefully crafted your point curve, then go to Basic and starting adjusting things like Exposure, etc., that generally invalidates all the work you've put into point curve (because the foundation on which that curve was based is now gone).

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:31 AM   in reply to Rob Cole

    I have actually just found the OP's issue in some of my own photos, in a very specific and very reproducable situation where it appears that this "auto recovery" could be a fairly major problem...

     

    In a studio invoronment in which you have a setup that is intended to be a pure white background and pure to nearly pure white reflective floor, but with the subject exposed normally, the LR4 version has parts not going white that should be, based on original in camera exposure.

     

    With the same image side by side, one with the 2010 process, one with the 2012 process, all other variables equaled as best I can figure (both at defaults with a linear tone curve) this issue is pretty obvious.

     

    When I adjust the "whites" slider to get those to match the 2010 version or at least to where I feel they should be, there is a noticeable and negative effect on midtones/skintones...

     

    As I have never posted (that I can remember) in these forums, before I post images as an example I'd like to ask what an appropriate format and size would be so that the modorators/engineers can best see this issue and hopefully try to do something about it? I would think that the issue might not be quite as obvious in a small websized jpg, but I also don't want to just post full resolution images...

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:32 AM   in reply to t.hall1982

    You could use YouSendIt.com to send an entire raw file to Eric Chan (manmanchan2000@yahoo.com). Eric is the Adobe engineer who is responsible for how the new controls work.

     

    https://www.yousendit.com

     

    It's free, but you have to sign up for their free service.

     

    Hal

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:37 AM   in reply to Hal P Anderson

    Ah, thank you; I use WeTransfer regularly; pretty much the same service... I'll try that.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:43 AM   in reply to t.hall1982

    t.hall1982 wrote:

     

    I have actually just found the OP's issue in some of my own photos, in a very specific and very reproducable situation where it appears that this "auto recovery" could be a fairly major problem...

     

    In a studio invoronment in which you have a setup that is intended to be a pure white background and pure to nearly pure white reflective floor, but with the subject exposed normally, the LR4 version has parts not going white that should be, based on original in camera exposure.

     

    With the same image side by side, one with the 2010 process, one with the 2012 process, all other variables equaled as best I can figure (both at defaults with a linear tone curve) this issue is pretty obvious.

     

    When I adjust the "whites" slider to get those to match the 2010 version or at least to where I feel they should be, there is a noticeable and negative effect on midtones/skintones...

     

    As I have never posted (that I can remember) in these forums, before I post images as an example I'd like to ask what an appropriate format and size would be so that the modorators/engineers can best see this issue and hopefully try to do something about it? I would think that the issue might not be quite as obvious in a small websized jpg, but I also don't want to just post full resolution images...

    Comparing RAW capbility and "in camera Exposure" JPEG will give you different views of dynamic range.  You kinda need to "calibrate" what you see on the histogram to guess at what the results will be in the RAW.  A good disertation on this is from Lee Varis at http://issuu.com/varis_photomedia/docs/digitalzonesystem-part-1?mode=e mbed&viewMode=presentation&layout=http%3A%2F%2Fskin.issuu.com%2Fv%2Fda rk%2Flayout.xml&showFlipBtn=true

    Part 2 is here: http://issuu.com/varis_photomedia/docs/digitalzonesystem-part-2?mode=e mbed&viewMode=presentation&layout=http%3A%2F%2Fskin.issuu.com%2Fv%2Fda rk%2Flayout.xml&showFlipBtn=true

     

    PV2010 and PV2012 process highlight recovery differently.  With 2010, you can get an initial view that shows highlight warnings (blown), which can still be recovered.  2012 will only show highlight warnings for truely blown (unrecoverable) highlights....i.e. 100 in one or more channels.  Everything below 100 will be slightly rolled off. 

     

    However, you can increase these highlights to a greater intensity as required with the other sliders. Check a number of MadManChan2000's posts for guidance.

     

    John

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:51 AM   in reply to t.hall1982

    that's madmanchan2000, not manman...

     

    Hal

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 8:56 AM   in reply to jrsforums

    jrsforums wrote:

     

     

    Comparing RAW capbility and "in camera Exposure" JPEG will give you different views of dynamic range.  You kinda need to "calibrate" what you see on the histogram to guess at what the results will be in the RAW.  A good disertation on this is from Lee Varis at http://issuu.com/varis_photomedia/docs/digitalzonesystem-part-1?mode=e mbed&viewMode=presentation&layout=http%3A%2F%2Fskin.issuu.com%2Fv%2Fd a rk%2Flayout.xml&showFlipBtn=true

    Part 2 is here: http://issuu.com/varis_photomedia/docs/digitalzonesystem-part-2?mode=e mbed&viewMode=presentation&layout=http%3A%2F%2Fskin.issuu.com%2Fv%2Fd a rk%2Flayout.xml&showFlipBtn=true

     

    PV2010 and PV2012 process highlight recovery differently.  With 2010, you can get an initial view that shows highlight warnings (blown), which can still be recovered.  2012 will only show highlight warnings for truely blown (unrecoverable) highlights....i.e. 100 in one or more channels.  Everything below 100 will be slightly rolled off. 

     

    However, you can increase these highlights to a greater intensity as required with the other sliders. Check a number of MadManChan2000's posts for guidance.

     

    John

     

    I'm not comparing RAW to JPEG, or using JPEG at all... and maybe I should clarify "in camera exposure" to "metered in studio". Obviously the different processes are treating rocvery differently, which is kind of the issue. In my images that I am referencing, the "auto recovery" seems to be pulling back a stop or more of the whites/highlites, which in some inscances might be nice and wanted; in these images it is not. When I try to correct for it, it has a negative effect on my actual subject; in some images it negatively affects the subject from the start.

     

    I will email Eric Chan as suggested so that he can see my particular issue, and hopefully help out.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 9:05 AM   in reply to t.hall1982

    I will email Eric Chan as suggested so that he can see my particular issue, and hopefully help out.

    Good idea. In the mean time some screenshots (compare the same image in PV 2010 at default to PV2012 at default as well as what happen when you lower whites in PV2012) would be nice. I think I know what you are talking about but would be nice to see it.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 1:49 PM   in reply to Hal P Anderson

    Thanks for the correction...

     

    but I believe it is  madmanchan2000@yahoo.com

     

    ....John

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 9:36 AM   in reply to t.hall1982

    t.hall1982 wrote:

     

    I'm not comparing RAW to JPEG, or using JPEG at all... and maybe I should clarify "in camera exposure" to "metered in studio". Obviously the different processes are treating rocvery differently, which is kind of the issue. In my images that I am referencing, the "auto recovery" seems to be pulling back a stop or more of the whites/highlites, which in some inscances might be nice and wanted; in these images it is not. When I try to correct for it, it has a negative effect on my actual subject; in some images it negatively affects the subject from the start.

     

    I will email Eric Chan as suggested so that he can see my particular issue, and hopefully help out.

     

    Whether metering "in studio" with a light meter or the cameras meter, you still need to "calibrate" to what you can expect in the RAW image....as processed by the raw converter.

     

    If you are comparing 2010 with 2012 in your statement of "...the "auto recovery" seems to be pulling back a stop or more of the whites/highlites...", this is only because 2010 would show revoreable highlights as blown, when this were not.  You can easily "blow" them back up with exposure, highlights, and/or whites, if you want.

     

    I am not saying you do not have a problem.  However, my experience is that you easily have 1+ stops of headroom in RAW, above what you would expect from the histogram.  Suggest you paruse Varis's article....if I am wrong, I apologize for my intrusion.

     

    John

     
    |
    Mark as:
  • Andrew Rodney
    1,388 posts
    Apr 16, 2009
    Currently Being Moderated
    Feb 9, 2012 9:38 AM   in reply to t.hall1982

    t.hall1982 wrote:

    I'm not comparing RAW to JPEG, or using JPEG at all... and maybe I should clarify "in camera exposure" to "metered in studio".

    Either way, exposing for raw is a bit different from exposing for the JPEG especially if you are using the camera meter without first nailing exposure for raw. This might help:

    http://www.digitalphotopro.com/technique/camera-technique/exposing-for -raw.html

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 9, 2012 9:57 AM   in reply to Jao vdL

    Jao vdL wrote:

    Good idea. In the mean time some screenshots (compare the same image in PV 2010 at default to PV2012 at default as well as what happen when you lower whites in PV2012) would be nice. I think I know what you are talking about but would be nice to see it.

    I'll give this a shot...

     

    Both examples the 2010 prcess version is on the left, 2012 process version on the right.

     

    Here both are set at defaults and linear tone curve...

    Screen shot 2012-02-09 at 12.18.11 PM.png

     

    and here the 2012 version on right I have adjusted the "whites" to match the 2010 version...

    Screen shot 2012-02-09 at 12.22.05 PM.png

    while it's easy to say "the one on right looks better than the one on left", there is a noticeable shift in the models skin tones by simply adjusting the whites...

     

    Another example, same scenario...

     

    before matching whites...

     

    Screen shot 2012-02-09 at 12.25.06 PM.png

    and after matching whites...

    Screen shot 2012-02-09 at 12.29.37 PM.png

     

    In the skateboard example, the "before adjusting whites" 2012 process version board is noticabley dull compared to the 2010 version (almost just like if I had used recovery in LR3, which I grew to despise and almost never use, becuse of this effect). After adjusting only the "whites", however, the tones of the baord have shifted considerably towards brighter highlights, well beyond what I would want.

     

    Hope this helps clarify what I'm talking about.

     

    Edited to add:

     

    What is also strange, is that in the example with the model, the "before" version of the 2012 process has slightly more, and in my opinion pleasing, contrast on the model, while having less blown highlights than the 2010 process; in the skateboard example, the 2012 "before" board has noticebly LESS contrast...

     
    |
    Mark as:
Actions

More Like This

  • Retrieving data ...

Bookmarked By (2)

Answers + Points = Status

  • 10 points awarded for Correct Answers
  • 5 points awarded for Helpful Answers
  • 10,000+ points
  • 1,001-10,000 points
  • 501-1,000 points
  • 5-500 points