Skip navigation
Currently Being Moderated

pv2012 first impression: boo!

Aug 1, 2012 2:54 AM

I spent a long time working out a default process preset for the 2010 process.

It had a lot of fill light and used the tone curve to give a rich black and put a bit of contrast back.  I cant sauy I ever had huge problems with artifacts.

Now I know pv2012 is supposed to be great for highlights, but no matter whether I start at the normal or overexposed braket I can get as much boost to the shadows in pv2012.  In the test image I'm working on I cant get the same separation between clouds and sky.  In fact in pv2012 when I expose correctly for the boosted shadows I have less detail in the highlights.

 

I know everyone is going to say I'm doing it wrong, But I've spent a couple of hours on this.  Even with some of the sliders at their max I still cant get the same effect (albeit a slightly stylised one).  Its as if with this process adobe have aimed for a "natural" feel.  Which is fine if thats what you want.  Personally I like that the fill light could be pushed until the image looked freaky.  Freaky can sometimes be good.

 

 

2010

2012-08-01_194415.jpg

 

2012:

2012-08-01_194445.jpg

 

And thats with highlights at -100 and blacks at +100m, shadows at +85

 

The dark area at the front of the hourse is darker, trying to tweak this with tone curves gives mud.

 
Replies
  • Currently Being Moderated
    Aug 1, 2012 6:50 AM   in reply to getho

    getho wrote:


    I know everyone is going to say I'm doing it wrong...

    Yep, I'm going to say you're wrong!  Or rather, I think you might not have given it enough time yet; a couple of hours isn't enough (I mean: to learn the new controls, not on one image).  I think it took me a couple of weeks to get results that were consistently better than I was getting with PV2010. 

     

    I'd say that in nearly all cases PV2012 can give results as good as and frequently better than PV2010.  It's much better at recovering detail in highlights and shadows without losing mid tone detail.  But the controls are markedly different in effect to those of PV2010, and take a bit of getting used to (as do the PV2010 controls, if you're starting out afresh). 

     

    Lots of posts here by Rob Cole and others sharing their experience of how to get the best out of PV2012. 

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 5:16 AM   in reply to getho

    One image doesn't show anything. No one ever said PV2012 was superior on every aspect of every image. If you have worked on many many images, and gone up the learning curve, and your conclusion is still the same, then I would say maybe you are onto something and your opinion would be more meaningful and well-supported. But everyone I have read says: once you get up the learning curve, PV2012 is superior, and this is my opinion as well. So, I look forward to hearing your opinion after you have used PV2012 for a while on a wide selection of images (not a few hours).

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 6:41 AM   in reply to getho

    You should be able to get the shadows up, and some highlight detail with PV2012, but the freaky look of large PV2010 fill values is not havable in

    PV2012, although PV2012 clarity will give you another freaky look you might like .

     

    Consider checking out this thread: http://forums.adobe.com/thread/972708?start=0

     

    If you want somebody to take a shot at your raw, please post with xmp that includes a snapshot of your favorite PV2010 settings - clearly indicated as such, or if you don't want to post, you can send via email - lossy DNG is ok.

     

    http://forums.adobe.com/people/areohbee

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 6:35 AM   in reply to getho

    I think you are trying to say that you like the slightly overdone tonemapped-HDR look and have found that impossible to achieve with PV2012 and wish the PV2012 sliders had more effect or went further.  Adobe tried to remove the gradient-reversal halos so maybe they did dial things back.  I use extra Clarity when going for the HDR-grunge look, but maybe that’s not quite what you’re trying to achieve with your example photo.

     
    |
    Mark as:
  • Andrew Rodney
    1,392 posts
    Apr 16, 2009
    Currently Being Moderated
    Aug 1, 2012 6:45 AM   in reply to getho

    First thing you need to do is learn to use the new tools!

     

    Start here with this freebee: http://mulita.com/blog/?p=3945

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 4:00 PM   in reply to getho

    getho,

     

    still waiting for that raw... - in the mean time, here are some things that might help:

     

    getho wrote:

     

    when I expose correctly for the boosted shadows I have less detail in the highlights.

     

    I have had the same problem with PV2012. The extra exposure needed to fill the bottom end, sometimes "overfills" the top end, and -highlights alone may be insufficient to remedy.

     

    With blacks=+100, and shadows=+100, if your shadows are still too dark, you probably do need a boost to exposure. If that over-brightens the highlights, and highlights=-100 isn't bringin' 'em down the way you want, then you probably also need:

     

    * -whites (instead of or along with -highlights), and/or

    * locals, e.g. gradients, with more exposure/contrast/highlight/shadow adjustments, and/or

    * -contrast (consider +clarity and maybe +vib/sat to go with that).

    * "U" shaped tone curve

     

    But note: It's very hard to assess a particular image without an example raw. PV2012 divvies up tonal ranges and defines slider characteristics on a per-image basis, so there is no universal answer.

     

    Also, extreme "fill" + highlight dampening often requires a tone curve - to bring down over-pumped shadows & mids, and bring out the detail of the highlights - consider a "U" shaped curve - which includes an upward inflection in the highlights.

     

    By the way, I agree that PV2012 is geared to image quality and accuracy, and the "special effect" aspect of PV2010 fill is a casualty - it's gone, gone, gone... You can get close to it by knowing how to work it, using locals, and heavy doses of clarity (which is very different than PV2010 clarity, for better and for worse), but you *can't* reproduce it exactly...

     

    Consider installing Cookmarks, then click on the link "Deep Fill, Ultimate" - you will have to tweak some things afterward, but should get you pointed in the right directon.

     

    I also occasionally do a "export back to catalog" thang which allows me to work on an one image using both PV2012 and PV2010 simultaneously (yes, it's klugy/klunky), for those cases when I want the PV2010 fill effect (or clarity), and still be able to use the other PV2012 editing tools too. That technique is somewhat automated via menu functions present in DevCorrect/2012, but will probably be redone and moved shortly to a new plugin to be named: "Edit In Lightroom".

     

    Cheers,

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 8:34 PM   in reply to getho

    The “splodge” is there in both of them, just less prominent in PV2010.  If you’re processing the DNG, then perhaps it’s an artifact of the lossy-compressed DNG and therefore original image detail, rather than an artifact added by LR processing.

     

    Can you post the original IMG_0470.CR2 file instead of a lossy DNG?

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 1, 2012 9:33 PM   in reply to getho

    getho wrote:

     

    http://pixelpics.com.au/stuff/test_0470.zip

    Are you hosting that file on a dialup? (~14K/sec download speed I'm getting). Sorry...

     

    Before (PV2010) & After (PV2012)

    (hint: to compare, keep hitting next button, instead of next-prev-next-prev...)

     

    Notes:

     

    PV2010 Settings:

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

    Fill 81

    Blacks 4

    Brightness 35

    Contrast 41

    Clarity 29

    Custom "S-shaped" tone curve

    No locals

     

    PV2012 Settings:

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

    Exposure +.8

    Contrast +25

    Highlights -80

    Shadows +80

    Whites +25

    Blacks -10

    Clarity +15

    No tone curve

    No locals

     

    Primary differences have to do with local contrast differences in process versions:

    * Fill vs. Shadows/Highlights

    * New Clarity vs. Old Clarity.

     

    Differences in highlights:

    * PV2010's are blown out, making for a more contrasty look when zoomed out, but void of detail when zoomed in.

    * PV2012's: full detail recovered, but look give a less contrasty sense when zoomed out.

     

    Differences in deep shadows:

    * PV2010's lack detail, making for a more contrasty look when zoomed out.

    * PV2012's have full detail, making for a nicer look when zoomed in.

     

    My opinion: image quality of PV2012 is better: more & better detail, more accurate tone/contrast. PV2010 version has wicked fill masking artifacts in some places, & blown highlights, & exagerated "what is wrong with this picture" contrast in some places, but has a certain charm due to fill-light character that I still miss sometimes.

     

     

    getho wrote:

     

    (thankfully 4.1 is a lot quicker than 3.6 for me )

    Now there's one we don't hear often .

     

     

    Cheers,

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 5:43 PM   in reply to getho

    getho wrote:

     

    Hey adobe can you make shadows and highlihgt go up to 150 please

    Work-around: make a preset with a local gradient which covers the whole image in uniform fashion, with highlights -50 & shadows +50.

     

    Note: additional oomph provided by h/s locals declines once global values are maxed out.

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 2:20 AM   in reply to getho

    getho wrote:

     

    cant see those images? (link wont open)

    Right-click and choose "open in new tab". Not sure why you have to do that. Ctrl-click on Mac I presume.

     

    I think you are attached to the look of PV2010/fill. I used to be too. But I've mostly gotten over it after a few months with PV2012. It's a psychological thing. Sometimes I'll leave a PV2012 image on screen that I didn't think I liked as well as the PV2010 version, then after a few hours come back and think: gee - that's a really nice picture, then switch to the PV2010 version and not like it as well. In other words it's partly about characteristics that are familiar and liked... For example, the grass and sky are brighter in the PV2012 version, but they should be - it was a sunny day and they were no doubt bright... So, it looks "right" after a while, but if switching back and forth with a predisposition for the 2010 version, it may not seem desirable... Not saying realism is necessarily the goal, but Lr4 is more true to life - one can exaggerate some things, but it'll never look exactly like what you've come to love about PV2010 fill. Blah blah blah...

     

    Note: PV2012 version had these changes too:

    Vibrance: 10

    Saturation 5

     

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 5:49 AM   in reply to getho

    getho wrote:

     

    AH right, yes your setting were a good match for the house - just the sky left looking a bit different.

     

    I hear what your saying - I just processed the rest of the bunch and some of the interiors benefitted from the new process. I liked how 2010 would give you areas of very flat tonality - I'm a big believer in using the limits of technology to define an aesthetic - so when film was grainy I used to just go with the grain.  When lightroom was all flat tonality and filltastic I went with that, so I guess I should just roll with pv2012, and see what happens. (Still feels a bit neat for my style though...)

     

    I got ya too - I went back and tried to match the sky without using locals (and without screwing up the tone elsewhere) - very challenging. I went ahead and put another PV2012 version up which looks more like your original PV2010 version, but it took some locals in the sky to do it. There are some aspects of PV2012 that are indeed hard to control, which is a complaint I have as well. I use the tone curve less in PV2012, because there is more flexibility to stretch and/or compress tones using the basics, but I use the locals more in PV2012 to even out "exposure" and such in cases where PV2012 and I aren't on the same page... - shadows and highlights in the local mix make that easier, and clarity attenuation is a common component, since sometimes I like the clarity at full strength in most, but not all places in a photo.

     

    Cheers,

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 9:13 AM   in reply to getho

    In the spirit of assisting, here are my observations with suggestions that have not been mentioned (ref. test_0479.dng file). Most of what follows pertains to both PV2010 and PV2012.

     

    1. I noticed you are using the Adobe Standard profile, but have made changes to the RGB primaries. On my hardware calibrated monitor the image appears to have a color cast, which does not appear natural.
    2. You have Grain set to Amount & Size = 2.
    3. Noise Reduction and Sharpening values are set much higher than optimal. I have a Canon 600D, which uses the same sensor and electronics as your 60D. The overall appearance at 100% view is like an oil painting with all fine image detail flattened. The Canon 60D and 10-22mm are capable of producing extremely high IQ, which is completely lost with these LR ‘Detail’ panel settings.
    4. The HSL Saturation sliders have some moderate adjustment settings with the goal to darken the sky. On my calibrated monitor the sky color does not appear natural. I realize this is subjective.
    5. Vibrance and Saturation have small opposing settings (-5, +6), which has virtually no effect on the image.

    Suggestions (for the above):

    1. If you would like to obtain more accurate color matching I highly recommend the X-rite ColorChecker Passport, which enables creating custom DNG profiles for specific lighting conditions. I use a CC PP with my Canon 300D, 600D, and 5D MKII cameras and the results are superior to the Adobe and Camera profiles proved inside LR. You can also use the CC PP target real-time on a shoot with difficult lighting conditions (i.e. mixed Natural & Tungsten).
    2. Probably unintentional.
    3. Adjust sliders in the Detail panel at 100% view. Start with the default Sharpening (25, 1.0). Next do the Color slider starting at ‘0” by observing the color noise in a shadow area. Slowly increase the Amount until all (or most) color speckles are removed. Then adjust the Luminance slider by observing a noisy midtone area. Increase it slowly until the noise is reduced, but still very slightly visible. REASON: You are viewing at 100%. With your 18Mp 60D and a large monitor (20”+) this is equivalent to viewing a 40” x 60” print at close range. Now switch to 1:2 view. If there is no obtrusive noise visible you are done. For your ‘test_0479.dng’ image I used Sharpening Amount 35, Radius 1.0, Detail 35, and Masking 0. For Noise Reduction, Luminance 25, Detail 50, Masking 0, and Color 25, with Detail 50.
    4. Using a ColorChecker PassPort (or similar) to create custom DNG profiles will help to keep all colors more accurate.
    5. With My Canon bodies I set Vibrance = 25 as part of my ‘default’ Develop settings. Your preference may be different, but it seems to work well with Canon raw files and ColorChecker PP custom DNG profiles.
    6. One other suggestion. I noticed you used an aperture setting of F18 with your 10-22mm lens. There is noticeable softening in the image due to diffraction from the small aperture setting. Most lenses perform best at two to three F stops down. I generally shoot at F8 and only stop down further under extreme depth of field situations (i.e. very close foreground subject). See this site to check your Canon Lens’ stop down performance:

    Digital SLR and Lens Image Quality Comparison - ISO 12233 Chart 100% Crops

    Here’s my PV2012 rendition of your test_0479.dng file with the above Detail panel settings. The other Develop settings are very similar to Rob Cole’s, but I changed WB to 5167 and 10.

    Double-Click on the image to see full-size:

       Original                                              My LRSettings

    PV2012.jpg

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 1:33 PM   in reply to getho

    getho wrote:

     

    Thanks for taking the time and trouble to look at this...

    You're welcome.

     

     

    getho wrote:

     

    I used your settings as a basis for a few 2012 presets and tried applying them to the next house. Results were much more mixed than with 2010. Some looked ok, some looked terrrrible. Whereas my 2010 preset would work on nearly everything - id just have to dial back the fill. And tweak exposure and contrast.

     

    Unfortunately, PV2012 does not lend itself as well to basic toning presets. Why? Because the tonal ranges and behavior of the sliders are not the same from image to image. For example, shadows slider on one image may affect only a small range of tones, whereas on the next image it affects a bigger range. Ditto for highlights slider. Fill light does the exact same thing on every image you use it - ditto for the rest of PV2010 sliders.

     

    My opinion of PV2012, in a nutshell:

     

    Pros:

    -------

    * Potential for improved image quality and great tone.

    * Max detail.

     

    Cons:

    -------

    * A big goof as far as predictability and learnability, and preset-friendlyness.

    * Limited ability to control it's characteristic look (same is true of PV2010, PV2012 just has some different characteristics...).

    * Max detail.

     

    Yes, "max detail" is on both lists - it's great when you want it (luckily for me, that's most of the time), but distracting in some photos, and difficult to suppress. For example, PV2012 tends to reveal all the gory deep shadow detail (read: auto shadow-recovery), yet sometimes I wish the shadows/blacks were just dark so-as not to draw attention. Likewise for highlights (read: auto highlight-recovery).  You really have to work for that in PV2012.

     

     

    I don't use absolute presets much for basic toning (although I do use relative presets sometimes, via Cookmarks plugin), but if I did, I think I would leave exposure out of them completely (maybe blacks too - discIaimer: I haven't given this much thought) - always set exposure (and blacks?) manually, then use the presets to vary amount of punch and other biases...

     

    Good luck with it.

     

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 2, 2012 9:21 PM   in reply to trshaner

    It was a little too "aqua-y" for my tastes as well. I assumed it was just the look getho was going for, as opposed to "correct white balance...", i.e. cyany skies instead of purply blue... Otherwise, I would guess getho's monitor is off. Prompted me to reset everything then process anew. Here's a version edited lickity split starting with reset-all, followed by auto-tone, followed by a few other adjustments. I did not look at the other versions, because I was curious what I would come up with sans influence of previous renditions, as much as possible:

     

    Natch from Scratch (hint: you may need to right-click (windows), or ctrl-click (mac) to open in new tab or window).

     

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 4, 2012 1:17 AM   in reply to getho

    Fair enough getho - sounds like a pretty good plan. Do keep us posted .

     

    getho wrote:

     

    In the develop module it takes 3-4 seconds when you move from one picture to the next before you can work on it - even with 1:1 previews rendered). Oy!

     

    Yeah, unfortunately rendering 1:1 previews only creates lib previews and a nearly worthless bit for the ACR cache to help develop speed. Anyway, you can start working on image as soon as the sliders are enabled, which for me is about a second, then it takes another 2-3 seconds for loading to complete (as evidenced by the loading indicator being extinquished). I finally turned off the loading indicator and start editing before loading is complete. Are you saying it takes 3-4 seconds for sliders to be enabled, or loading to complete?

     

    R

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 5, 2012 4:37 PM   in reply to getho

    Ah. Yeah that's a little long. I don't understand why it's using swap memory either. I thought it was only supposed to be used when no more ram. Hmmm.........

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 4, 2012 8:42 AM   in reply to getho

    Ghetto, my intent was to provide feedback that might be useful to you. I even stated, "I realize this is subjective," concerning the HSL slider settings. Since I know very little about your experience with photography it's understandable that you may feel I was being condescending. That was not my intent at all, but merely suggestions given "In the spirit of assisting you."

    getho wrote:

     

    For better or worse, my bread and butter is all about quality/speed. Hence the initial point of this post...

    I agree with your statement, "I'm rather in the school of "there's no right way to do it". I am also in the school of "quality and speed," and always looking for ways to save time processing my images in LR. I provided suggestions towards that objective only after carefully reviewing your test_0470.dng image. It appears you have experience with both film and digital photography (as do I), so I appreciate your viewpoint and feedback.

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 4, 2012 9:11 AM   in reply to getho

    getho wrote:

     

    2-3 seoncds before sliders become available, 4-5 seconds before loading indicator goes off.  when I click on an image I can here my swap disk scratching away.  (why its using swap memory , though I dont know - I've got a i920 @3.6Ghz, 12gb ram).

    I'm running  LR4.1 on a very similar performance system (Windows 7, i7-860, 12GB RAM, 1TB 7,200 rpm drive, 1920 x 1080 display) with no issues. What OS, drive(s), display(s) resolution, and Camera Raw Cache size are you using?

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 4, 2012 10:18 AM   in reply to trshaner

    trshaner wrote:

     

    getho wrote:

     

    2-3 seoncds before sliders become available, 4-5 seconds before loading indicator goes off.  when I click on an image I can here my swap disk scratching away.  (why its using swap memory , though I dont know - I've got a i920 @3.6Ghz, 12gb ram).

    I'm running  LR4.1 on a very similar performance system (Windows 7, i7-860, 12GB RAM, 1TB 7,200 rpm drive, 1920 x 1080 display) with no issues. What OS, drive(s), display(s) resolution, and Camera Raw Cache size are you using?

    I'm also using a similar system (i7-930, 12G RAM, 1920x1200 display - but I've a slowish SSD for C: with the ACR cache on it, but a slowish 1TB hard drive for catalogue and photos).  It can be 10 seconds to get into Develop module the first time, but then going from picture-to-picture in Develop, or going back to Develop, it's about 2-3 seconds for the loading indicator to go off (with 12M pixel images).  That's just slightly slower than with LR3.6. 

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 4, 2012 11:22 AM   in reply to CSS Simon

    My LR4.1 performance with Canon 5D MKII 21.1 Mp raw image files.

     

    Develop Module

    I'm seeing 4 sec. to get into the Develop module the first time, and then about 1.0 sec when switching between Library and Develop modules. The Develop module 'Loading' indicator takes 3 sec. when switching between images, but  the sliders after available after only 1.0 sec. and I can see their effect onscreen real-time. There is no slider delay or sluggishness after the initial 1.0 sec. delay when switching between images.

     

    Library Module

    Everything in the Library module is like greased lightning. The 'Loading' indicator appears for about 4 sec. in the Library module when going to 1:1 view, but only after making changes in the Develop module. After that initial Library preview rebuild it's back to greased lightning. Switching between images is instantaneous (less than .25 sec.).

     

    Even when using the Spot and Brush tools the performance remains the same. The only thing different I did with my 1 TB single drive system was to create a separate 200GB partition (J:) for the Camera Raw Cache, Lightroom Catalog and Previews. All applications, OS and image files are on the primary C: drive sector. I have no idea if this helps LR4's performance. But it does seem to reduce drive fragmentation. which would slow down the system.

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 5, 2012 5:12 PM   in reply to trshaner

    trshaner wrote:

     

    ... but the sliders are available after only 1.0 sec.

    Given my limited experience, this seems to be about normal, and is close to what is normal for Lr3.6 as well, maybe it's a tad longer - dunno. In any case, 2 and 3 seconds for sliders to be enabled on a capable computer (multiple cores with enough ram and a healthy hard drive...) means you are one of the many being bitten by one of the many performance bugs in Lr4.1.

     

    I seriously doubt doing things like defragmenting and partioning drives, or even buying additional/faster drives for further distributing/optimizing access will help much. Maybe a little, but not a lot. I mean the time it takes to fetch the raw file, and/or a preview from disk, or an ACR cache entry..., should be only a small fraction of a second, and the amount of catalog data that needs to be fetched in order to prepare develop is miniscule (and accessed using random-access - fast). Granted, there are dozens of other little files that need to be accessed as well, but still all of this should be done in a fraction of a second, the whole second delays are Lightroom wandering around lost with its ________ up its _________, if you know what I mean (or waiting on some other confused piece of dependent software...).

     

    Rob

     
    |
    Mark as:
  • Currently Being Moderated
    Aug 6, 2012 8:06 AM   in reply to getho

    getho wrote:

     

     

    I've got a 3 monitor setup main is 2560x1440, 2nd is 1920x1200 running off 2 nvidia cards.

    win7 x64, ssd boot drive

     

    There have been many reports of "slow" develop sliders on dual display systems, and on single high resolution display systems (i.e. 2560x1440). Systems with higher performance processors, graphics cards, SSDs and more memory are not immune to this slow down.   The only real solution that I am aware of is to make the Loupe image smaller, regardless of whether it's on a single or dual display. I tested this on my single 1920x1080 display by opening a 'Second Window' using the '2' button in lower lefthand corner. I then made the 2nd window as large as possible (1000x1500). This is only about 36% larger than my normal Loupe size (733x1100). Develop slider response went from virtually instantaneous to about a .5 sec. delay for the primary Loupe (733x1100) and 1.0 sec. delay for the 2nd window Loupe (1000x1500). At a 1920x1200 2nd Display Loupe window size the affect on the Develop sliders is going to be much worse. With this "dual window" setup the Library module remained very responsive.

     

    See my suggestions in this post:

     

    http://forums.adobe.com/thread/1044035

     
    |
    Mark as:

More Like This

  • Retrieving data ...

Bookmarked By (1)

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