Need your opinion on Technique

anonymouscubananonymouscuban Registered Users, Retired Mod Posts: 4,586 Major grins
edited April 11, 2009 in Finishing School
Not sure if any of you have been in the same boat as I am right now. I have quite a few digital-darkroom tools at my disposal but have never really taken any formal courses or siminars on using them. I am also relatively new to the whole digital scene.

When I do post on my images, I make the adjustments based on what looks best to my own eyes. However, I question whether I know what just looks good to me but may be over processed to others. I'm hoping you guys can help me out a bit with a critique of this image, just from a post perspective, and give me direction on what I can do to improve.

Here is the image straight out of the camera. All I did was convert it from RAW to JPG.

505137809_TvWu2-L-1.jpg

Here is the image after post:

505139484_PvXiG-L-1.jpg

Here are the steps I took in post, all done in LR2 & CS3:
  1. I adjusted the WB in LR because it seemed a bit yellow to me
  2. Then I correct the exposure since it was a been underexposed, again in LR
  3. I then drop the image into CS3 and adjusted the curves to increase contrast... seemed a bit on the flat side
  4. I created a dup layer and selected the dog using the Quick Selection tool, then shapened only the dog to give more detail in the fur
  5. I then created another layer and inverted the selection and blurred the background using a Gaussian Blur. I set the opacity of the blurred background to around 80% and then cloned a bit of halo around the dog that was created when I added the blur. I blurred the background because I found it a bit distracting after I corrected the exposure, especially the white moulding.
That's it. I would really appreciate your feedback. It looks good to me so maybe I nailed it. But again, I'm wanting some feedback because I am not sure if I am just overprocessing because I have the tools. :scratch

Thanks guys.
"I'm not yelling. I'm Cuban. That's how we talk."

Moderator of the People and Go Figure forums

My Smug Site

Comments

  • JustPlainMeJustPlainMe Registered Users Posts: 190 Major grins
    edited April 4, 2009
    I'm in the opposite situation from you; I've been doing photo correction and scrapbooking for a while, but only recently have I started paying attention to my photos.

    I think what you did looks pretty good; I like the background change behind the dog. I think, though, you could also have done an levels adjustment layer mask around the darkest part of his face (snout, eyes) and maybe lightened him a little, given it a tiny bit of fake fill-flash, just to bring out some of the detail. There is a lot of contrast between his head and snout, and I think the head is possibly overexposed next to the darkness of the nose.

    But that's just my two cents!

    You ask a great question; photos that look okay on my monitor, to my eye, sometimes are not when I print them or display them on another monitor. Sometimes I like an effect, but I walk away from it and come back and realize it was way overdone. (And this was before my RAW days, I've messed up many a photo getting a little too overzealous with blending modes!) So much, though, is in the interpretation of the photographer/processer; what do you want to see? Exaggerated contrast, details, colors? You can do whatever you want with it. When it looks just right to you, that's probably what matters in the long run.

    Sarah
    Please ignore my opinions! And if I ask for constructive criticism, please give it to me. I have really thick skin! :huh
  • TonyCooperTonyCooper Registered Users Posts: 2,276 Major grins
    edited April 5, 2009
    To effectively critique this we'd have to know how accurate the image is. Is that the color of the dog's fur? Is that the color of the dog's scarf?

    The image looks over-processed to me. If the dog's fur is really that light in color, then it's less over-processed than I'd guess. If the scarf is really that bright a blue, then it's less over-processed than I'd guess.

    The main problem with the image is in the lower right where its difficult to see where the dog ends and the background begins. The corrected image doesn't really improve this.

    The idea of the alterations you've done in Photoshop is to correct the image to agree with what the subject actually looks like. Only you know if you've accomplished this.
    Tony Cooper - Orlando, Florida
    http://tonycooper.smugmug.com/
  • anonymouscubananonymouscuban Registered Users, Retired Mod Posts: 4,586 Major grins
    edited April 5, 2009
    I'm in the opposite situation from you; I've been doing photo correction and scrapbooking for a while, but only recently have I started paying attention to my photos.

    I think what you did looks pretty good; I like the background change behind the dog. I think, though, you could also have done an levels adjustment layer mask around the darkest part of his face (snout, eyes) and maybe lightened him a little, given it a tiny bit of fake fill-flash, just to bring out some of the detail. There is a lot of contrast between his head and snout, and I think the head is possibly overexposed next to the darkness of the nose.

    But that's just my two cents!

    You ask a great question; photos that look okay on my monitor, to my eye, sometimes are not when I print them or display them on another monitor. Sometimes I like an effect, but I walk away from it and come back and realize it was way overdone. (And this was before my RAW days, I've messed up many a photo getting a little too overzealous with blending modes!) So much, though, is in the interpretation of the photographer/processer; what do you want to see? Exaggerated contrast, details, colors? You can do whatever you want with it. When it looks just right to you, that's probably what matters in the long run.

    Sarah

    Sarah,

    Thanks for the feedback. I tried to pull out as much detail in his mask; I will work on the levels to see if I can get more. You made a very good point in that images can look very different on different monitors. I have three monitors at work and two at home. My monitors to home are calibrated, the ones at work are not. I do my post at home and then sometimes view them at work just to see the difference. It is dramatic sometimes, which mucks things up a bit but I go with what I see at home on the calibrated monitors.
    "I'm not yelling. I'm Cuban. That's how we talk."

    Moderator of the People and Go Figure forums

    My Smug Site
  • anonymouscubananonymouscuban Registered Users, Retired Mod Posts: 4,586 Major grins
    edited April 5, 2009
    TonyCooper wrote:
    To effectively critique this we'd have to know how accurate the image is. Is that the color of the dog's fur? Is that the color of the dog's scarf?

    The image looks over-processed to me. If the dog's fur is really that light in color, then it's less over-processed than I'd guess. If the scarf is really that bright a blue, then it's less over-processed than I'd guess.

    The main problem with the image is in the lower right where its difficult to see where the dog ends and the background begins. The corrected image doesn't really improve this.

    The idea of the alterations you've done in Photoshop is to correct the image to agree with what the subject actually looks like. Only you know if you've accomplished this.

    Tony,

    Thanks for taking the time to post. I completely understand your ponit. Without seeing my dog IRL, then you can't know how accurate the post work is. Very obvious now, but not when I posted.

    I think the colors of the fur and scarf are pretty accurate to what they are IRL. The scarf has since been thrown away but the dog, Meatball, is sitting right here with me while I type this and I can see that his fur is very light.

    Again, the problem with snout detail being lost in the background may be an issue of monitors. I can see very distinctly where the the lower snout ends and the background begins on my monitor in the processed image. Not so much in the original. But then again, who's monitor is more correct???
    "I'm not yelling. I'm Cuban. That's how we talk."

    Moderator of the People and Go Figure forums

    My Smug Site
  • ruttrutt Registered Users Posts: 6,511 Major grins
    edited April 11, 2009
    Two observations:
    1. Take your foot off the gas. Just because you can take it as far as you did, doesn't mean you should. Save an original version (perhaps with Image->Duplicate) before you start and blend it back in as the final step to soften whatever you did in case you went overboard. Often a more subtle edit works better than your full on "best" effort.
    2. When you increase contrast as much as you have done here, you have to make the colors more vivid somehow. I think the easiest way to do this is to work in LAB and steepen the A+B curves after you adjust the L curve. You can find out about this technique here and even more here (especially the first chapter.)

    510125368_xCvxf-O.jpg
    If not now, when?
  • anonymouscubananonymouscuban Registered Users, Retired Mod Posts: 4,586 Major grins
    edited April 11, 2009
    Thanks for the feedback Rutt. I found the LAB technique tutorial earlier this week. I've read through it a couple times but have yet to actual walk through it with an image. I plan to though.

    I will also try your suggestion of blending in the original image to regain some of the color and to just step it back a bit.

    This is really helpful because I had a feeling I was push the PP a bit too much.

    Thanks,
    Alex
    "I'm not yelling. I'm Cuban. That's how we talk."

    Moderator of the People and Go Figure forums

    My Smug Site
Sign In or Register to comment.