Post Processing Test

[FrankMi]

I have been trying to learn PD CS5 layers and masking. For a test I am using an image I took in 2004 with a Sony Cyber-Shot DSC 42. I shot most of the images back then at a whopping 2 MP. I wanted to not only learn what I could do with PP, but also to determine, at the screen ratios we use in CiC, how much of a degredation I would see when compared to my DSLR and images posted by others using much better cameras.

I am hoping to get feedback in two areas. First, what things I can do in PP to improve the image I am using in the test and two, what are the biggest differences we can see in the images that get posted at these restricted sizes compared to today's DSLRs.

Here is the original image SOTC with only size reduction and output sharpening applied. The original was 1600 x 1200 pixels.



And here is the same image after PP using primarily levels and masking.



Any other C&C feedback is, of corse, welcome as well! Thank you for taking to time to view the images and provide comment.

It might be helpful to view in the lightbox.

[MiniChris]

I've noticed as much in my own works as in other's work, there is a tendency to over-saturate or blow out of details in whites and lights. The lady's white hat, the man's blouse and the tablecoth in the PP image are like that. Try taking the PP image and overlaying it with the original, then doing some light painting to get a nice combination of the two value ranges. Later, you can add a slight curve, color balance and hue/sat balance adjustments if needed to pop the colors a bit, without losing the detail.

[FrankMi]

I've noticed as much in my own works as in other's work, there is a tendency to over-saturate or blow out of details in whites and lights. The lady's white hat, the man's blouse and the tablecoth in the PP image are like that. Try taking the PP image and overlaying it with the original, then doing some light painting to get a nice combination of the two value ranges. Later, you can add a slight curve, color balance and hue/sat balance adjustments if needed to pop the colors a bit, without losing the detail.

Thanks Chris, for the suggestions. Here it is with the changes. Did I go far enough? I'm a little reluctant to go too far until I get a good sense of what works and what doesn't.

Any other suggestions?

[MiniChris]

I do think it works better, but no amout of PP will ever recover blown out whites...there is one more technique you might try on her cap and on the endcap of the tablecloth which involves placing a clear layer mask over the image and "painting" in a low value of blue (knock down the opacity to about 6-8%). It does require great patience and experimentation with both color, flow and opacity.

[FrankMi]

That sounds intriguing, Chris. When you say a 'low value of blue', what might the RGB values be for that for a starting point? By 'flow' do you mean that you would you paint just the places where the white has lost detail? Thanks!

[MiniChris]

just where the white has lost detail...but while an overpaint won't bring back lost detail, it will help to soften the blownout look. I generally start with using the eyedropper tool to find the best color (a warm color for a warm blowout, cool for one like yours), then using an opacity of 8-10% and flow value of 30-40%, very soft brush (unless there is a hard edge involved) and slowly build up the color. You can change the opacity of the brush, or the overall opacity in the layer.

This is a technique designed to soften a hard light, not to replace a lost detail. Nothing replaces lost detail due to a blowout of excessive light/exposure. I only use this when I want to keep an irreplacable photo op (like something with the grandkid). On raare occasions, I have used it to build up an edge or provide some contrast, but always in very small areas and low opacity applications. You can also try some blend modes with the mask that will help to improve the contrast value.

[FrankMi]

Thanks Chris, for the expanded explanation. I'll give it a try!