Happy Mother’s Day to my mom (pictured above), who doesn’t get on the internet much and probably won’t see this anyway 🙂 I’m not one for lots of sappy words on the internet so I’ll just say that I love her and I appreciate her greatly.
I recently posted our Thanksgiving Day family portrait and today wanted to show how I modified it. The only direction I gave to the family for the picture was “wear something solid-ish on top, and something denim on the bottom”. As you see in the picture at this link, we all ended up in rather muted colors except my youngest son who had a bright yellow shirt on. I was busy thinking about how to add fill to the shot, position us reasonably without taking all day to do it, etc. (I should have spent a bit more time on the positioning). So, when the bright yellow shirt was pointed out to me I thought to myself, “Whatever…it won’t matter”. Of course, when editing the photos it bugged me to death and I wished I had changed it.
The solution? Photoshop’s “Replace Color” adjustment. I used the tutorial linked below as a starting point to learn about it and experimented from there. Other than choosing the new color, the key setting for me was the “fuzziness”. This determines how aggressive the automatic selection is. What I found is that because of the variation in saturation throughout the shirt I had to slide the fuzziness way up which causes other parts of the image to also be selected (trying to automatically select the shirt’s colors reveals how much variation is really there). I thought the checkbox for ‘Localized Color Clusters” (not shown in the tutorial but exists in CS5 at least) would help minimize the selection but I didn’t see a lot of difference once the the fuzziness was increased much. I also used the +/- eye droppers to add/subtract from the selection. Finally I needed a bit of manual masking to only change the shirt and not other areas of the image. Something which is more solid in color would far easier to use this tool with. The resulting photo is above — a 5-minute edit. I will probably do another version and use a color picked from someone else’s shirt so that it matches even better with the rest of us. When I look at the new image I kind of think it doesn’t look right because *I* know that I made the edit but in my brief survey of people who didn’t know about it, not a single person noticed anything.
Here’s the link to the tutorial I started with.
We *tried* to take some portraits of my wife and daughter but not everyone was cooperating. Eden was a bit fussy when we posed her but I snapped off some frames anyway. This is image is one — the ONLY one — worth keeping. Despite being the only good image I call it an outtake because it’s not at all the image I was after. I like the expression on my wife’s face and Eden’s outstretched arms but it has a few technical issues. For starters, because I was shooting near wide-open and my wife was moving back and forth to rock the baby, the focus is a bit off. We’ll try again soon.
This was shot with two lights: a Canon 580EXii at about 1/16 power in a small softbox at camera left for the key light and a Canon 430EXii high, behind my wife at camera right for hair/highlight (1/64 power and gel’ed with some ND to kill more of the power). The background is a sheet we hung in the hallway (yep, I need to get some backgrounds). I started by setting an exposure which killed the ambient. Using my older daughter as a test subject I then added the key light followed by the hair light. The background is not lit because my intent was to make it pitch black.
Before even shooting this my intent was to process in black and white but I haven’t even attempted to go that route in processing yet. I tweaked some areas in Lightroom then brought the image into Photoshop. I used masked curves to brighten the hair, eyes (a tad), and a few areas of skin. I also used curves to darken a few areas. One final curve dropped the red channel ever so slightly. I sharpened the hair and used noise reduction on the rest of the image. That’s all I can remember anyway…
Here’s another version of the same image which I processed slightly differently. I can’t personally decide which I like best although I lean toward the one at the top of the post which blends subject/background relatively seamlessly.