Real-life retouch: a Paris bridge

As promised in my last mess of a post, I want to start doing real-life retouch examples of things I am actually working on. Right now, I'm still sorting through the 3000+ photos I took in Paris in July. The one above was taken on the beautiful Pont Alexandre III. I think it has potential, but I don't like how it's reading. As I take you through the steps that lead me to something I like, you will notice that:

1. I believe in using a very intuitive approach
2. curves adjustment layers + masks are my best friends

First, I identify what I don't like about the photo: the statue is too dark and I would like it better if the highlights stood out a little more.

My solution:
1. I add a curves adjustment layer.
you can add a curves adjustment layer from the fill/adjustment layer pull-down menu at the bottom of the layers palette or, if you have CS4 or later...
you can click on the s-shaped curve on the top row of all those boxes above your layer's palette. The curves adjustment layer shows you a histogram (the mountainous graph that maps your image information from dark[left] to light [right]). You can see that I have a lot of information on the dark side of things. To fix that, I click on the diagonal line (I happen to click close to the lower left corner—i.e. down in the darkest area, although it doesn't matter a lot in this case), which gives me a control point (a little dot I can use to drag the curve around). I don't use a formula, I just experiment and see what looks good. Here's the curve I end up with:

Had I pulled it in the opposite direction, I would have made the image darker.

2. "paint with light"
The sky is fine, it's the statue part that needs to be lightened. To undo the effect on the sky, I simply brush it out. First, I hit "b" to get the brush tool. Then I adjust my settings:
Once I select the brush tool (b), the top toolbar gives me options (click on the dot to the right of "brush" to see them) to change the size of the brush, the hardness, and so on. I start with a huge size (1600 px) because I'm making broad sweeping changes. However, I often change size on the fly by using the right (bigger) and left (smaller) brackets on my keyboard. I put the hardness all the way down to zero because I want completely soft, gradual transitions (fyi—you can also adjust hardness as you go by using shift with the left and right bracket keys). One more setting that I adjust as needed: flow. It's like controlling how fast ink is coming out of the brush. I want to paint in changes gradually so I can see if I like where it's headed, so I start with 17%—a fairly random choice.

One thing to keep your eye are those white and black boxes (if yours are not white and black, hit "d" to bring them back to default). Whichever color is on top is the color you are painting.
I want to paint with black because I am blacking out the effect on the sky. Toggle between white and black by hitting "x." As I said earlier, I like intuitive retouching. I simply paint until I like what I see. You can hit the backslash key to make your painting visible in red like this:
The red shows where I have been painting out the lightening effect on the curves adjustment layer. Here's what the black looks like in the curves adjustment layer mask over in the layers palette:
3. Are we done yet? No, one more curve.
For a moment, I think I'm done, so I go to the pull-down menu (the little triangle at the top right of the layers palette) and choose "flatten image":
Goodbye separate curves layer. Now I can save and quit...or can I? No, I'm still not quite happy. I would really like to see the highlights on the statue pop a bit more, so it's back to curves again. Instead of lightening everything and painting out everything but the highlights, I pull the curve down to darken everything...
And then I paint the highlights back to their lighter state:
It may seem like I just undid most of the lightening from my earlier curves adjustment—and maybe I did—but this is a try-as-you-go process and as long as it's moving me toward something I like then I'm happy.

4. Meh. The color isn't doing anything for me.
This isn't exactly the most spectacular sunset in the world, and although I like the gold, I don't like it enough to convince me that this wouldn't be better in black and white. Another motive: black and white sells better than color in home decor, and I've got bills to pay.

And here is where the "real-life" retouch gets hypothetical for a moment. The truth is, I like to do black and white conversion in Silver Efex Pro, but if I were to use Photoshop, I might pull up the black and white conversion from the Image Adjustments menu and start to play with the sliders:
Confusing? Yes, so I like to work on each slider, pull it to the extreme to see what part of the photo it affects, and then make a more subtle adjustment. Look at the "yellows" slider, for example. If I pull it all the way to the left to -200%, the gold virtually disappears:
If I pull it all the to the right (300%), the gold becomes shiny beyond belief:
Trial and error has shown me a way to make the gold stand out. So at this point, I would set it to something more reasonable (140%). There is also a "tint" checkbox that I decide to try for fun. I end up setting the hue to 35 and the saturation to 1% just to give a slight warmth to the photo. Where did I get those numbers? I just dialed the saturation up high, moved the tint around to find a good color, and then dialed the saturation all the way down. In Photoshop, I now end up with this:
It's a fairly nice black and white. But as I mentioned earlier, I usually do black and white in Silver Efex Pro, so in reality, I take the color version of the photo, and redo the black and white in that program instead. What I'm really in the mood for is high contrast and some grain, so instead of the more subtle black and white above, I end up with this:
It's all a matter of preference. I'm kind of waffling between the two versions. The latter version is perhaps too contrasty online, but print always tones things down a little. Either way, I like the end product way more than the beginning.

And there you have it. My first walk-through of a real life retouch.
Helpful? Not so much? Want more or not?

Photoshop staples, or "Yes, it is overwhelming."

This photo has nothing to do with the post, but hey! it's pretty and this is possibly the most schizophrenic post I've done, so why not?

First, the intro that I wrote before becoming overwhelmed and deciding that I needed to watch a movie instead of tackling the impossible...

So, let's take a hypothetical reader who just installed Photoshop on his computer—we'll call him "Mark." (You see, readers. This is what happens when you leave comments. I might actually do a tutorial with you in mind.) Photoshop can be overwhelming. Too many menus and functions. You can buy a book, and probably should, but choose wisely. Some books try to be all things to all people and end up jumping from stylized typography to drawing tools to curves until you just give up. I've always been puzzled by the tutorials in most Photoshop books. "Let's cut out this clown's head, paste it in the middle of a desert, make flames come out of its mouth, add some text with cool 3D effects and drop shadows, and turn it into a web page!" OK, I made that up. But only because I don't have one of those books on hand. Next time you're in a bookstore, flip through a few and you'll see what I mean. Ask yourself if you would ever do any of those projects. Yes, they're supposed to teach principles, but why not teach them in more of a real-world scenario? For me, "real world" means photography, so I prefer books that focus on photography such as Martin Evening's weighty tomes. They're not exactly curl-up-by-the-fire reading (unless you're the type of person that reads encyclopedias for fun), but they are a great reference tool.

I still stand by my reading suggestions (in whatever edition is the most current), but for more immediate satisfaction, I will identify 3 staples of my own real-world Photoshop use. If my "starters" tutorials are little recipes, this is a look into the pantry.

Lasciate ogni speranza...
And now, here's the part where I interrupt myself and explain why it is indeed overwhelming after all.

I thought I was going to boil it all down to 3 elements. If I had they would have been:

1. Layers
It always starts with command-J (or Ctrl-J for PC), which duplicates your background layer. I know it sounds extremely basic, but when I first started doing Photoshop, I would often forget which layer I was working on. Had I actually continued the course of this post, I would have explained the difference between making changes to duplicate layers v. making changes to adjustment layers, and by about 1 a.m. I would have just told everyone to buy the book Layers.

2. Curves
I use curves more than anything. My "adjusting tone and color tutorial" shows a technique that need not be limited to photos of happily engaged people sitting in dead grass. If you experiment with the individual color channels and with the "all channels" s-curve (see step 8 of the same tutorial), you will be working on a technique I use in most photos. When I attended an Amy Dresser webinar, I saw her lasso highly-feathered selections dozens of times just to make subtle curves adjustments to each.

3. Masks
If you can't selectively apply adjustments, there's really no point in doing Photoshop. Quick masks are a good place to start because it is a very intuitive technique.

Had I really gone into detail with each of the above, I would have an endless post, so, yes, it is overwhelming. I still think the layers, curves, masks thing is important, but my desire to show "real-world" retouching has led me to conclude that I should start sharing start-to-finish retouching that isn't based on one tutorial. Nothing could be more honest than to walk you through my actual process. And so, I'm adding a new feature to take-out photo: the real-life retouch. Once every couple of weeks I will show the steps I take from RAW to retouched. Sometimes I do very few adjustments, sometimes I do dozens. I'm not promising an Amy Dresser webinar or anything, but I like the idea of de-mystifying the process and I'm guessing that it will lead to more specific "starters" tutorials as well. It may turn out to be underwhelming, but sometimes underwhelming is exactly what we need.

Refine: le mieux est l'ennemi du bien

A quote usually attributed to Voltaire (which he attributes to "a wise Italian")—Le mieux est l'ennemi du bien—has always disturbed the perfectionist in me. A few of the English translations are:

• Better is the enemy of good
• Better is the enemy of good enough
• The perfect is the enemy of the good

When I was living with my dad and his artist friend in France, I heard them throw the phrase around, often in reference to her work in ceramics. If you've ever worked at a potter's wheel, you know how quickly a perfect vase can suddenly turn into a muddy heap just because you thought you could make it a little thinner, a little taller. In that context, "better" destroys "good." In other contexts, the enemy line is not so clear. In fact, I often have trouble deciding which side I'm on. Why should I fight for good when I can be a martyr for better? Besides, if I learned anything from watching Six Million Dollar Man episodes as a child it is that "better, stronger, faster" is always right.

In Voltaire's moral tale, La Bégueule, a foolish woman who likes pretty, shiny things learns the perils of always trying to trade up. Improving talents, knowledge, and moral character is always desirable, notes Voltaire, but the rest can lead to a life of illusion. In Photoshop, where, let's face it, we are often creating illusion, the quest for pretty, shiny pictures can quickly go too far. I won't moralize, but I will say a few words about refinement in retouching.

In my dodge and burn tutorial, I try to show how very subtle skin retouching is always better fake, plastic-looking skin. The shortcut skin smoothing methods are useful as well, but the problem is that when you see Barbie-doll skin, it can be hard to dial things back down to "good enough." For example, I have seen a lot of people do something like this...

when this is plenty good enough:
Self-restraint is hard when you have just learned how to make people look perfect. I look back in horror at some of the work I did in my first year or two of retouching. But since then, I have developed a rule:

Do the correction and then dial it down.

It's the Photoshop equivalent of Coco Chanel's advice, "When accessorizing, always take off the last thing you put on." How do you edit accessories in Photoshop?

1. Do your adjustments on separate layers
2. When you're happy with an adjustment layer, go to the "opacity" box in the layers palette and lower it from 100 to 80 or below. Look at step 3 of my "fix dark photos" tutorial and you'll see what I mean.

That's it. That's probably my top rule of refining photos. I always do it. The only possible exception to the rule is when you are looking for an extremely stylized effect. Otherwise, always turn it down. This applies to just about everything:

• sharpening
• contrast
• saturation
• etc.

One of my pet peeves with tutorials is that they almost always encourage excess just to make sure you can see the before/after difference. In my dodge and burn tutorial, I was worried that people wouldn't even notice the before/after difference, but I didn't exaggerate the "after" because I wanted to reinforce my preference for moderation and at least a modicum of verisimilitude.

It's OK to fight for good even though better might be brighter, sharper, more colorful, and more dramatic. Trust me, your photos will end up looking more refined.

Revisiting old photos

The theme this month is "revise, reshoot, refine," but I can't help but add "revisit" to the list. If you've been reading this month's posts, you know that the theme is inspired by the writing process. When it comes to writing academic papers, I have a fairly tedious method. First, I cast the net wide. Let's say, for example, that I want to write about scientific instruments in the 17th and 18th centuries ("Why would anyone want to do that?" you say, and yet I did). I don't like to know exactly what I'm going to say in advance because it doesn't seem very honest. So I read book after book and take notes. I feel like I have to reach some sort of critical mass before I stop collecting notes and actually start writing. One of my best experiences in the "casting the net wide" phase of research was when after a couple of months in the same French library, the librarian let me browse the hidden stacks of not-yet cataloged rare books. For me, it was as if I had tamed the fox in The Little Prince. This is a complete digression, I know, but before I start making things apply to photography, I have to say that going through neglected books, papers, and ephemera is hands-down my favorite part of research (Don't get me started on my week at the Cooper-Hewitt with 12 boxes of papers related to the origins of industrial design or I'll never get this post done.) I like to think that it's the same feeling William Eggleston (one of my favorite photographers) has when he takes a picture of a ceiling fan or a stack of garbage bags.

Did I lose you yet? Let me bullet point my tedious research/writing process and then skip to how this applies to revisiting old photos:

• The initial phase leads to perhaps 100 pages of typed notes
• I then move on to the "notes of my notes" phase, where I pull out possibly relevant quotes and get it down to maybe 36 pages.
• Finally, I write the actual academic article (that will be read by far fewer people than this hasty blog post) and use only a fraction of those notes.

The parallels with photography seem clear, don't they?

• First, I take way too many photos (unlike Eggleston, by the way, who thinks that if you take more than one picture of the same thing, then it's just too painful to have to choose)
• Next, I rate my photos in Aperture. One star means "Why am I keeping this? I should really just delete this. Help! William Eggleston was right!" Two stars—i never use two stars, what's the point? Three stars means the photo has no major technical flaws. Four stars means I think this could be the one, but I don't want to commit until I have seen all the options. And five stars means I think the photo is worthy of actual post-processing attention.
• Post-processing doesn't mean I'll ever bother to print. A very small fraction of photos end up on my wall or in an album. And the ones that make it aren't always the best ones; they're just the best ones for a certain context.

But what happens to the rest? Are they worth revisiting?
This struck me when Becky from "Life with Kaishon" included the following photo of mine in her interview of me:

I actually love this photo, but I had completely forgotten about it. I took it two years ago as Eva was sleeping on my chair. She's five now, and the blanket is still a huge source of comfort for her, although most of the time it has to stay in her bed.

In my hunt for the original file in Aperture, I opened a folder of photos I took for her birth announcement and there it was—the blanket, making its pink and pristine début as a backdrop for our equally pink and pristine baby girl. I immediately made a diptych of the two photos:

There's an empty 8x10 frame on our wall that I have never gotten around to filling. Now, thanks to some digging through forgotten photos, these two become part of a printed composition.

The moral of this story is that a new context might give you a renewed interest in some images that would otherwise fall into oblivion. Take a little time and open one of those folders of neglected images. Wander the uncataloged stacks of your photo library. See what inspires you now. See what a little cross-referencing can do. Give something a chance to make it into print.