I get a lot of questions about work flow. In today’s post I go through, step by step, my workflow on a landscape image. Its not the ONLY workflow I have; it really depends on the scene, but its one I used quite often with some recent images. Many of the images I shot on my recent trip to Death Valley, from Badwater to Furnace Creek to Zabriskie Point, particularly the images shot in Alabama Hills, required a very technical approach. The location is amazing, but I did not feel that HDR engines, like Photomatix, were working sufficiently. So, without further delay, I’ll walk through an image shot on the trip.
General Comments: I almost always bracket +- 2 stops. This guarantees that I’ll have a full range of information for every pixel, i.e. full detail in shadows and full detail in highlights. I may not use all that data, but I have it somewhere in a raw file, providing me the best possible data to start with in my post production workflow. Often, I can use two raw conversions of the same frame to get what I want, but in this example I’m actually using two different frames at different exposures and going old school post production.
Step by Step:
1. Once I’ve selected the frames I open both in ACR and make my global adjustments. Most important is to work out the chromatic aberration (CA). Doing it at this stage with save a lot of work later. ACR has a menu selection to handle CA and you can find it in ACR; I also make use of another useful tool, the Parametric Tone Curve tool in ACR. Play with it, you may like the level of control it provides you;
2. What ever global adjustments I choose to make, one issue I always keep my eye on is the historgram, specifically the Red channel in the histogram. It is easy to lose detail in the Reds (Red channel stacks up on the right end of the historgram). In the image below I’ve created a situation where I’ve lost detail in the Reds, which is in the sky.
3. I open images as 16 bit, 300ppi files. If you’re an image quality freak, you’ll get this. If your larger concern is computer RAM limitations and storage, then you’ll probably use 8 bit. A 16 bit file can take quite a bit more punishment, especially when working in black and white; nonetheless, its what I do. Its a better starting point;
4. My next step is to stack the images in Layers. My choice is to place my foreground image as the Background layer and stack my sky frame on top of that.
4. Now things start to get a little more technical and require some elbow grease. I want to select out the sky from the foregound. As always in Photoshop, there are a 1000 ways to skin a cat and I’m open to ideas of how to better do this. I use a Magic Wand tool set to the default tolerance of 32 and make a quick rough selection. Then I zoom in to at least 100% or more, shift the tolerance to 20 or lower, and make refinements to the selection.
Step 5: Once I’m happy with the rough selection of marching ants around the sky, I make Layer 1 active and click on the Add Layer Mask icon. With the Layer 1 mask selected and active, I invoke the Refine Mask tool, found in Select/Refine Mask menu. This tool requires a bit of experimentation, but with some practice you’ll find what works best on your selection. I also view the image at around 300% to get a good view of what Refine Mask is doing. With that being said, I’m finding a starting point on a clean horizon selection like the one in this image that I use the Smart Radius selection and set the radius to about 2.5 pixels, but in this image I can take a lot more and get a better refinement. As it turns out, I can push this quite far, in the range of 40 + pixels.
Step 6: With the mask the way I want it, I take a potentially important next step. I reactivate the marching ants selection of the mask. This is done, on a Mac, by holding down the Apple key and clicking on the mask. With the selection active I choose Select/Save Selection. This places your selection in a new Channel. Don’t worry about what that means. Just know that your selection is saved and you can recall it at anytime. You may want to do this to make local adjustments to the sky at a later point in the workflow. Why do I do this? Because I don’t like a lot of layers. At this point I tend to flatten my image and work from here, but I still have the selection for sky stored in Channels and recall it later, if necessary.
Step 7: At this point I have a complete exposure of raw data. I now clean noise from the file. There are some different schools of thought, but it seems the dominate school is to clean noise early in the editing process. There are many, many different opinions on how/what software. I use Noise Ninja because that’s what I’ve always used and it works well with my files. I’m now ready to begin some additional global and localized editing for color and contrast.
Step 8: I switch the color mode to Lab. I won’t get into the pro’s and con’s of Lab color mode (Image/Mode/Lab). I like what it does and I like the efficiency with which I can get it done. I use Lab for color seperation, that is, to push colors away from each other, increasing color contrast (vs. tonal contrast) in the image. Its a nice boost to the image. This is also a great place to do some tonal contrast. If you want, you can search the internet or read Photoshop Lab Color by Dan Margulis. Great book. Nice read during these winter months. Word of caution: Its easy to go to far to fast with a curves adjustment in Lab. Experiment.
WARNING: WATCH YOUR RED/YELLOW, ie. ORANGE colors and make sure you don’t start loosing detail in them. If these colors are stacking up on the right side of your historgram, you’ve gone to far.
Step 9: Switch back to RGB color mode, check the warm tones for detail. Here I’ll finish most of my color work using a Selective Color adjustment layer. Play with it. Its very powerful. As you can see in the image below, I’ve held detail in the Reds and used the ability to darken blues, which brought out detail in the warm tones of the clouds.
Step 10: Local contrast work with a variety of tools, including Curves, Brightness and Levels adjustment layers.
Step 11: Ok, now were getting down to brass tacks. I’m not a big fan of “master” files, as many folks will use these for different reasons. I’ll use them, but I’ll throw them out later to save disk space. So, I might have saved off a master along the way in this process, but I just don’t find myself going back to them very often. If I rework an image, I’ll go back to the starting point. I do this with my fine art work and with my commercial work. So, here I flatten the image and get ready for final finishing steps. Zooming back in on the image I see that the horizon where I masked in the sky is still a little, well, less than perfect. I’m going to sharpen for output here, today using PhotoKit Sharpener. My experiments with it show it to be fairly good, but harsh on noise and places like the sky – foreground interface. You know what I’m talking about here – HALOS. I don’t like them, don’t tolerate them and will work my tail off to keep them from anywhere in my image. So, now I’ll sharpen but then I’ll go back in the final step, step 12 and clean it up a bit. Hey, if you’ve made it this far then read on. Sharpening routines sharpen everything, including NOISE. PhotoKit is no different. So, I invoke the sky mask we saved earlier to mask off the effects of sharpening from the sky. The sky is critically sharp already and doesn’t need any noise enhanced in it.
Step 12. Well, here we are the final, and most annoying task. Not everyone will do this, but I do. Sharpening has left a small halo at the sky – foreground interface and I’ll use the clone tool to clean that up. It is not as bad as it sound, but I recommend a good strong beverage while you do this. It will be over with in no time.
Ok, no more steps. That’s it. More or less. I did a few more things to finish off the image you see at the top of the post, but you can figure that out. Hopefully this was helpful. If you have questions or suggestions for better/more efficient technique, please leave your comments in the comment section for all to see. THANK YOU.