Time for another in my Before and After series: I’m taking some of my best images and processing them through my workflow step by step so everyone can see what came out of the camera and how it is turned into the final image. (Look here to see all of the tutorials in the series)
The purpose is to explain my processes and techniques in the hopes that all of you will be able to borrow some of them for your own work, and because I hope to get feedback from you on techniques I haven’t discovered yet that will improve my own images.
I’d love to hear what you think of these articles and whether you find them possible. If you have questions on any aspect of them, please leave a comment or drop me an email.
Here’s today’s image as I originally published it. We’re not going to reproduce this, we’re going to start over with the raw and process it from scratch, so the final result will be different (and perhaps better). We’ll find out together…
In the last article I said I’d reprocess one of my bird photos this time. I’ve chosen a personal favorite that’s an older image that I think can benefit a lot by reprocessing with the knowledge I’ve discovered since I worked on this image. This image is from 2010 and was shot at Redwood Shores, a well-known location here in the Bay Area for winter ducks and shorebirds. I used a Canon 7D, my trusty 100-400 lens, and it was shot at 400mm at 1/800 at F8, ISO 400. Good light, fast shutter to freeze action, and moderate depth of field.
The first thing I notice on this shot is that it’s dark, and the duck’s head and tail are too dark. We can fix that..
Processing Cinnamon Teal
Here’s the image in its raw unprocessed form
Camera Calibration and Lens Correction
First thing you should do on every image is set your camera calibration and lens correction. Camera calibration is a subjective thing and different vendors and bodies will have different choices available. Adobe Standard will be there for every camera. With Canon bodies I find I prefer Camera Faithful almost all of the time. What this value does is give “hints” to the RAW rendering engine on how to interpret the data when rendering the image.
Lens correction profiles allow you to tell Lightroom what lens was used. If it’s a lens from a major vendor chances are a profile exists for it. This helps Lightroom auto-correct for certain lens flaws like chromatic aberration and distortions, especially around the edges of the frame. It’s possible to manually make changes as well, but for most images I don’t; the big exception being buildings where I need to straighten up walls and edges.
If your lens has a profile you should use it. If it doesn’t, leave this disabled.
Setting color balance is the next step, up in the Basic section. I’ve found that you can get decent color by using Auto White Balance but that I can rarely get the best possible color by leaving the choice to the camera. It gets close, but it rarely gets it perfect. Shooting in RAW mode allows me the option of deferring the white balance decision to when I process the image, though, so since I shoot almost 100% of the time in RAW, I almost always shoot 100% of the time with AWB set in camera. When I bring the image in to process, I’ll choose a specific white balance that I think best represents the lighting at the time.
This image was shot mid-day but the look of the water and the general light tells me it was overcast; the shadows are soft, the light is a bit greyish and diffuse. Looking at setting a base color balance I check between the Daylight, Shade and Cloudy settings. Shade is definitely too warm. Either Daylight or Cloudy would work, but Cloudy warms up the image while Daylight gives the duck what I feel are truer colors, so that’s what I choose. The color balance as set is pretty good this time, so I could have just left it.
Black and White Point
The next change is to set the black and white points. This will let you adjust the overall contrast of the image and help create the strong blacks and whites that can create impact in an image. Since the duck is dark and the overall image is dark, I’m going to start by raising the white point a fair amount (+56), but where I might leave the black point alone in a well-balanced image since it’s close to touching the left side of the histogram, in this case I’m going to pull back from the black side and raise the black point to lighten the image and set black to (+31). I may adjust that later, but this is a good starting point.
The duck looks better already.
Exposure, Shadows & Highlights
Next up we adjust with the exposure, shadows and highlights sliders. Exposure sets the overall light levels for the image, while shadows and highlights adjust the light and darker areas independently. The shadows slider in Lightroom 5 has become a close friend of mine because it allows me to bring out the detail in the shadows amazingly well, and because of that I rarely turn to the tone curves any more (sssh: don’t tell photographers who live and die by adjusting curves, with Lightroom 5 I find it’s rarely necessary).
If you look at the histogram, while the important details of the image (the duck) are quite dark, much of the image is very light. The exposure of this image was done by the camera and was probably on evaluative metering; the large amount of water convinces the camera the image overall is bright and it exposes for the water, leaving the darker duck under-exposed.
This image is a tricky one to expose well; there’s a lot of dynamic range here to squeeze into the image. One option would be to spot meter on the duck to get it properly exposed, but then the water is going to be much brighter and start blowing out highlights and you risk losing data in it. if I were shooting this image today I’d probably do that but use exposure compensation to pull the exposure back so the water wasn’t blown out. Another option would be to leave the exposure settings as is, but use exposure compensation to boost the exposure by half or 2/3 of a stop.
It’s a bit of a hobson’s choice, do you want to try to fix the duck’s exposure or the water? You’re really not going to get both right in camera, something’s going to need to be adjusted. My normal opinion is that I’d rather try to raise up the shadows than try to bring down highlights because you usually have more and better data to work with in preserving detail — but the side effect of that is a higher risk of noise coming out during processing. This is a common challenge with bird photography because you’re not working in a studio so you don’t control the light, you can’t do much to clean up massive dynamic range images in the field, and you can’t boss the model around to get them in a better position photographically. They tend to just duck-laugh at you and fly away. An option for some is to use fill flash using a hot shoe flash and something like a better beemer, but I’ve worked with that and it just adds more complication to getting the shot and personally, I don’t like the look of images taken that way as much. I’d rather work harder to find flattering light than trying to get all of the gear to cooperate, and flash just adds more gear to haul and curse at when it’s batteries run down at the wrong moment.
Because of the bright water, I won’t touch the exposure slider at all. Instead, I’m going to do the adjustment with the shadows slider. I’m trying hard to bring up the dark parts of the image without adjusting the water, since it’s very close to blowing out the right side of the histogram. Fortunately, we can do a lot with the shadows, and it makes a huge difference. Exposure stays at 0, Shadows go to +54, and to try to calm down the bright waters, I drop Highlights to -28.
One thing to note here. One reason I raised the black point in the previous step instead of leaving it alone is that it changes the starting point for the shadows slider by raising all of the dark tones. This effectively is pushing capabilities of the shadow slider even further towards lighter tones. if that doesn’t make sense, find an image with really dark colors in it like this one, and try raising out the shadows with the black point set so it’s almost touching the left part of the histogram, then try again after raising the black point. It’s almost as if you’re amplifying the effect of the shadows slider this way, and it gives you more oomph for bringing the dark parts of an image into balance with a bright overall image.
Looking at the duck at 100%, you can see it’s really been lightened up and looks pretty good, but needs more work. The dark head is still too dark, but if I push it further using these sliders, the breast feathers start blowing out. I’m going to need to tweak this more selectively. And if you look closely, you’ll see that the processing has brought out that nemesis of photographers everywhere, noise….
Clarity and Vibrance
Clarity can be thought of as a “smart sharpener”, while Vibrance is an intelligent saturation adjustment. A bit of clarity helps most images; I typically set it around +20 and it does some localized contrast adjustments much like sharpening does; on good images, it can sometimes replace doing any sharpening in the detail module although I typically do both clarity and sharpening.
Vibrance has almost completely replaced the saturation slider in my workflow. the Saturation slider adjusts all instances of the colors the same, while vibrance will adjust lower-saturated colors more and high saturation colors much less, making it a more foolproof way to tweak the colors on the image. For this image I’ll use my normal +20 for clarity and vibrance.
The clarity darkens the duck a bit, and the vibrance also intensifies the breast feathers a bit; and having seen that, I’ve decided to change my mind on exposure, and I’ve boosted it by half a stop, which lightens the duck’s headThe breast feathers may need a bit of a tweak later, but after seeing how the clarity and vibrance tweaked things, pushing the exposure a bit saves me even more detail fiddling later. On the other hand, the color noise is also enhanced (check the bird’s dark butt).
I still think the head is too dark. If I lighten it further now, I’ll mess up the body, so this is as far as I want to push it right now (but we’re not done yet). I hadn’t intended it as such, but this is turning out to be a great example of how sometimes the best way to fix thing is through as series of changes that all layer on top of each other, and how if you plan for that, you can take care of the big problems globally without messing anything up and then make more localized adjustments later. Learning that not everything has to be fixed with a big, global brush is a big step towards taking control of the Lightroom process.
Using HSL as part of that more localized fix is a huge part of that ability to work on limited parts of the image, so that’s next.
HSL — Hue, Saturation, and Luminance
Now we move to the HSL section where we can start adjusting colors with more precision. This section allows us to take specific color channels and change the color itself (hue), how saturated the color is, and how bright the color is (luminance). You can make some very dramatic improvements to an image this way, but if you’re not careful you can make the image look like it was snipped out of a bad sci-fi movie, so I’ve found a delicate touch is necessary.
One critical tool to learn how to use in the HSL area is the color picker. Each of the three color boxes have what looks like a donut in the left corner; if you mouse over it it’ll show up and down arrows. Click on it to turn it on.
You can tell one is enabled because it stays white and the arrows stay visible when you move the cursor away from it. You’ll need to learn to remember to turn it off, because it’ll stay enabled even when you try to go off and do something else, leading to wondering if your computer is haunted. It took me a few times to teach myself to always turn it off when I’m done.
The way you use it is to find the spot you want to adjust, and click on it, then drag the cursor up to increase the values down to reduce them. It will adjust all of the colors that are part of the spot you clicked on; the side of the duck is both orange and red, for instance, and has more orange in it. I feel it’s a bit over-saturated, so I’ve clicked on the the side of the duck and dragged down, changing the red by -6 and the orange by -10. This is a LOT easier and more powerful than trying to guess colors and move the sliders individually. Once you get the hang of it, it’s a big help in doing area-based color adjustments quickly and easily. Since the only significant places in the image that have red and orange in it are the duck, it’s an easy way to clean up the colors of the bird. Note that the color changes are global, though — if red and orange existed elsewhere they’d be adjusted as well, which may or may not be what you want. There are ways to do even more localized adjustments which we’ll cover later.
Now I like the color and exposure of the duck’s body. The water looks good, too. I still think the duck’s head is too darl, but I can’t fix it here. In some cases I might finish this in a plug-in like Viveza, but in general I limit the time and work needed to process through plug-ins for my best images, which this one, while nice, clearly isn’t. So we’re going to do the full job in Lightroom, which both makes it non-destructive and keeps it less complicated and usually, faster. It’ll be good enough for the kind of uses I’ll have for an image of this quality, since it won’t be one I’ll ever print or use in high resolution situations.
As much as I love how well Lightroom 5 processes images, I feel that almost any image can be improved further by careful application of some time and a plug-in or two — but I also feel that most images don’t warrant that extra work. I’d rather save that extended processing for the best images and not try to take something that’s pretty average and hope that some ‘magic’ will make it a better image than it is. Some images deserve a lot of careful and time-consuming processing. Most, in all honesty, don’t. One of the things I had to learn as I matured as a photographer was to know which images deserved that time and not waste time hoping a bit of lipstick will turn the pig into something prettier. That time is better spent going out and taking new, better images. Time is most often than not going to be your most limited resource. My belief is that a big part of maturing as a photographer is learning how to use that time efficiently and to maximize your effort on the images that will benefit from the extra effort.
Now we’ll fix the duck’s head. To do that we’ll use a local adjustment using Lightroom 5′s radial fill. I’ve really come to like this feature, which is new to LR5. It’s the second tool from the right in the toolbar below the histogram, next to the adjustment brush. In Lightroom 4 and earlier, this adjustment would have been done with the adjustment brush, but the radial tool works for many of these operations and is even easier to use… So we select the radial fill and then drag out an oval covering the duck’s head, and then adjust the feather. I’ve found the best way to see how the masking and feathering will impact the image is to add a color to the filter while setting the location and feather of the fill, and then remove it when I do the adjustment.
The radial fill gives you most of the power of the Lightroom processing engine, but it only affects the area defined by the mask. Even better, you can have multiple fills in a single image, the same way you can create multiple adjustments with the brush. This is a great tool for those small, localized adjustments and it has a surprising amount of power behind it. So first, I remove that color, and then I go about brightening the head.
You could do that two ways, by increasing the exposure, or by increasing the shadows. I went with the shadows because I wanted to minimize changing the eye, which is brighter. If I set the shadows to +30, the duck’s head now has the same tonal value as the body, but a Cinnamon Teal’s head should be darker, so I back that off to +15. I then decided the eye had been brightened a bit more than I like, so I added a second radial fill with no feather on the eye, and set the exposure to -.83. I also gave it a clarity boost at clarity: +45 to sharpen it. That’ll let me use more gentle sharpening in the next step.
I did consider doing noise reduction in this step, but I decided it would just complicate sharpening later, so I left it alone. here’s a look at the duck at 100% compared to the unprocessed version.
Overall, I think that’s about as far as I need or want to take this image. Now some final cleanup and we’re done.
The image is done, but we need to do some cleanup with final sharpening and noise reduction.
Disclosure: I’ve really struggle with sharpening. I’ve finally gotten decent at it, but it’s taken forever, and I still don’t think I’m anything remotely like “great”. For almost all of my web-based images I use some very basic Lightroom presets borrowed from Jared Platt that work quite well for me. When I start getting into my best images, I’m usually doing at least some processing in the Nik plug-ins and so I’ll use Sharpener Pro on them. I’ll definitely use Sharpener Pro when working on custom prints, although sometimes I’ll sharpen in Photoshop using a high pass filter. But for the typical online image I don’t worry about it excessively, especially given many online services (like Flickr) will ‘help’ your image with some magical sharpening and other tweaks when you upload to them. What really matters online isn’t to get the sharpening just right, but to not over-sharpen and not create visible artifacts, and it’s easy to get too aggressive and do that.
I’ll start with some moderate sharpening, which is Amount = 26, Radius = 1.0, and Detail = 50. Depending on the look, I might soften the detail to 25. Sometimes I’ll ramp up the sharpening beyond this but that’s fairly rare. For this image, my defaults look okay. The only real way to see the effect of sharpening is with the image at 100%.
Now I apply a mask, which removes the sharpening from areas of low detail. The effect of this is to reduce the noise in those areas. I tend to apply masking more aggressively than many because I find I like the reduced noise — and in general, I rarely if ever add artificial grain to my images, preferring the noise-free look. Like sharpening, masking and noise reduction also need to be done at 100%. Hold down the option key and the image goes monochrome with light areas being affected by the sharpening and dark areas being excluded. I turn up masking until most of the low-texture areas have been excluded. In this case masking is set to 67.
Then I’ll do my noise reduction. There are two types of noise in an image, luminance, where nearby pixels are lighter or darker but of the same hue/color, and color noise, where random pixels have random colors in them. In this image, the luminance noise is minimal, but I’ve added a lot of color noise to the image.
Lightroom has separate controls for both types of noise. I’m going to start by reducing color noise, and then add a luminance reduction if I find I need it. As with sharpening, I’m using some presets from Jared Platt here. I add a small amount of color reduction (=25, detail and smoothness =50), and it makes a huge difference. I decide to add a bit of luminance and the difference is subtle but noticeable (=10, detail = 25). These screen grabs show the changes and the final values:
I think that pretty much does it.
Here is the before and after from the RAW image to the final.
And since you never exactly reproduce the same image when you reprocess it, here’s my original version of this image compared to the new one.
I don’t analyze images before reprocessing them for this series, I pick them because I think there are some interesting techniques I can show off with them and dig in. In this case, I hadn’t realized this image had some rather serious weaknesses because of the way I shot it and re-cropped it — things I doubt I’d do to an image today. This is all part of the maturation process, but in reality, I remember this image a lot more fondly than it’s overall quality returns. That said, it’s images like these that can sometimes be the best learning tools because turning them into something useful isn’t easy, and reprocessing this was an interesting challenge. I rather like the result, although I wouldn’t use this image other than on-line. I definitely wouldn’t try to print it.
The Cinnamon Teal is quite a pretty duck, also. I really need to go out and get some better images of them…
Thanks for reading, and for your feedback. Next time — I’m not sure. Any suggestions of an image of mine or techniques you’d like me to talk about?
I’ve also considered processing images by others; if you have an image you’d like to see processed by me, drop me a note and we’ll see if it can be done.