Read Sharpening Your Photographs Chapter 7.indd text version

Chapter

7

Multipass Sharpening

The traditional approach to sharpening is to apply sharpening globally just before the file is sent to the output device. Bruce Fraser argued for a three-pass sharpening approach in his online articles for creativepro.com and for multipass sharpening in his book, Real World Image Sharpening with Adobe Photoshop CS2. You can still read Photoshop experts who argue in favor of tradition and against more than one sharpening pass. The logic goes something like this: "When you sharpen more than once, any artifacts you create in the first pass can become obvious in the second pass." It's certainly true that multipass sharpening requires care. It calls for a gentle hand on the sharpening settings. Especially in the first pass. If you're the sort of retoucher who thinks anything much below Amount = 500 is for wimps, you should stay away from multipass sharpening. Done carelessly, it sure certainly make a mess. It's also true that any visible sharpening artifacts you create in the first pass will get enhanced in subsequent passes. That's why the first pass requires care. We're not going to apply sharpening sufficient for a printed photo during the first pass.

Sponge boats at Tarpon Springs, FL.

Multipass Sharpening

184

Adding a final dose of sharpening is what output sharpening is all about. We're not looking to make individual features stand out, either. That's the job of creative sharpening. All we want to do is restore sharpness lost during digital capture. The concern about presharpening -- that it can introduce artifacts that we can't be remove later -- is the reason why I advocate turning off all sharpening by your digital camera or scanner. The camera doesn't know significant photo details from the insignificant. It simply can't provide optimal sharpening. The best it can do is aim for average results.

The Philosophy Behind Multipass Sharpening

I don't know if Bruce Fraser was the first person in print to propose sharpening in multiple passes. If he wasn't the first person, he certainly deserves credit for popularizing the idea. Bruce advocated two passes in earlier writing, later he recommended three passes, and then argued for multiple passes in his book Real World Image Sharpening with Adobe Photoshop CS2. His thoughts on sharpening in multiple passes certainly had an impact on my own thinking, and I want to be sure to acknowledge that. Bruce noticed that we sharpen photos for three main purposes:

USM Sharpening: 100, 0.7, 0. Applied twice.

185

Chapter 7

· · ·

We lose sharpness during digital capture. We sometimes apply sharpening creatively to emphasize specific features of a photo. We eventually convert most of our digital photos into a halftone dots, stochastic dithers, or continuoustone dots for output on some sort of press or printer, which also softens their appearance.

The basic challenge is rather obvious when you break sharpening down into those three main purposes. Rarely will one sharpening setting satisfy all three needs. The idea behind multipass sharpening is to apply optimal settings for sharpening at the appropriate places in our photo editing workflow. It is very likely, for example, that the settings to restore sharpness lost during digital capture will be different from the settings used to output that same photo to a dot matrix printer. We take care of the sharpness lost during digital capture close to that point by making it one of the first steps in editing a photo. Output sharpening occurs after cropping and resampling, just before we send the photo to some sort of output device. You'll find that you need less aggressive settings with multipass sharpening. Applying sharpening twice at the same setting does not result in a simple sum of the sharpening effect. Instead, the sharpening effect compounds. Let's take a quick and easy example. When we apply USM 100%, 0.7, 0 twice, we get a boost in perceived sharpness equivalent to USM 300%, 0.7, 0.

USM Sharpening: 300, 0.7, 0. Applied once.

Multipass Sharpening

186

(1.0 + 1.0)2 - 1 = 4 - 1 = 3, which is 300% Apply the filter three times and you cube instead of squaring the sums. Here's what happens with USM 50%, 0.7, 0 applied three times: (0.5 + 0.5 +0.5)3 - 1 = 3.375 - 1 = 2.375, which is 238% When the sharpening compounds though multiple passes, the artifacts do not compound. We get artifacts as if the multiple rounds of sharpening were summed. This means we get nearly identical artifacts for applying for USM 50%, 0.7, 0 once or twice. But we get a sharpening effect equivalent to Amount = 238%. That's a very cool benefit: perceived sharpening increases much faster with multipass sharpening than objectionable artifacts. The concern about applying sharpening more than once has some merit when sharpening is applied globally and without the benefit of Blend If settings to protect the brightest highlights and darkest shadows along the edges. Sharpening without benefit of layer masks and Blend If settings is undesirable, whether you sharpen with one pass or multiple passes. Sharpening in multiple passes without layer masks and Blend If settings is extremely risky. We're going to approach multipass sharpening with enough smarts to use layer masks and Blend If sliders.

Unsharpened crop.

187

Chapter 7

Capture Sharpening

Here's another mantra. The goal of capture sharpening is to restore the sharpness lost during digital capture and nothing more. It's a gentle round of sharpening. Don't expect to see a large increase in the perceived sharpness. The photo will still require additional sharpening before output: even if the intended output is a file for display on a Web page. The need for capture sharpening varies. Just like the need for noise reduction. When you push your digital negatives to the limit to make the largest enlargements possible, the greater the need for sharpening tailored to capture source and for noise reduction. When you intend to downsample a digital image, you might be able to avoid noise reduction, capture sharpening, or both. Especially when you start with a high resolution digital photo or high resolution scan of medium or large format film. Enlargement factor is an important consideration. When we apply capture sharpening, it has to be tailered to the content of the image. The choices we make for mask edge width and Radius setting are critical. The sample photo for this chapter is a 6MP Canon D60 image. It does have large areas of uniform color, so we might think about a wider Radius setting. The photo also has lots of small details, however. The nets, the booms for the nets, the rigging, etc. The image would have more pop if those fine details had a little more emphasis. The first decision is mask width. The rigging is a fine detail, so I decided to try a narrow mask width. 1. Open Cap'nNathan.psd in Photoshop. The file is located in the \MultipassSharpening folder. 2. Add an edge mask to the photo. If you use my TLR Edge & Surface Masks actions, select Narrow Luminosity. If you use one of the Javascripts from the TLR Professional Mask Toolkit, select Narrow and Edge. You can also make the mask youself using the directions in chapter 4. 3. Add a duplicate layer to the photo.

Toggle off the visibility for the other layers, and you can see the effect of the mask width.

Multipass Sharpening

188

The red highlight shows the boom that I used to gauge the capture sharpening settings. 4. Load the alpha channel for the edge mask as a selection. 5. Create a Reveal Selection layer mask. 6. Toggle the visibility of the Background layer off. You can judge whether the mask is too wide, too narrow, or just right. I selected one of the net booms to visually target my capture sharpening settings. Remember, the monitor is a dangerous guide for sharpening when you intend to send the photo for output to other devices. With capture sharpening, we're only trying to restore sharpness lost during the digital capture process. We don't want to sharpen aggressively enough to add "pop". We're going to have at least one more round of sharpening and each additional round of sharpening that affects the same pixels builds up sharpening much faster than just the sum of their Amount settings. If we make the halos too wide or give them too much contrast at this point, we'll have serious problems later. I watch carefully for the development of obvious artifacts during capture sharpening. I strongly recommend that you load Capt'nNathanCaptureSharpened.psd into your photo editor and watch the highlighted area as you toggle the capture sharpening layer off and on. You won't see a "grab you by the socks" sharpening effect. Going to 100% magnification will help you see the effect as you toggle. It's a subtle sharpening effect, and that's all that's required. Capt'nNathanCaptureSharpenedCS3.psd has the same settings. It uses Smart Filters so Photoshop CS3 users can experiment with the USM settings.

189

Chapter 7

Unsharpened original.

Blend If settings for capture sharpening a medium resolution digital camera photo.

Multipass Sharpening

190

Pixels that will remain unchanged as a result of the Blend If slider settings.

The example of capture sharpening on page 192 uses a single layer for all of the capture sharpening. The USM settings were Amount = 150, Radius = 0.6, Threshold = 0. Without an edge mask to restrict the sharpening to the edges, these settings would be much too aggressive for capture sharpening. Opacity = 65% for editing headroom also attenuates the sharpening effect from Amount = 150. I applied a layer mask with narrow edges. The Blend If slider settings I recommend for single layer capture sharpening are displayed in the screen shot at the bottom of page 190. These are not the same settings from Chapter 3 that were used to protect the extreme shadows and highlights when sharpening is done with one pass. With capture sharpening, we have to be mindful that subsequent sharpening passes will heighten the effect. We, therefore, want to constrain capture sharpening more to the midtones and further away from the extreme shadows. Th e Blend If slider settings on page 190 will allow no changes to the pixels with brightness levels less than 24 or greater than 208. Th e pixels from the underlying layer(s) will instead punch through and replace any of those pixels in the sharpened layer. If any pixels in the sharpened layer are pushed below 16 or above 224, again, the pixels from the unsharpened layer will replace them in the sharpened layer. The figure above shows a duplicate layer with nothing but the recommended Blend If settings. The Background layer was replaced with a bright red fill layer. The bright red areas mark out the pixels that would allow any underlying layer to punch through as a result of these Blend If settings. You can also see the affected pixels in the figure at the top of the next page. The reason for the subtle sharpening effect with Amount = 150, Radius = 0.6, Threshold = 0 becomes obvious in this case when you toggle off the

191

Chapter 7

Capture sharpened pixels.

Capture sharpening: One layer USM (150, 0.6, 0)

Multipass Sharpening

192

visibility of the Background layer. Take a look at the figure at the top of page 192. The sharpening effect is very tightly constrained within narrowly defined edges. This is an image where separate sharpening of the light and dark pixels can be a benefit. If we get too aggressive with the light contour, we can easily create visible artifacts, even if we use an edge mask and and carefully set the Blend If sliders. The trees and the red roof could get some more definition, however, if the sharpening for the dark contour is done with a slightly heavier hand. The details in the tree leaves are predominantly dark contours, also. The detail in the red roof, too. We can apply heavier sharpening settings for the dark contour, if we sharpen it separately. I settled on Amount = 150, Radius = 0.6, Threshold = 0 for the light contour. I watched for visible sharpening effects on the warning lights at the tops of the masts for the boats towards the center of the photo. I wanted to avoid any obvious halos at this stage. For the dark contour, I preferred Amount = 200. My attention there focused on the tree at the left edge of the photo. I wanted to restore sharpness to the leaves. The Blend If settings for multipass sharpening of the light and dark contours are different from the settings we discussed in chapter 6. Again, when we apply capture sharpening, we want to keep the effect focused more on the middle tones and keep it away from the extreme shadows and highlights. We need to leave some headroom for output sharpening. If we push the edge pixels too far during capture sharpening, when we do get to output sharpening, lots of pixels along the edges, especially along the light contour, can clip and become noticeable. We discussed surface sharpening in the last chapter. It's a possibility, but we'd want to weigh our options carefully here before applying any surface sharpening.

Capture sharpening: Light (150, 0.6, 0) and Dark (200, 0.6, 0)

193

Chapter 7

Blend If settings for the light contour of the capture sharpening halo.

Capture sharpening: light contour pixels.

Multipass Sharpening

194

Blend If settings for the dark contour of the capture sharpening halo.

Capture sharpening: dark contour pixels.

195

Chapter 7

There are some textures in the trees, bushes, red roof, and the boats that might benefit from a light application of surface sharpening. The big issue is that there's also a lot of sky in this photo. When a large portion of the photo contains sky, water, or shadows, we risk exaggerating any noise that might be present if we get even remotely aggressive with the surface sharpening settings. The Threshold setting can minimize the sharpening of noise, if you do apply surface sharpening, but care is still required. Photographer to photographer, I would almost certainly forego surface sharpening for this photo. Any surface textures that I might want to sharpen could wait for creative sharpening. I decided to put my intuition to the test, however, and added a surface sharpening layer. The surface sharpening settings were Amount = 100, Radius = 0.6, Threshold = 5. If you toggle the surface sharpening layer on and off for Capt'nNathanCaptureSharpened.psd or Capt'nNathanCaptureSharpenedCS3.psd, you can see that the surface sharpening layer does bring some more definition to the foilage. I carefully examined the sky gradient for evidence of emerging noise. I even used the inverted Difference layer maneuver from chapter 5 to check and see if noise was emerging. I applied Auto Levels to boost the contrast in the inverted Difference layer, just in case the emerging noise was so slight that it might be hard to see. Threshold = 5 kept the minimal amount of noise in the sky under control. To keep things focused in this example, I didn't apply any noise reduction. My experience tells me that multipass sharpening will make even barely visible evidence of noise in the originalevident by the time I finish with output sharpening. If I was coverting the photo to a B&W or sepia toned print, I might welcome the emergence of a little noise for its film grain-like appearance. With a color photo, I would run something like Neat Image or Noise Ninja before applying capture sharpening to a photo like this, especially if I intended to use surface sharpening.

Capture sharpening: surface sharpening pixels.

Multipass Sharpening

196

Capture sharpening: Edges and surfaces, Background layer turned off.

Capture sharpening: Dual layer USM plus surface sharpening.

197

Chapter 7

The dynamic range and the resolution of capture devices differ. This affects the candidate settings for both USM sharpening and the associated Blend Ifs for the sharpening layer(s). I've supplied Blend If and USM settings for five different categories of digitsl capture devices. The Blend If settings are generic to different mask widths. The USM settings are not. The Amount and Radius settings vary depending on the width of the sharpening mask. · · · · · High Resolution Digital Camera. 8 million pixels and above. Medium Resolution Digital Camera. 4 to 7 million pi. Low Resolution Digital Camera. 3 million pixels and under. 35mm Film Scan. Small and medium format film scans. 4x5 Film Scan. Large format film scans.

If you're feeling bewildered with all of these options or you prefer to make a quick and easy action for capture sharpening, you can work with the generic settings on page 190. You'll likely get better results if you start with settings that have been tested on digital photos from the familiar devices in the table on this page and the next. The list is not comprehensive, so you might need to experiment. For example, if you use medium format film and scan it, you might try the 35mm film scan settings, the 4x5 film scan settings, or something more intermediate. You'll notice that Amount tends to increase and Radius decrease as the edge mask narrows. These settiongs for Amount might seem excessive, especially if you've become accustomed to working with the Unsharp Mask filter without an edge mask.

This Black This White Underlying Underlying Black White

Image Source High Resolution Digital Camera High Resolution Digital Camera High Resolution Digital Camera Medium Resolution Digital Camera Medium Resolution Digital Camera Medium Resolution Digital Camera Low Resolution Digital Camera Low Resolution Digital Camera Low Resolution Digital Camera 35mm Film Scan 35mm Film Scan 35mm Film Scan 4x5 Film Scan 4x5 Film Scan 4x5 Film Scan

Contour Light Dark Single Light Dark Single Light Dark Single Light Dark Single Light Dark Single

L 70 30 30 10 10 10 90 0 0 10 0 0 0 0 0

U 70 50 50 40 40 40 90 0 0 30 0 0 0 0 0

L 100 70 100 140 90 140 140 90 140 180 180 180 180 180 180

U 195 70 195 250 140 250 250 90 250 250 250 250 250 250 250

L 25 25 25 90 10 10 10 10 10 70 10 10 90 10 10

U 40 50 50 90 40 40 40 40 40 70 30 30 90 30 30

L 80 70 80 140 90 140 140 140 140 180 70 180 180 90 180

U 200 70 200 250 90 250 250 250 250 250 70 250 250 90 250

Blend If settings for several digital capture sources.

Multipass Sharpening

198

Image Source High Resolution Digital Camera High Resolution Digital Camera High Resolution Digital Camera High Resolution Digital Camera Medium Resolution Digital Camera Medium Resolution Digital Camera Medium Resolution Digital Camera Medium Resolution Digital Camera Low Resolution Digital Camera Low Resolution Digital Camera Low Resolution Digital Camera Low Resolution Digital Camera 35mm Slide Film 35mm Slide Film 35mm Slide Film 35mm Slide Film 4x5 Slide Film 4x5 Slide Film 4x5 Slide Film 4x5 Slide Film

Mask Width Wide Medium Narrow Extra Narrow Wide Medium Narrow Extra Narrow Wide Medium Narrow Extra Narrow Wide Medium Narrow Extra Narrow Wide Medium Narrow Extra Narrow

Amount 275 300 300 325 150 150 150 175 65 65 60 60 300 325 350 400 375 375 450 450

Radius 1 0.8 0.6 0.5 1 0.7 0.6 0.5 2.5 2.5 2.5 2.5 0.8 1 1 1 1.7 1.2 1.3 1.2

Threshold 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

USM settings for several sources of digital capture. The settings on this page assume that you're using a sharpening layer that uses an edge mask, reduced Opacity to something like 65%, and the appropriate Blend If settings from the previous page. The combination of layer mask, reduced opacity, and restrictive Blend If settings work together to keep the sharpening focused. Even with these rather substantial Amount settings, the resulting capture sharpening is still mild. The goal is to restore only the sharpness that was lost through the digital capture process and no more. We still have the possibility of creative sharpening to consider and a final round of output sharpening to apply. The sample files Capt'nNathanCaptureSharpened.psd and Capt'nNathanCaptureSharpenedPS3. psd contain separate layer groups for the generic Blend If settings and the capture source-specific settings from the table on the previous page. You might have noticed that the USM sharpening settings I used for this chapter were identical to those listed above for a Medium Resolution Digital Camera with a Narrow edge mask. This way you can easily compare the generic settings and the capture source-specific settings. These settings are the result of trial-and-error. I've found they work well for me as a starting point. You can reduce the sharpening (or increase it) by adjusting the Opacity setting for the sharpening layer. I rarely find that these setting need much adjustment, but feel free to experiment with them. There's nothing "magic" about these numbers.

199

Chapter 7

Creative Sharpening

Capture sharpening and output sharpening are a combination of science and art. Creative sharpening, in contrast, is all about bringing out your artistic vision. You take control over precisely where to apply additional sharpening (or blurring) and how much to apply. We're going to use the Photoshop Brush tool to apply sharpening and blurs with pinpoint precision. As Bruce Fraser notes in Real World Image Sharpening with Adobe Photoshop CS2, creative sharpening isn't required for every image. Commercial studio photographers, and wedding photographers, for example, often need to process hundreds of images over the space of a few days. There's just no time to be brushing in sharpening effects or blurs on every image when you need to prepare a wedding album. On the other hand, even the busy wedding photographer often has a few "special" images they want to emphasize in their wedding album, and some of them might benefit from some extra attention via creative sharpening. Capture sharpening is applied globally. We do use an edge or surface mask to avoid sharpening noise and Blend If sliders protect some pixels. That said, however, we don't make any localized selections by hand to limit capture sharpening. We're restoring what was lost during digital capture and that tends to affect the entire photo. Creative sharpening is all about localized changes. We can use Photoshop's collection of selection tools, work with the Brush tool, or manipulate channels to limit where sharpening gets applied. Creative sharpening tends to be hands-on.

Bringing more definition to areas of the photo. In this case, attention is on the tree leaves.

Multipass Sharpening

200

It's the Brush tool we typically turn to during creative sharpening, and that'll be the focus in this chapter. We'll save the selection tools like Color Range for later. There are just a handful of steps to apply effects locally with the Brush tool: 1. Make a Merge Visible duplicate layer. Always apply creative sharpening on a separate layer. My recommendation is that you set Opacity to 65% for editing headroom, the Blend If sliders to protect the extreme shadows and highlights as I described in chapter 3, and the Blend Mode for the new layer to Luminosity to prevent color shifts.

Tip A graphics tablet is helpful accessory for creative sharpening. You can easily obtain a smooth blending of sharpened pixels.

2. Sharpen the photo. This is key! Slightly oversharpen the feature(s) that you want to creatively sharpen. 3. Add a Hide All layer mask. The layer mask will be filled with black. The sharpening effect you just applied will completely disappear. 4. Set the Foreground to White. If you have not changed the Photoshop defaults, pressing the "d" key will make the foreground white and the background black. 5. Select the Brush tool. Set the brush's Size, Hardness, Opacity, Fill, and other options. 6. Paint the sharpening back into the photo with the Brush tool. If you're unfamiliar with the Brush tool and its many options, I strongly recommend Deke McClelland's Photoshop CS Bible: Professional Edition or Photoshop Artistry for Photographers Using Photoshop CS2 and Beyond by Barry Haynes, Wendy Crumpler, and Sean Duggan. I suggest you select a brush with a soft edge. Set Opacity low: something like 10-15% to start. That way you can brush in the sharpening with successive strokes. You have more control when you allow the effect to build up gradually. If you do get carried away, it's easy enough to reverse the effect. Just switch from painting with white temporarily and paint with black. Again, if you haven't changed the Photoshop defaults for foreground and background color, you can switch back and forth between white and black with the "d" and "x" keys.

201

Chapter 7

Bruce Fraser puts a lot of emphasis on using a Hide All layer mask. That works well when you want to add sharpening to some small, localized areas, and that's often the case. We decide a feature or two could benefit from a little more sharpening. Sometimes, however, we want to apply a good dose of sharpening to most or all of the photo and then reduce the added sharpening in just a few places. In that case, it's easier to use a Reveal All layer mask (which is filled with white) and then paint with black. Reducing the layer's Opacity setting to something like 65% is important in versions of Photoshop prior to CS3 because we still have a round of output sharpening to apply to the image. We know where we want to apply sharpening. That's why we're having a go at creative sharpening. We have an idea about how much sharpening we want to apply. The difference between pixels on the monitor and pixels in print complicates matters. The creative sharpening effect is going to be enhanced by the output sharpening. We need a way to adjust the creative sharpening later. With pre-CS3 versions of Photoshop, we're pretty much limited to the Opacity setting and the Blend If sliders. High Pass filter sharpening can be toned down or stepped up with a Soft Light or Hard Light blend instead of an Overlay blend. Photoshop CS3 gives us more flexibility to refine our creative sharpening later. When we use Smart Filters, we can easily adjust USM, High Pass Filter, and Smart Sharpen settings later. Even during a later editing session. I was interested in the trees, so I made sure it was visible in the thumbnail display for the Unsharp Mask filter when I applied capture sharpening (see page 200). I chose Amount = 135, Radius = 0.6, Threshold = 0, and I started with Brush Opacity set to 10%. Then I brushed in the sharpening effect. I followed the same precautions I would for a capture sharpening layer. Luminosity blend. Opacity at 65%. Same Blend If slider settings, too. We need to be mindful of the same concerns with creative sharpening and capture sharpening: a final round of output sharpening awaits, so leave some editing headroom. You can see the result in Capt'nNathanCreative Sharpened. psd in the \MultipassSharpening folder. There's also a CS3 version, if you want to try manipulating creative sharpening with Smart Filter layers. Creative sharpening is brushed onto the photo through a layer mask.

Multipass Sharpening

202

You can check and see where the creative sharpening has built up by turning off all of the underlying layers, including the Background layer. If you want to see the full size version of the layer mask instead of just the tiny thumbnail on the Layers palette, alt/option-clicking on the layer mask's thumbnail will toggle the layer mask on and off. You can see examples of both maneuvers on the next page. Viewing the layer mask can point out where the transitions between sharpened and unsharpened pixels are sharper than we might wish. There is always a trade-off between allowing the effect to build up with successive strokes and working with the fewest strokes possible. There are a couple of quick and easy ways to soften the transitions without pressing the "x" key and repainting parts of the selection with the brush tool. One is to click on the layer mask, which will make it active, and then apply a light Gaussian Blur. This has the effect of feathering all of the selections. The other quick solution is to grab the Blur tool and brush it over a few troublesome spots on the layer mask. Adjusting edge masks and alpha channels is one of the few occasions when I use the Blur tool. We can apply capture sharpening, using separate layers for the lighter pixels, the darker pixels, and surface textures. I do that sometimes. You need to weigh the additional creative control that comes from separating the sharpening effects for the light and dark contours against the increase in file size and the the increasing demands on system resources. While I typically sharpen the light and dark contours separately during capture sharpening, sometimes even adding a surface sharpening layer, I resort to separate sharpening for the light and dark contours less often with creative sharpening. I'm all for theory. Theory might incline us to separately sharpen light and dark contours for both creative sharpening and capture sharpening. We don't need to be a slave to theory, however. While I tend to obsess more than some with my photo editing (that's the nature of fine art photography, I guess), I don't waste time or system resources

Creative sharpening: The effect of brushing in additional sharpeness to the trees and the red roof.

203

Chapter 7

Hide All layer mask with the creative sharpening effect painted in. I alt-clicked on the thumbnail for the layer mask to make it visible rather than the photo.

Creative sharpening: The Background layer turned off, so only the sharpening remains visible.

Multipass Sharpening

204

needlessly. If I think I can bring out more detail by sharpening the dark contour more heavily than the light (or vice versa), I'll add separate sharpening layers for each. If I'm in doubt, I'll check quick and see if sharpening those two contours separately helps. Same with surface textures. If I think I can restore some of the sharpeness with a separate sharpening layer during capture sharpening, I'll give it a go and evaluate the results. At some point I have to make a judgment: is the proliferation of layers really giving me an appreciable advantage? If not, the additional layers go to the trash can icon. It's also possible to use an edge or surface mask to further restrict the creative sharpening effect. Applying a layer mask is not sufficient for painting in just edges or the surfaces. A layer has only one layer mask, and we can convert it to an edge mask or we can use a tool to paint sharpening in/out. To do both simultaneously, we need an extra step. We need to apply the edge mask. 1. Make a Merge Visible duplicate layer. Set Opacity = 65%, the Blend If sliders, and the blend to Luminosity. 2. Sharpen the photo. This is key! Slightly oversharpen the feature(s) that you want to creatively sharpen. 3. Add Reveal Selection edge layer mask.

4. Click on the Layer | Layer Mask | Apply submenu item. The layer will now be transparent where it was previously black. Since black conceals in a layer mask to protect them from changes, transparency will ensure those pixels will be unaffected and the underlying visible layer(s) will punch through. 5. Add a Hide All layer mask. Now you can use the Brush tool to paint in your creative sharpening, and what you paint in will be restricted to the mask you applied in step 4. When we apply a layer mask, we lose the ability to edit it later unless we step backwards through the History palette. That's why I don't delete the alpha channels for my edge and surface masks. I can always delete the layer and start over again with the alpha channel for my edge/surface mask. We can apply blur creatively in the same way as we can sharpen. We can use a Hide All layer mask and then paint the blur effect in. Or, we can use a Reveal All layer mask and paint it out. At the top of the next page is sample of street photography from the Campo Fiero market in Rome, Italy. The depth-of-field (DOF) was wide enough to allow a little too much of the background detail to be in focus. At least, that was my artistic judgment. I decided the sharp features of the woman would be even more dramatic if the background was less obvious. Photoshop CS added the Lens Blur filter to help with these sorts of DOF issues. It simulates photographic lens blur better than the humble Gaussian Blur filter. So you might be inclined to use it in this case, and you should with Photoshop CS, CS2, or CS3. But what if you're working with Photoshop 7 or earlier? There is no Lens Blur filter. You could easily add a creative blur layer instead. I cheated a bit. I didn't just use the Brush tool for this example. I'm much too impatient to try to brush in all of the fine details around the woman. I used the Magic Wand tool to jumpstart my layer mask. Then I worked with the Brush tool from there.

205

Chapter 7

The photo before creative blur ws applied.

Creative Blur applied to the photo. Even with three successive rounds of Gaussian Blur = 3.0, the effect is not so profound when layer's Opacity = 65% and the photo is downsampled.

Multipass Sharpening

206

Here's the layer mask. No blur to the woman or the trash bin she's leaing against. Progressively more blur is added to details further behind the woman.

The work with the Magic Wand took a couple of dozen dabs with the tool. I moved back and forth between Tolerance = 10 and Tolerance = 5 as I shift clicked the tool just outside the edge of the woman. I knew I'd have to clean up the selection, so I aimed to get nearly all of her outline intact. Then I switched to Quick Mask mode to refine the selection with the Brush tool. If you own Photoshop CS3, you can use the Quick Selection tool. You can isolate the woman in just a minute or two with Photoshop's newest selection tool. it's another reason to upgrade to CS3 You'll notice that the layer mask for creative blur is not just black and white. Black conceals, white reveals. That's the selection mantra for masks. However, if I applied the same blur to all of the background, that might appear artifical. For the benefit of readers with Photoshop 7 or earlier, I implemented a "poor man's" DOF blur. I blurred the remote background more and the near background less. I used white for the most remote background, L*a*b with Lightness = 25 for distant features, L*a*b with Lightness = 50% for closer features, and black for the woman and the trash bin. Since the woman and the trash bin were approximately the same distance, it would have looked strange to have her in focus and the trash bin blurred. If you do this with a photo where the subject is not at the front, like she is here, then remember the photographer's rule of thumb for depth of field is that it extends onethird in front of the subject and two-thirds behind the subject. You can see the result in RomanWoman.psd in the \MultipassSharpening folder. There's a Photoshop CS3 version that includes Smart Filter layers. If you toggle the creative blur layer on and off, you'll see that the blur is a subtle effect here. You have complete control over that. Subtle or go wild. The creative layers are all about enhancing your artistic vision. You can also restrict creative blurring to surfaces with a layer mask.It's usually a good idea to keep creative blurs away from the edges we've been laboring to sharpen. Just follow the same steps on page 205. Add a surface mask

207

Chapter 7

and apply it. Then add a new Hide All layer mask to paint in the blur effect. The final creative enhancement that I want to discuss is local contrast enhancement. It's typically applied to the entire photo. If local contrast enhancement is applied to the entire photo, isn't that an inconsistent statement?! Shouldn't it be called something like global contrast enhancement? The brief answer is, "No." Credit for this technique goes to Michael Reichmann, who in turn shares it with Thomas Knoll. The name for the technique was Michael's choosing, too. "Local" in this case refers to " . . . smaller adjacent areas in the image . . ." The technique aims at increasing the contrast in "small scale" differences, or what many people refer to as high frequency features. Local contrast enhancement is done with the Unsharp Mask filter. The technique sounds counter-intuitive at first. We use a high Radius setting together with a low Amount setting. Something like Amount = 20% and Radius = 50. The result is from such a combination is a general contrast boost. I've seen actions that use this same technique and refer to it as haze reduction. Local contrast enhancement is a good choice when a photo can benefit from haze reduction. You should read Michael's article. It's available on the Web. You can find it at this URL: http://www.luminous-landscape.com/tutorials/contrast-enhancement.shtml. While you're at his site, I suggest you look around. Michael's site has a lot of tutorials and reviews for the digital photographer. It's a great site to bookmark in your browser. What's missing from Michael's discussion is a more technical explanation for the effect. Dan Margulies discusses this technique in more detail, although he calls it HIRALOAM sharpening in his book Professional Photoshop, Fifth Edition. HIRALOAM is an acronym for HIgh RAdius LOw AMount. Dan explains it this way: coupling together large Radius and small Amount results is a generalized contrast boost because the large and dark sharpening contours begin to overlap.

Insight Another way to increase the apparent sharpness for a subject is to decrease the sharpness for a photo's competing features.

Multipass Sharpening

208

The test target from chapter 2 is the top image. Directly underneath is the same test target with USM 30, 50, 0. The bottom figure shows the affected pixels. Our test target from chapter 2 is helpful for understanding how local contrast enhancement works. Just keep in mind, the dark swath in the affected pixels are not necessarily pixels whose brightness is going down. Dark pixels are affected pixels. The more they are affected, the darker they become. The effect itself on a particular pixel could be an increase or decrease in brightness, depending on how sharpening affects that pixel. We can see that a whopping Radius setting does not result in stark sharpening halo, as it did in our examples in chapter 2, where Amount was held at 100 as Radius increased. The low Amount setting prevents an unattractive halo from emerging. Instead, we get a pleasing enhancement that can actually increase the perceived sharpness of smaller details in a photo. That's right. High Radius can actually increase the perceived sharpness of smaller details when Amount is kept low. Ratchet up the Amount setting, though, and that same Radius will obliterate small details and give us really ugly sharpening halos. I agree with Michael Reichmann's general advice. I find this technique to be so helpful, I use it on nearly every photo I retouch. You can try different settings. I find that Amount 20% to 30% works well; couple that with Radius between 30 and 60. Let's look back at our photo of the sponge boats at Tarpon Springs, FL. I used USM Amount = 30, Radius = 50, Threshold = 5 for the local contrast enhancement. The Threshold setting keeps noise from emerging in the sky, since we're applying the effect to the entire photo. I also set Opacity = 65%, Luminosity blend, and the Blend If sliders for single layer multipass sharpening to protect the extreme highlights and shadows. Before we move long to a discussion of output sharpening, there are a couple of creative sharpening caveats to consider:

209

Chapter 7

Creative sharpening applied. Local contrast enhancement not yet applied.

The result from local contrast enhancement. USM 30, 50, 5.

Multipass Sharpening

210

·

Don't get carried away. The monitor is an imperfect guide to judge prints. If your photo looks over-sharpened at the creative sharpening stage, it will almost certainly look oversharpened after output sharpening is applied. Work on the image at its native resolution. You might your photo as a 4" x 6" print today. Then someone asks for an 8" x 12" or a 12" x 18" print. If you do your resampling after creative sharpening, you end up with a master image that can be reused for different size prints. I always save a master file that includes noise reduction, capture sharpening, edits to tone and color, and creative sharpening. Then, when I get ready to print, I resample (if necessary) and apply sharpening appropriate to that print size.

·

Output Sharpening

Many digital photographers undersharpen their printed photos. They use the monitor as their guide. As soon as the sharpening halos become even slightly noticeable, they pull back on the sharpening settings.

Insight Sharpening for printer output will look oversharpened on your monitor. You need to hard proof or you'll under-sharpen your prints.

You can use your monitor as a sharpening guide when you're preparing photos for display on the Web or some other multimedia display on a computer. That's it for the monitor as a good guide, however. If you want optimally sharp prints, you need to print samples and examine the results with a critical eye. Bruce Fraser claims that output sharpening is much more "by the numbers" than capture sharpening or creative sharpening. He comes to that conclusion because the relationship between image pixels and output dots is fixed. Once you figure out how a device converts image pixels into output dots, according to Bruce's argument, you can devise optimal sharpening settings. Real world sharpening is more complicated. If you read Real World Image Sharpening with Adobe Photoshop CS, you'll notice that Bruce I doesn't use a mathematical formula to arrive at any of his output sharpening settings. If it really was as mathematical as Bruce implies, he would mapped those pixels with some sort of formula. When he describes the source of his own output settings, he describes it as trial-and-error. I'm not disparaging the settings that Bruce offers readers. Mine are trial-and-error, too. I invested man-months in the sharpening

211

Chapter 7

settings yu find in this eBook, and I continue to invest more time in their refinement. I'm sure Bruce Fraser put a lot of time and effort into his settings, too. I wish output sharpening was a simple as a mathematical formula. It would have saved me a lot of time. Optimal output sharpening requires more than a set of "magic numbers" for the Photoshop sharpening tools. You need to start with an image that not ready to "fall apart" once additional sharpening is applied. Noise needs to be considered and handled properly. Capture sharpening needs to be gentle and just enough to restore sharpness lost during digital capture. Creative sharpening needs to be done with an awareness that additional sharpening is still to be applied. It gets more complicated. Output sharpening needs to consider the output device, the paper type, ink type, output resolution, and viewing distance. Yes, you can quantify all of that into an equation. But just what is the equation for a 12" x 18" print from Epson 9600 printed at 360 dpi on Epson Radiant White Watercolor Paper using the OEM Epson ink to be viewed in a gallery with tungsten lighting. I don't know! I don't know anyone who does know. Change just one of those variables, like using Arches Infinity smoothi finish fine art paper, and what's the new equation? Bruce assumes away viewing distance. He talks about "reasonable" viewing distance. Well, that's not all that helpful. You can print a 4" x 6" print at 300 dpi and view it at arm's length or a little less. You print the same image on the same printer at 300 dpi on the same paper at 12" x 18" and view it at a few feet. The same sharpening setting will almost certainly not be optimal for both because sharpening halos that are noticeable on a 12" x 18" print at "reasonable" viewing distance will be too wide for a 4" x 6" print in a photo album. The fixed relationship between image pixels and dots on a print makes output sharpening "repeatable." I put double quotes around repeatable because we can come close. If we're consistent about our capture sharpening and creative sharpening, we can develop some candidate settings for our preferred printer and the media we prefer and those settings can give us predictable results for a given resolution and print size. Notice, I said "candidate settings." They are not the magic numbers that Bruce Fraser and other authors publish. There's nothing magic about these numbers I'm sharing with you. They're trial-and-error. They work well for me. But "work well" doesn't mean they're opti-

Insight Don't pull back on output sharpening just as halos emerge. You need sharpening halos at this point to give your photo more apparent sharpness.

Multipass Sharpening

212

mal for every photo I print. I use them unchanged for many of my prints. For othes, I finesse the settings. Use these numbers as starting points for developing your own settings. Output sharpening is all about getting the sharpening halos to have the final width and definition we need for enhancing the apparent sharpness of our photograph. It compensates for final image size, resolution, viewing distance, and output device. Output sharpening makes two big assumptions: · Sharpening for content is done. We no longer worry about high frequency v. low frequency details. Creative decisions that affect sharpening are also behind us. The image is at its final resolution. Any resampling has already been done.

·

I am extremely reluctant to upsample images. I have read all sort of reports from people who take digicam images and make poster size prints and claim there are no visible artifacts. I guess it depends on how experienced your eye is or how far back you stand. I've rarely been happy with upsampled photos. I only go there when I have no choice. Downsampling is less of a concern.

Blend If settings for single output layer. Settings for separate output sharpening of the light and dark halo contours canbe found on the next page.

213

Chapter 7

Multipass Sharpening

214

Output Sharpening Techniques

I use two different techiques for output sharpening: · · USM Sharpening. USM sharpening works well for halftone printing. USM followed by High Pass. USM and High Pass perform sharpening differently. Start with USM, fade it with a Luminosity Blend, set the blend mode to Overlay, and then apply High Pass sharpening. This results in halo details that inkjet and contone printing can preserve. It works well when sharpening for the Web, too.

You don't need an edge mask for output sharpening. Without an edge mask, you will likely need to use the USM Threshold setting to reduce or eliminate the sharpening of any noise, however, since it's common for noise that was barely noticeable in the unsharpened original to become very evident at the output stage: especially if you didn't use any noise reduction. I find that Threshold = 5 or something close to that will reduce light noise. The Blend If slider settings are broader for output sharpening. The settings for output sharpening with a single layer appear on page 213. Nearly all pixels from the underlying layer are eligible for sharpening. Only the very brightest are excluded. The extreme shadows and highlights are protected from sharpening. Blend If settings for separate light and dark contours appear on page 214. Pixels lighter/darker than 90 are sharpened separately. If you followed my suggestions and you reduced the Opacity for your sharpening layers to something like 65%, you have some control over the fial sharpening result. That way, if you find that capture or creative sharpening was too aggressive, you can reduce their effect of those passes by reducing their Opacity setting(s). Photoshop CS3 users have another option, if they use the new Smart Filters option for their sharpening layers. They can go right back and adjust the individual sharpening settings.

Halftone Printing

Let's start with the simpler case: halftone printing. Simpler in the sense that we can use USM adorned only with the Blend If settings above and a Luminosity blend for the sharpening layer. I also recommend setting Opacity to something like 65%. We have three factors to consider: paper type, screen frequency, and screen multiplier. Ink bleeds more on uncoated paper, so a slightly higher USM Amount setting is usually required. Screen frequency and screen multiplier are press-dependent. Your printer will need to specify these. You'll notice in the settings below that a smaller screen frequency typically gets a smaller Radius and larger Amount setting; larger screen frequency gets the opposite, larger Radius and smaller Amount settings. Candidate settings to get you started with halftone printing can be found on page 216:

215

Chapter 7

Output sharpening applied: 150 lpi halftone screen, 2x multiplier, coated paper. Frequency 85 lpi 85 lpi 85 lpi 85 lpi 120 lpi 120 lpi 120 lpi 120 lpi 133 lpi 133 lpi 133 lpi 133 lpi 150 lpi 150 lpi 150 lpi 150 lpi 175 lpi 175 lpi 175 lpi 175 lpi Multiplier 1.5 1.5 2.0 2.0 1.5 1.5 2.0 2.0 1.5 1.5 2.0 2.0 1.5 1.5 2.0 2.0 1.5 1.5 2.0 2.0 Paper Type Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Coated Uncoated Amount 205 220 205 220 210 225 200 215 200 215 175 190 165 180 135 150 185 200 165 180 Radius 1.0 1.0 1.0 1.0 1.2 1.2 1.3 1.3 1.3 1.3 1.3 1.3 1.7 1.7 1.5 1.5 1.7 1.7 1.7 1.7 Threshold 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

Multipass Sharpening

216

Since this is an eBook, I have no idea how you are viewing the pages. Most users, my guess, will view the eBook on a monitor. Some will prefer to print the book. (I confess, I tend to prefer hard copies to scrolling pages on my monitor.) The halftone output sharpening example will look very different, depending on how you view the eBook. You have the sample image to experiment with. For the example, I used the settings for 150 lpi, 2x multiplier, coated paper. USM with Amount = 135, Radius = 1.5, Threshold = 5. If you're viewing this eBook on a monitor, the example is going to look rather scary. Now, tell the truth! Would you have pushed the image to this point, if you were sharpening by eye and using your monitor as a guide?! Nah! The typical digital photographer stops well before they reach this point. We actually want visible halos by the time we apply output sharpening. Sharpening halos is where the increase in perceived sharpness comes from. I know I'm flogging the heck out of this horse, but this bears repetition. Output sharpening for prints will look oversharpened when you view the photo on a monitor. If you want to experiment with this photo, it's named Capt'nNathanUptputSharpened.psd, and you can find it in the \MultipassSharpening folder. There's also a CS3 version for you refine the settings. You'll see that I sharpened the light and dark contours separately. That's my work habit. It gives me options once I hard proof.

Inkjet Printing

Most of the readers of this eBook will do most or all of their printing on a desktop inkjet printer. Inket printers don't use halftone dots. They convert pixels to printed dots in a different way. The technical term is error-diffusion screening. I mention it only to show off my own technical knowledge and so you have something to enter in a Web search engine (if you want more details about how inkjet printing works). The practical consequence for our discussion is this: we need slightly different USM settings for inkjet sharpening and we need to follow that up with some High Pass filter sharpening, too. There are several steps to inkjet output sharpening. I use the same steps for contone output shapening and for Web/ multimedia sharpening, too. It's just the sharpening settings that vary. Here are the steps: 1. Make a Merge Visible duplicate layer. Set Opacity = 65%, the Blend If sliders appropriate for output sharpening, and the blend to Normal. 2. Apply USM sharpening. 3. Immediately click on the Edit | Fade . . . submenu item. Fade the sharpening to 65% Luminosity blend. This is the critical step. You need to do ths immediately after you close the USM dialog. Do anything else and you lose the option to fade the result. 4. Change the layer blend to Overlay. Now we're getting prepped for some High Pass sharpening. 5. Apply High Pass filter sharpening. This will build up some extra contrast in our sharpening halos.

217

Chapter 7

Output sharpening applied: 300 dpi inkjet, glossy paper.

We have fewer factors to consider for inkjet sharpening. The candidate settings below are adjusted only for print resolution and paper type. Don't let the high Amount settings scare you! Remember, we have a Fade maneuver half-way through the sharpening. The difference between halftone sharpening and inkjet sharpening isn't all that great. The halos for inkjet sharpening will have more contrast. The way that inkjet printers convert pixels into dots means that inkjet prints can reproduce some finer distinctions in the sharpening halos more precisely compared to halftone printers. As a consequence, inkjet prints can benefit from a little extra contrast along their sharpening halos and that's exactly what the added application of High Pass filter sharpening provides. Resolution 180 180 240 240 300 300 360 360 400 400 Paper Type Matte Glossy Matte Glossy Matte Glossy Matte Glossy Matte Glossy Amount 290 275 315 300 340 325 365 350 390 375 Radius 1.0 1.0 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 Threshold 5 5 5 5 5 5 5 5 5 5 High Pass 2.2 2.2 2.1 2.1 2.0 2.0 1.8 1.8 1.7 1.7

Multipass Sharpening

218

Contone Printing

Some digital photographers prefer to have prints made for them on a continuous tone printer, like the Fuji Frontier or the Durst Lambda printers. My candidate settings use the same technique as inkjet printing, just different settings. We only concern ourselves with resolution, too. Things are getting a little simpler as we proceed! I doubt anyone will print this book with a contone printer. So it doesn't really make sense to include a sample of the Capt'N Nathan with contone output sharpening settings. You've seen enough scary examples to make the point that sharpening for print looks raher ugly on the screen. Let's move along to sharpening for the Web.

Resolution 80 150 200 267 300 400

Amount 250 275 325 325 350 350

Radius 0.8 0.8 0.7 0.7 0.6 0.6

Threshold 5 5 5 5 5 5

High Pass 1.8 1.8 1.9 1.8 1.5 1.6

Sharpening for the Web

When you sharpen photos for display on the Web, you can let your monitor be your guide. Just keep in mind, it's still not a perfect guide. Monitors do differ. If your photos are destined for a Web page, it's best to use monitor settings that are closer to those of the average viewer, somewhere between 75 and 100 dpi. The candidate settings I'm going to offer are designed to sharpen downsampled photos for display on Web pages. They start at 1024 pixels along the longest edge and go all the way down to 100 pixel thumbnail size images. OK, a rather big thumbnail. We'll see that the 100 pixel thumbnail needs some extra aggression in its sharpening settings. You have to work to bring out detail when you make an extreme downsample. Resolution 100 250 480 640 800 1024 Amount 250 115 115 125 135 150 Radius 3.0 1.6 1.5 1.2 1.0 0.8 Threshold 5 5 5 5 5 5 High Pass 3.5 1.3 1.2 1.0 0.8 0.7

219

Chapter 7

Using Adobe Camera Raw 4.1 for Capture Sharpening

Photoshop CS3 includes a new version of Adobe Camera RAW (ACR) for processing the "raw" files from many digital SLRs. ACR 4.1 is not compatible with earlier versions of Photoshop, but if you own Photoshop CS3 (or Photoshop Elements 5.0) you can consider using ACR for your capture sharpening. Earlier versions of ACR were not up to the task, in my opinion. You could not control the sharpening sufficiently. ACR 4.1 upgrades the sharpening controls, adding one called Clarity, and has much improved masking. While you'll have more control over the capture sharpening if you do it yourself once the photo is in Photoshop, you can consider ACR 4.1 as a worthy candidate for capture sharpening. You need to weigh the ease of use in ACR 4.1 against the added control from using USM and edge/surface masks. The problem with using ACR 4.1 for capture sharpening is that is very difficult to see the capture sharpening effect. You can hold down the alt/option key while your drag several of the sliders and the display will change to give you some helpful previews. But I find the 100% view that ACR 4.1 requires to view the effects of the control to be less than helpful with large photos. When you capture sharpen using layers, you can toggle the layers off and on. I find this very helpful. It helps me judge whether I have restored the sharpening lost during during capture and not overdone the sharpening. ACR 4.1 let's you toggle the preview off and on, but you have to be careful. Once you load the image into Photoshop, those settings become your original settings for the Preview toggle. If you want the preview to go back to the RAW file, you'll need to zero the sliders, click Done, and then load it again into ACR.

There are more options on the Sharpening tab in ACR 4.1. Hold down the alt/option key while moving the sliders produces helpful previews.

Multipass Sharpening

220

You also lose the capability the USM dialog provides for boosting the magnification and measuring the sharpening halo width. You can boost the magnification to 400% from the ACR 4.1 dialog (half what you can do with the USM dialog). The Masking control is really nice. I wish Photoshop offered it as a separate filter. With a setting of 0, you get a white mask. All pixels are sharpened. Drag the slider to 100 and you get a nice edge mask that's pretty much the equivalent of a medium width edge mask. Between 0 and 100, you sharpen more or fewer surface details. The Clarity slider is on the Basic tab instead of the Sharpening tab. What Clarity does is implement Michael Reichmann's local contrast enhancement: USM with low Amount and high Radius. I like having LCE available in a RAW processor, since I apply it creatively to nearly every image. This saves me the overhead of a layer. You control the sharpening effect with the Amount, Radius, and Detail sliders. These function similar to the USM settings, although the numeric values for Amount and Radius are more restricted and Detail is not the equivalent of Threshold. Amount goes to 150, although you'll find that 150 for ACR is a stronger sharpening effect than it is with USM. Radius goes to 3. The sharpening in ACR 4.1 is intended to be capture sharpening, not output sharpening. That probably explains the throttle on Radius, but you can still hideously oversharpen with ACR 4.1. The Detail tab is helpful for bringing out smaller image features. Jeff Schewe claims that a value of zero applies an extreme dampening to the sharpening halo. This is one slider where it helps a lot to hold down the alt/ option key to see the preview as you adjust it. With a little practice, you can obtain satisfactory results for capture sharpening. I start with the Masking slider. Then I use the USM Radius settings from this chapter. I watch the fine detials as I adjust the Detail slider. The Amount slider gets pretty aggressive over 100.

The Masking slider is really nice. At 0, you sharpen every pixel. Pull the slider to 100 and you get a medium width edge mask.

221

Chapter 7

My candidate value for the Capt'N Nathan photo were Amount = 55, Radius = 0.6, Detail = 90, and Masking = 20. One other recommendation: when you open a photo, use Open as Smart Object... It is always best to leave yourelf with flexibility when you apply capture sharpening. When you use layers, you can adjust the layer's Opacity setting or you can use Smart Filters with Photoshop CS3. Opening a photo from ACR as a Smart Object gives you the option of revising the settings later. The noise reduction in ACR 4.1 is also improved over previous versions. I still cannot recommend it as a replacement for dedicated plug-ins like Noise Ninja and NeatImage. The surface masking is also limited. You can sometimes see the transition along edges between pixels with noise reduction applied and those without. To be fair, Thomas Knoll and the team responsible for ACR 4.1 didn't intend the noise reduction as a replacement for thirdparty noise reduction plug-ins.

It's easier to judge the effect of the Detail slider by holding down the alt/option key. I used it to watch the details in the trees.

Multipass Sharpening

222

The preview for the Amount slider is in grayscale when you hold down the alt/option key.

The preview for the Radius slider looks more like High Pass filter sharpening when you hold down the alt/option key, highlighting the effect of the Radius setting.

223

Chapter 7

Multipass Sharpening

224

Information

Sharpening Your Photographs Chapter 7.indd

42 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

862880


Notice: fwrite(): send of 204 bytes failed with errno=104 Connection reset by peer in /home/readbag.com/web/sphinxapi.php on line 531