Archive for November, 2007

Photoshop Tip: Two Pass Sharpening Workflow

Tuesday, November 27th, 2007
Nearly all digital images need sharpening to some degree, and I apply various techniques and amounts of sharpening throughout my workflow. The most important sharpening stage, so that the final image looks crisp and focused, is done as a final step to the image once it has been sized for its intended use. When images are downsized they lose sharpness, as detail along edges is lost in the reducing process. When images are enlarged, the lack of sharpness inherent in digital images is magnified. Using Photoshop’s digital sharpening filters makes it possible to increase contrast along fine edges to bring back lost image sharpness. There are many ways to do this, but I have recently developed a technique that I think does quite well. It is important to note that digital sharpening is not a substitute for lack of sharpness in the original image. Proper focus, shutter speed and technique is required at the point of capture to enable good digital sharpening.First, open an image in Photoshop. Then, flatten all layers (if any exist) and size the image for its intended use (print, email, screen saver, etc.). For help on how to size images you can refer to my articles on SIZING FOR PRINTING or SIZING FOR EMAIL. Once the image has been sized it is time to apply sharpening to make it look appropriately sharp. The technique I have developed makes two sharpening passes. The first pass uses the Unsharp Mask filter set to add a small degree of sharpening to larger edges and details to help with the overall crispness and “pop” in the image. The second pass uses the Smart Mask filter set to add a greater amount of contrast to the very finest edges and lines in the image for razor sharp fine detail. Versions of Photoshop prior to CS2 don’t include the Smart Sharpen filter, so the second pass can also be made with the Unsharp Mask filter with good results.

Make sure that you are viewing the image at 100%, or, if it is a larger sized print, 50%. To complete the first, large detail sharpening pass go to Filter>Sharpen>Unsharp Mask. Select a Threshold of 1, a radius of .7 to 1.0 for web images and 1.5 to 2.5 for prints, and an amount in the 50 to 75 % range. I find it helpful to check and uncheck the preview box to dial in the right amount of sharpening. Select OK when you are finished. Now to sharpen the fine edges and details, go to Filter>Sharpen>Smart Sharpen (or Unsharp Mask again if your version of PS is pre CS2). Set the Remove box to Lens Blur, the radius to 0.1 for screen and web size images or .2 to .4 for prints, depending on how large the print will be, and finally, adjust the amount to 100 to 140% or until the desired sharpness is achieved.

I find that this two-pass sharpening technique does a great job of targeting sharpening to both the larger and finer details separately, especially for images that will be displayed on a screen. As with any sharpening technique, knowing how far to go without going too far is the key and some experimentation is needed to determine the best combination of adjustments for each individual image. Below are samples of a screen sized image. The first is not sharpened, the second is sharpened appropriately using this technique and the final image is over sharpened in my opinion.

sharp sample

Photoshop Tips: Sizing Photos For Printing

Tuesday, November 20th, 2007

Sizing digital images for various uses may be one of the most puzzling problems that the average digital photographer runs across these days. Do your eyes glaze over when trying to understand terms like resolution, pixels and dpi? Have you ever sent an image of your dog in an email that is so large that it fills up the recipient’s email in-box? Have you ever had an image enlarged and printed only to find out that the file size was too small and the printed photo is so pixilated that it looks like a graphic from an old Atari video game? If you answered yes to any of these questions then read on. Hopefully I can shed a little light on the matter.

In a previous article I provided some pointers for sizing images for email and the web. You can check out those tips by clicking HERE

Sizing images for printing is very different than sizing images for a computer screen. A concept that is difficult to wrap the mind around is the fact that the size you see an image on your screen is not the size at which it will print, unless you print at screen resolution, which generally does not have enough detail for good quality prints. For example, if I size an image to 4″ by 6″ inches at 80 dpi (dots per inch) it measures 4″x6″ on my monitor (this will be different from monitor to monitor depending on what the resolution of the monitor is set at). Screen resolution is usually between 72 and 96 dpi. At 80 dpi the 4″x6″ image looks great on my screen. When I print it, the print does indeed measure 4″x6″, but with only 80 dpi it is very blocky and pixilated. In order to get a detailed 4″x6″ print I would need to set the dpi much higher. When viewed on my 19″ screen, a 4″x6″ image at 300 dpi is too large to fit, even though it still prints as a 4″x6″ print (with much greater detail than the 4×6 @ 80 dpi). So, as you can see, how big an image appears on your screen does not necessarily indicate how large a print it will make.

Most cameras come with software that provide some help with sizing. However, this software is often oversimplified, giving the user very little control over the size and resolution of the final print. I prefer to use Photoshop, or some other advanced image editing software to size my images so that I have the most control. However, more contol also means more complexity, not something everyone wants. However you size your images, the requirements for getting good results remain the same.

Different methods of printing require different amounts of resolution detail, or dots per inch. If you don’t care that your print has jagged edges and visible pixels, then go ahead and print at 72 dpi. Some tests indicate that 150 dpi to be the minimum for good print results, but there are many factors that come in to play. For chemical process printing (non ink jet printing) 200 dpi is generally plenty of resolution. Most publishers (magazines, newspapers, etc.) have their printing standardized to 300 dpi and want all submissions at this resolution. The current generation of high quality ink jet photo printers are able to resolve detail in the 300 to 360 dpi range, although tests have shown that 240 dpi is almost indistinguishable and that 200 dpi gives great results. Knowing how the image will be printed will help you determine what dpi resolution to select when sizing the image.

Another consideration in selecting the dpi resolution at which to print has to do with the distance the print will be viewed at. Small prints that will be viewed close up benefit from higher print resolutions (240-300 dpi), while large prints that will be viewed from a few feet away can make due with lower print resolutions (150-200 dpi). Billboards that are viewed from hundreds of feet away are printed at 72 dpi and lower.

Once you know what dpi resolution you will select, the next thing to do before sizing is determine the native size of the image; in other words, the size at which it came out of the camera. In Photoshop it is easy to determine this. Open an image and then go to Image>Image Size. A dialog box will open that indicates the pixel dimensions as well as the dimensions in inches and the dpi (dots per inch). Next to the pixel dimensions it also indicates the total pixel count in thousands (k). It takes one million pixels (1000 k) to equal one megapixel. If the number in the dpi box already matches your desired print resolution, simply look at the inch measurements and they will show the height and width that the image will print. If the dpi is not correct for your desired print method, make sure that the “Resample Image” box is un-checked and then change the dpi to what you want. Once the dpi matches your desired print resolution you can now look at the image dimensions in inches to determine how large it is. For example, if you set the dpi to 200 and the dimensions in inches are 10 inches by 15 inches, the file will make a 10″x15″ print without any changes to its size. If you want to make a print that is 10″x15″ or smaller, then you are in good shape. Downsizing an image doesn’t create any resolution problems. Let’s say you want to make an 8″x12″ print. Simply re-check the “Resample Image” box, leave the dpi at 200 and then type 12 into the box that currently indicates 15 inches. If the “Constrain Proportions” box is also checked (which it should be) then the box indicating 10 inches will automatically change to 8. Click “OK” and your image will automatically be sized to print at 8″x12″ at 200 dpi. Remember not to save the image now unless you do so with a different name. Otherwise you will have permanently saved the image at the smaller size.

Now, let’s suppose that you want to enlarge the image so it prints bigger than its native size of 10″x15″ at 200 dpi. When enlarging, or upsizing, images you need to be a little more careful than when downsizing. To downsize, the software can simply remove pixels to shrink the image, which will not degrade how the smaller image appears to your eye. In order to enlarge an image, the software must add pixels. The act of removing or adding pixels is called interpolation. When interpolating an enlargement, the program uses complex math to decide how best to add pixels to make the image bigger. The current generation of interpolation software contained in Photoshop or third party applications, such as Genuine Fractals, is amazingly good at doing this. Lots of pixels can be added to enlarge an image without a noticeable decline in image quality. However, the computer can not add detail that wasn’t recorded by the camera, so at some point an enlargement will begin to show digital artifacts left by the computer doing its best to add large numbers of new pixels.

To enlarge an image in Photoshop, follow the same steps as when downsizing. Open the image in Photoshop and go to Image>Image Size. Uncheck the “Resample Image” box and enter the desired dpi resolution. This will show you how big the image will print at that dpi without any change in size. Using our previous hypothetic example, the image will print at 10″x15″ at 200 dpi. If you wish to print the image at 16″x24″ at 200 dpi, you know you will need to use interpolation to enlarge the print. As before, simply recheck the “Resample Image” box and change 10 inches to 16 inches. With the proportions constrained, the 15 inch box will automatically change to 24 inches. Click OK and Photoshop will enlarge the image accordingly.

How much can you enlarge an image and still maintain fine image detail? The answer is largely dependant on your own personal taste, the material the image is being printed on and the distance at which the image will be viewed. I have enlarged some 8×12 inch @ 300dpi images to 20×30 inches @ 300dpi with very good results, especially when viewed at normal viewing distance. The master print lab that I use has a very handy chart that helps visualize just how much various sized image files can be enlarged with good results. The chart can be viewed by clicking HERE.

So far, all of the sizing examples I have given have assumed that the image is being enlarged or reduced while maintaining the same dimensional proportions of the original. For example, an 8″x12″ image will perfectly size to 4″x6″, 10″x15″ or 12″x18″ because the proportions have not changed. However, it will not cleanly size to 5″x7″, 8″x10″ or 11″x14″ because the proportions are different. In order to size an image for printing and also crop the image to new proportions at the same time, use the Crop tool. Open an image in Photoshop and then click the Crop tool on the tools palette (or press “c” on the keyboard). With the Crop tool selected some cropping control boxes will appear below the menu tabs at the top of the page. Here you can enter your desired print dimensions and dpi resolution. If you wish to print an image at 8″x10″ at 300 dpi, simply enter those figures in the appropriate boxes. Then, move the crop tool to one corner of the image and drag to select the area that will be cropped. Once you have created the cropping area, you can use your mouse to move it around and fine tune your selection. You will notice that because you are changing the proportions of the image, some of the image will be cropped out. Once you are happy with your cropping selection, double click inside the selection or press the Return key and the image will be cropped and sized to your specifications.

Photoshop Tips: Sizing Images For Email

Monday, November 19th, 2007

One of the great advantages of digital photography is the ability to easily share your photos with others. In addition to making traditional style prints, digital images can be posted on websites, uploaded to digital personal organizers, displayed on cell phones and emailed. In my business I email photos on a regular basis. I email photo submissions to publishers, proofs to commercial clients and archived image samples to stock photo buyers. But it isn’t just professional photographers who email lots of images. I also email shots of my kids to the grandparents and photos I took out skiing to my friends. Emailing digital images is most likely the number one thing that people using digital cameras do with their photos and yet many people don’t have a clue how to optimize their photos for best email performance.

The most important consideration when emailing photos is image size, especially for those of us with dial-up Internet connections, but also for those with broadband. If you set your camera to take images at a size that’s right for email then they are too small to make good prints, but if you try to email full size images they take a long time to send and don’t even fit on the screen for viewing. The solution is resizing. Most cameras these days come with software that can easily help you resize your images for email. It is also easy to resize images in image editing software such as Adobe Photoshop. However you go about sizing your photos, here are a few tips that I follow to make sure that I’m not the guy emailing photos so large that they overfill inboxes, take several hours to download and can’t be viewed without scrolling around. These tips only apply to photos that are intended to be viewed on screen but not printed. Print size files need to be large to give good print quality.

First, I like to size my photos so they are about 500 pixels on the longest side and have a resolution of 72 DPI (dots per inch). This enables them to be viewed on smaller monitors and even within the email message window without opening them full screen. 72 DPI is important because that is about the maximum resolution that a monitor can show. Images for print often need 300 DPI or more, but this is a lot of extra information that doesn’t make the image look any better on the screen. A 300 DPI image that is 500 pixels on its longest side contains 16 times more data than the same size image at 72 DPI.

Second, I save the resized images to be emailed with a different name in a separate email folder. This way I don’t change the original image file and I can safely delete the resized images from the email folder when I don’t need them any more without accidentally losing the originals.

Third, I save the email images as JPEG (.jpg) files and set the quality to medium (about 5 on the Photoshop scale) to compress the files even more. These steps ensure that I am emailing the smallest files possible and that they send quickly, don’t fill up all the space in someone’s inbox and can easily be viewed on screen.

One more tip if you are using software like Photoshop is to apply unsharp mask (Filter>Sharpen>Unsharp Mask) to the final sized image before saving and sending it. When you downsize an image it loses some detail that can be brought back with a little sharpening. Good luck and happy emailing.

Getting Our Images To Match Our Vision

Thursday, November 15th, 2007
Donna, a fellow photography enthusiast recently emailed me for some suggestions on two of her photos. Both photos clearly show a strong sense of purpose and concept, but she felt that the results did not match what she envisioned when she took them. Her questions reminded me of my own struggles to create images that live up to my vision, percpetion and experience. I thought the content of our diologue might prove helpful to other photographers who are inevitably confronted with the same challenges.
Hi Sean,
I couldn’t resist sending you a couple of photos that I have taken recently. I am learning to try to see things from a different perspective.

donna1

The first one is so you can tell me how I could have made it better. It was such an interesting contrast. The bright autumn tree standing at the far edge of the old, subdued cemetery. I was using my little Fuji and it just doesn’t do as well with dim light. I took some shots up closer to the tree, but I like this one best. It just didn’t quite come off as I had hoped.

donna2

The other is by the Lithia fountain in the park. It was shot hastily when I was there with guests. I wish you had been with me on both occasions, I think you would have come up with some really good versions of what I was attempting.Thanks for your help.

Donna

Hi Donna,

I see what you are going for with both shots and both are very worthy subjects. You obviously have an eye for elements that make an interesting image. In the first there is the juxtaposition of color and mood, and a sense of perspective that draws the eye along the headstones, through the cemetery to the orange tree. In the second I can almost feel the warm glowing light transmitting through the trees and bathing the autumn scene in soft, magical tones.

I think you did particularly well with both of these images for the circumstances. There are a couple of composition choices that I might have made differently, but that is hard to say without actually being there. For example, in the cemetery image, it feels like there is too much empty space in the lower left corner. My instinct would be to move a bit and to adjust the composition, but then the headstones might not line up or the tree might lose its central focus. It is hard to know without being there.

I think most of what isn’t meeting your expectations in these images are functions of lighting, atmosphere, equipment, technique and post processing. The same cemetery shot taken at sunrise on a misty morning, with warm light casting long rays and shadows across the scene and lighting the tree would give it some of the drama you were perhaps feeling when you were there. A dense foggy day would give another mood, but also add to the image.

I have made many trips to, and shot hundreds of photos of, the Rogue Gorge in a variety of conditions. I was never happy with a single photo, even though I felt there was something there worth capturing. So, I kept going back. Finally I happened to be there on an amazing morning last month when the elements all came together and I made a photo that captured something of what I had been visualizing. It took me five years to get that photo and I will probably never again be able to get such a shot (although there is always the chance I’ll experience something different but just as good). I really can’t overstress the value of timing, perseverance and luck in outdoor photography.

Crimson Gorge

The equipment, techniques and post processing used in creating a photo also have a great deal to do with getting a shot to match your vision. The same Rogue Gorge shot wouldn’t have been possible with my handheld pocket camera. I was using a camera with a large, high resolution sensor and a high quality lens. The exposure lasted for several seconds so I was shooting on a tripod. I used a polarizing filter to help cut the glare off the water and the wet rocks and a graduated filter to try to hold back the exposure of the sky. I also took two different exposures and blended them in Photoshop because the sky was so much brighter than the dark foreground that the camera couldn’t record the entire range of light in one shot even with the graduated filter. All told, including learning the techniques, I may have spent hundreds of hours creating that one shot.

So, I guess what I’m getting at here is that I think you are completely on the right track with your photos. Keep looking, seeing and photographing things that get your attention when you find them. This is how you develop your personal vision. I think developing that vision is the most important step, but it is only one step in getting your photos to match the vision. The other steps are a little harder to pin down. Learning the camera and processing techniques takes time and practice, getting the equipment takes money and finding the right conditions takes perseverance, patience and luck. In the end, however, it is the vision that matters the most. The best part is you don’t even need a camera to explore your vision. I’m always visualizing photos in my mind, as I drive or walk down the street. I often hike through the woods without taking photos, instead holding up my fingers to help me frame and visualize my “virtual photographs”.

I think every photographer is constantly struggling to create photos that best express his or her inward impression of the experiences he or she has with the surroundings, but I don’t think anyone fully achieves that. I’m content knowing that my images are incrementally getting closer to my vision, but that I’m not in any danger of putting away my camera because I have done all there is to do.

Best,

Sean

Luminosity Masking

Saturday, November 10th, 2007

Apparently luminosity is a term that is used incorrectly by Adobe Photoshop and photographers in general. I won’t go into the correct astronomy definition of the term. However, the inaccurate concept of luminosity as it pertains to photography is still particularly useful. In photography, luminosity is considered to be the measure of the density of luminous intensity coming from or through a surface. In other words it is an indicator of how bright a surface will appear. Because of the limited ability of film and digital sensors to capture as much tonal range of light as the human eye can, photographic images are often challenged by areas that appear to have too much luminosity (over exposure) and areas that have too little luminosity (under exposure). In a scene where the human eye can easily see color and detail on a sunlit mountainside and in the shadow of a tree at the same time, the detail and saturation in these areas of a photograph of the scene would be compromised. Luminosity masking is a masking technique in Photoshop that allows one to accurately select and adjust the luminosity of different tonal values within an image. There are many ways of working with luminosity in Photoshop, including blending different exposures of the same image, dodging and burning, screening and multiplying and using curves or levels adjustments. While all are useful in a variety of situations, they also have their limitations because of the difficulty of confining the adjustments to specific areas of an image and easily blending the adjusted area with unadjusted areas so it looks natural.

Luminosity masking makes it possible to adjust the luminosity of varying degrees of dark, light and mid tone areas in an image in a way that is gradual and completely blended. This is done by converting luminosity channels into highly detailed masks that can then be used for targeted curves adjustments. It sounds complicated, which it is to some degree. However, the power, subtlety and precision the technique provides makes learning it a worthwhile effort for the serious photographer.

The following are a links to two photographer sites that have well written tutorials on luminosity masking. Tony Kuyper’s tutorial at www.goodlight.us is particularly excellent and has an accompanying set of actions that can be downloaded. Bob Johnson’s site www.earthboundlight.com has a tutorial on luminosity masking that takes a slightly different approach towards the same goal.