Tuesday, October 5, 2010
The Facts About Upsizing Images
How big can you go?
Resize For Output
Once you’ve reduced noise and otherwise optimized the image, it’s time to resize the image to its final output size. In theory, you could leave this task to the printer, but you’ll get the best results by enlarging the image before printing. At a basic level, you can utilize the built-in tools for interpolating your images, which all applications you may use for printing will offer. In the case of Aperture and Lightroom, the work is done for you when you specify the output size for an image. In Photoshop, you can choose several different algorithms when enlarging an image using the Image Size command. The only two you really need to consider are Bicubic and Bicubic Smoother. For significant enlargements, you’ll generally get the best results with Bicubic Smoother, which is designed to avoid the appearance of jagged edges in the enlarged image. It’s wise to test both options before making a final decision for a given image.
In addition to the enlargement options found within the software tools used by most photographers, there are third-party solutions. One of the best among these is Genuine Fractals from onOne Software, which uses a sophisticated approach to ensuring maximum quality and detail in the final image.
The final step in preparing an image to be printed is sharpening, and here again you can have a significant impact on the quality of the final print. This requires that you balance the need to accentuate fine detail with the importance of not creating objectionable halos in the final image. Part of the challenge, of course, is that you need to sharpen not based on what you see on your monitor display, but rather based on the behavior of the printer you’re using. In most cases, that calls for slightly (very slightly) oversharpening the image to compensate for dot gain, which is the spreading of ink on paper. It can be helpful to test various sharpening settings on some sample images, cropping the image after it has been enlarged to avoid wasting too much paper and ink. Then try different sharpening settings so you can evaluate what the image should look like on your monitor in order to ensure the best results on paper.
The Bottom Line
Taking all of this into account, the real question on the minds of many photographers is simply, “How big can I print?” The answer: pretty big. Assuming you’re working with a relatively recent model of digital camera, and that you’ve been careful to optimize the results throughout your workflow, you can very comfortably double the height and width of a photo (quadrupling the surface area). In other words, a relatively modest DSLR with a 10-megapixel image sensor enables print sizes of around 16x24 inches. A top-end DSLR with a 20-megapixel image sensor produces images you can comfortably enlarge to about 30x40 inches.
With the very best equipment, flawless technique and a careful workflow, you can go even larger, especially if you don’t expect viewers to get nose smudges on your prints. Under ideal circumstances, it isn’t out of the question to enlarge to triple the height and width. That translates to print sizes of around 24x36 inches for a 10-megapixel capture and around 40x60 inches for a 20-megapixel capture.
As equipment and software improve, we’ll certainly continue to push the limits of just how big we can enlarge our photographs. Most of the benefit is likely to come from image-sensor improvements, both in terms of higher resolution and lower noise levels. But software tools and printer technology also will continue to contribute. With time, perhaps the only question we’ll need to ponder is whether we really want (or need) to print as large as technology will allow.
Page 2 of 3