Tuesday, October 5, 2010
The Facts About Upsizing Images
How big can you go?
When photographers think about producing a large print, it seems they focus their attention on the photograph itself. They’ll think about final output size, detail in the image, sharpness and other factors related to the photographic print. This makes perfect sense considering that from the perspective of the photographer, what they’re enlarging is the photograph. But enlarging a digital photo is a rather sophisticated process, and understanding what’s going on behind the scenes can provide insight into just how large a particular image can be printed.
When you enlarge a photograph, you’re not just taking something and making it bigger. Rather, each individual pixel must be evaluated and then new pixels need to be placed in between existing pixels. Through this process, new information is being added where there wasn’t any information before. In other words, we’re increasing the amount of information when enlarging a photo, and the quality of that information determines the quality of the enlargement.
PictureCode Noise Ninja; Adobe Photoshop Creative Suite; Adobe Photoshop Lightroom 3; Nik Software Viveza 2; Nik Software Dfine 2.0
It may come as a surprise to many photographers that a big jump in megapixel resolution actually results in a relatively modest increase in the native (uninterpolated) output size you’ll achieve with a higher-resolution digital camera. A digital camera offering 10 megapixels, which is around the lower end of today’s DSLRs, would produce a print of about 8x12 inches without any interpolation, assuming a 300 pixel per inch output resolution. Upgrade to a top-of-the-line digital camera of 21 megapixels, and that native output size is about 13x19 inches. At a glance, the increase in output size doesn’t seem commensurate with the higher resolution (or price) of the top cameras. But, of course, resolution is only part of the story when it comes to producing printed output, and different cameras enable different degrees of enlargement in the images they produce.
In the context of a digital photograph, the information we’re enlarging ultimately goes back to the signals gathered and processed by the image sensor on the camera. The quality of that signal directly relates not only to the quality of the final image, but the size to which the image can be enlarged. The cleaner that signal, the more detail (and less noise) will be contained in the image. This is why the top digital cameras are capable of producing what may be considered extreme enlargements, despite the relatively modest increase in output size you may expect based upon the native resolution. While all digital cameras make some effort to minimize noise, the most advanced digital cameras also employ the most advanced technology for maximizing signal and minimizing noise, from microlenses in front of each photosite to highly engineered amplification to sophisticated noise-reduction algorithms.
After the capture, you also can improve the quality of the information in the image by using noise-reduction software. Even if the image doesn’t seem to exhibit any significant degree of noise, if the image will be enlarged to a particularly large size, it can be beneficial to apply at least some degree of noise reduction. The latest versions of Apple Aperture, Adobe Lightroom and Adobe Photoshop all include very good tools for reducing noise in an image. You also can use third-party tools such as Nik Software Dfine, PictureCode Noise Ninja, Neat Image and others. Bear in mind that optimal noise reduction in postprocessing is always a compromise and requires some finesse. Applying too much noise reduction will result in a loss of detail, sharpness and color within the image. It’s important to balance the strength of noise reduction with the actual effect in the image, as well as use the various options available to mitigate the effect in areas of fine detail and to apply some degree of sharpening to compensate for any softness that’s introduced.
Page 1 of 3