Popular opinion says that the bigger your image files, the more chances you have to sell them, and eventually, the more cash you will get. But this is only in theory…
In practical terms, and contrary to what popular forums say, the best you can do is to buy a camera that will give you a decompressed RAW file of 50 Megabytes or more. There are affordable cameras for pro and semipro shooters that will do this. However, if your finances are not yet there because 14 cents an image does not allow for lots of expending, and your existing camera only allows you to get 12.1 mega pixels that give you a meager decompressed 30.3 megabytes file, the question is: do I want to upsize the file to 50MB or more?
If you say yes, how are you going to process substantial shoots of several hundred images? In batch processing? Are you aware that your results are going to be for many images unpredictable and time consuming? Why would you want to expend time hidden in your digital dungeon when the rule of thumb in order to make some money today (and let’s not forget about the 14 cents an image) is to generate lots of images… Volume, volume, volume!!! is what will guarantee good positioning in most ranked search engines and more selling possibilities as a result – if the picture is a good one, of course.
But if all the above does not convince you, and you choose to strain your eyes in front of your computer screen, at least strain them doing your upsizing, technically defined as interpolation, correctly. You have two options for doing this:
You can send the resulting file from your camera, any camera, and our technical team will interpolate the images if necessary, or
You can interpolate your files yourself and send the results to us and wait to see what we say…
But what is interpolation? Here are the three basic concepts:
Interpolation is increasing the size in pixels of digital images, therefore allowing a bigger reproduction of an image that was not originally produced to yield such reproduction size. Interpolation is also known as image resize or even image upsize.
Image resize is not a mathematic miracle that happens when you need to increase the total number of pixels. You can obtain a bigger size image file at the expense of losing definition with this operation. A bigger file is obtained by carefully adding the best approximation in color and intensity based on the values of surrounding pixels, in areas where the damage caused to the image is going to be lower. Unfortunately the damage will occur.
The most frequent types of interpolation algorithms are the nearest neighbor, bilinear and bicubic. The Bicubic is the standard algorithm in many image editing programs used by professional photographers as it produces sharper images and hence better overall quality.
There are many programs that allow you to interpolate your images but there are subtle differences among them; therefore, every photographer has to carefully evaluate which one is the best that suits her/his needs, as the result can vary depending on the interpolation algorithm and the type of subject. For example, resizing an image of an architectural masterpiece is different from resizing a view of the Amazon jungle.
So, should I interpolate my files? The decision is finally yours, of course, but if you do so it’s important that you:
Know the technical characteristics of your files to know how far up you can go.
Check the image at 100% of visualization to verify that the result is correct, and that it has not become pixelated or is displaying artifacts.
Think that sometimes less is more, and good quality files are better than large defective files. If you interpolate do it well. Otherwise just stick to the nominal file size that your camera can give you.
Remember that it is always cheaper to buy a better camera than to chain yourself to your computer screen upsizing your images; please use your time intelligently and get out and shoot more images as volume is really what is needed when photographers are accepting 14 cents for their images…ok, ok, even for $3 dollars per image!
Here is an example of what a badly interpolated file looks like:
The image received was certainly not really crisp – an otherwise ordinary image that does not have
excuses not to be razor sharp…
…but an enlarged detail of that picture at 100% of visualization shows the problem.