I would like to give you a very brief overview about manipulating the TIFF Image File. "TIFF" is either an abbreviation for the Toronto International Film Festival, or Tagged Image File Format - we will be looking at the latter.
First of all, what is TIFF? Here is a technical page that should answer some questions, for those who are deep into computing. I hardly understood a word of it to be truthful - all the talk about "Tags" and "Offsets" simply had me slack-jawed...
So here's the Dave Definition, intended for those of you who really know something about this to crap all over - please, PLEASE add your comments below - I don't get many comments here, and I want some! Now is the day!
- TIFF is an uncompressed image file format which is quite old; it first appeared in the mid 1980's seemingly to provide some flexibility in creating files within Fax Machines. It is also capable of compression in a loss-less manner, meaning the file size can be reduced without compromising image quality (which is good). It is also capable of bit-depth of up to 32 bits (which is very good). Finally, it is "tagable" and this is the part I don't understand at all, but evidently, it means something along the lines that certain extra bits and bytes of data can be placed as needed within all the other data that makes up the lines and colours of the picture. I hate to ask - who wants to ad extra data, and how is it done?
Then a little light came on. You'll recall when we discuss the RAW files that are generated by "good" cameras are really not picture files at all, but rather all the data about the picture, for which special RAW conversion software is needed to generate a preview of the image, and allows very heavy manipulation to really enhance it's appearance, and finally exports the data as a real picture file (usually JPEG). Well, maybe, TIFF is a "meet you half way" kind of file which is a real picture file, but at the same time can be made to carry some extra data payload about the picture too.
Well then, whether I'm right or wrong, lets go with that. So, it turns out that my Epson V500 scanner puts out 8-bit or 16-bit TIFF files, as well as 8-bit JPEGs. I always keep it in 16-bit TIFF mode, so I've got plenty of samples to work with. But what are we dealing with here, really? To answer my question - "who adds the extra data tags?" The answer must be "my scanner", because it sure ain't me. "What is my scanner adding?" I have no idea. Finally, the only question that really matters - "Do the data tags that my scanner is adding create any value to the final product?" Well, maybe, to a limited extent. So, I set out with a simple experiment. First, I scanned a specific image in JPEG mode, and here it is:
Next, I scanned the same negative in TIFF-16 mode -
The only difference I can see is slightly better contrast - look at the snow in the foreground, or at the trees against the cloudy sky - you'll see it there best. Also I should mention the JPEG file size is 4.3 MB, but the TIFF is 24.9 MB! A file size of six times bigger is hardly worth a little extra contrast, is it?
However, there is an up side, I think. When you import the TIFF into your image editing software (Lightroom, Photoshop, Photivo, etc.), it turns out that the TIFF format allows a better range of enhancement. Here are two TIF to JPG conversions I made using Photivo: