Yesterday, my colleague Meg and I had an animated discussion about tiff (as one does): we neither of us like it much as a long term, or even short term, image format, and we were asked to explain why. Tiff is, after all, used or recommended for many cultural heritage collections, and has been around since 1992 (https://partners.adobe.com/public/developer/en/tiff/TIFF6.pdf). Even though it is a proprietary Adobe product, the documentation for the format is readily available, almost as though it were open source. Meg, who is currently undertaking research into early digital music, and is building a Trautonium, has more digital street cred than I, but on this we were united.
Meg and I are also united on the issue of metadata, but for Meg, the metadata or tags created in generic tiff files were ‘junk’ data. The tags from which tiffs derive their name are generally not well used or understood by those creating the images, and seemed that many don’t even realise that tags are being created. Meg also identified that the image format to which the tags applied was as varied as the entire run of image formats, and that they were often compressed, usually without the imager/creator knowing that it was part of the settings, or what compression level had been chosen (it’s in the tags, but you have to be able to open and read them). The file sizes are enormous, in large part due to the tag fields, even if there is no meaningful data attached.
I was more concerned about the proprietary nature of tiff. Having originally been involved in the digitisation of maps at the SROWA, the masters of which were saved in a form of .tiff, I have been aware that the original ImageViewer software had been summarily removed by Kodak during one of the Windows update, and that reading multi page tiffs on a Mac was almost impossible (you save it as a pdf, then open in Preview. Easier to just create a .pdf). I, therefore, greeted with some enthusiasm the report from the Digital Preservation Coalition, Tiff or Jpeg 2000?, in 2010. While there has been some desultory debate since then, the recently presented Library of Congress blog, ‘Comparing formats for still image digitsation’ parts one and two (which, yes, was in an email mailbox – boom tish!), are providing me with much food for thought.