Friday, December 28, 2007

JPEG2000 (J2K) vs. TIFF

I've just been looking at a presentation Ronald J. Murray, from the Library of Congress, did at MCN 2005 on J2K. In it, Murray compares J2K with TIFF, using information that is familiar to some about photography (e.g., Halide-Based Imaging Systems).

In the presentation, he notes that TIFF is like taking a photograph, where you may make some adjustments to the image when it is taken, but then fix the image and make it permanent. Derivatives are made by making copies of the master.

He then talks about J2K as being a format where much data is captured, then
File management and delivery system meters out image quality (up to a maximum established or available) on demand
Notice that "creating" different quality access images with J2K doesn't require creating derivative images (files), but instead relies on the delivery systems for displaying on-demand what the user wants.

You can view Ron Murray's complete presentation here.


Technorati tag:

1 comment:

Anonymous said...

Oya Reiger has a fantastic in- depth paper on all the large scale digitization projects, on page 21 she has a table of what output each of the various projects provide as output:

http://www.clir.org/activities/details/lsdi.pdf