Previous PageTable Of Contents../index.htmlIndexNext Page


Overview of TIFF File Format

GDK maintains facsimile data in a TIFF (Tagged Image File Format) compression format. These files have a TIFF header of no fixed size that describes the data in the file, and helps to distinguish facsimile image files from other types of image files.

GDK uses a subset of the available set of image-description tags recommended in the Microsoft Version 6.0 TIFF specifications.

NOTE:
GDK supports send-only faxing of multi-page TIFF files.

A TIFF file format offers these advantages:

GDK supports the following compression formats:

TIFF Type 3 is compatible with ITU-T Group 3 T.4 recommendation. Either 1-D or 2-D (two-dimensional) modified Huffman encoding compression can be used.

Each image line of a 1-D file is encoded as a TIFF Type 3 1-D line with EOL codes after each scan line, and the TIFF file is terminated with six EOLs to indicate the end of the page. The fill order can be MSBF (most significant bit first) or LSBF (least significant bit first).

The 2-D files also are called "Modified READ" (MR) files. The first image line of these files is encoded as a TIFF 3 1-D line. A certain number of lines, which is usually one to three, follow the 1-D line and encode only the differences between the current line and the previous one. Following the 2-D lines is another 1-D line. Then, there are more 2-D lines based on this new reference line. A flag that indicates whether a line is complete in itself (a 1-D encoded line) or is based on the preceding line (a 2-D encoded line) is embedded in the end of line code.

This encoding scheme results in approximately a 15 to 20 percent reduction in file size over 1-D encoding in most cases. This can vary greatly, depending on the type of image. A failure in one line affects only a small portion of the document.

TIFF Type 4 files also are called "Modified Modified READ" (MMR) files and require Error Correction Mode (ECM). This type is compatible with ITU Group 4 T.6 recommendation. An imaginary white line precedes the first line. Every line in the file is based on the differences between the current line and the line that preceded it; the first line presumes a blank line preceded it. This byte-oriented compression scheme results in approximately a 20 to 40 percent reduction in file size over one-dimensional encoding in most cases. A scan line error can corrupt an entire image from the point of the error forward.


Previous PageTable Of ContentsTop Of PageIndexNext Page

Click here to contact Dialogic Customer Engineering

Copyright 2000, Dialogic Corporation
All rights reserved
This page generated January, 2002