Range Encoding for Data Compression
deegen1.github.io/rangecomp/index.html
1 Users
0 Comments
3 Highlights
0 Notes
Tags
Top Highlights
biggest problem that an actual range encoder implementation faces is the use of finite precision arithmetic
our goal is to keep range in the bounds norm/2 < range <= norm. Whenever range is equal to or below norm/2, then we consider the high bit of low settled, so we output it and shift down.
naccurate scaling has the advantages of being easy to implement with fixed precision integers as well as being fast. Its disadvantages lie in its loss of accuracy (hence the name) and inability to decode malformed input streams.
Ready to highlight and find good content?
Glasp is a social web highlighter that people can highlight and organize quotes and thoughts from the web, and access other like-minded people’s learning.