What file compression application is the most space-efficient?

Question

I have a bunch of directories that I want to keep, but I also want to clear out storage on my drive.

What is the most efficient (in terms of compression ratio) compression application? I don't care if it's slow as I will not be accessing these files often.

The only constraint is that the program must run on OS X.

Solution 1

the program must run on OS X

I'd say your best bet would be any one of the 7-Zip derivatives for Mac OS X.

7-Zip is free open source software and the compression ratio is marginally better than Rar.

StuffIt also does compress very efficient, however, the price of $80 is rather steep, for that money you'll get a 1 TB hard disk. :)

What file compression application is the most space-efficient?

Here's an extensive test comparing about 250 different archiving programs (not all of them available for Mac OS X though) in various scenarios:

Lossless data compression software benchmarks / comparisons

Solution 2

The storage ratio depends entirely on the data you're compressing. Different algorithms have wildly differing performance.

Comments

[RJFalconer]

I took the liberty of changing the title; I don't think you're actually interested in the algorithm used.

[Charles Stewart]

The data that site uses is representative only under certain assumptions, one being that executables are .exe/.dll win32 binaries. Different sample data may yield different results...

[Josh K]

I really don't like either of those, my preference would be for RAR or another platform independent format.

[Admin]

7-zip is VERY platform independent ... it also supports regular ZIP if you prefer this format.

[Josh K]

Does it? Learn something new every day.

[Admin]

@Josh K - Yes, 7-Zip can compress as 7z, Tar and Zip ... and of course the software is available for all platforms.

[John T]

@Molly Tar is not a compression format, it's an archiving format.

[Grzegorz Adam Hankiewicz]

Concatenation of smaller files yields better compression because usually lossless methods generate a "compression dictionary" per file. If files are small, the dictionary has to be recreated over and over and the compression is not that significant as using one for many where it can be reused. Compare the compression of zip (individual file compression) to that of tar+zip with same compression level of a directory of many small text files.

[John Riselvato]

How is this an answer? It's more of a comment really.