For text compression benchmarks it's done http://mattmahoney.net/dc/text.html http://mattmahoney.net/dc/text.html Matt doesn't do this on the Silesia corpus co

For text compression benchmarks it's done http://mattmahoney.net/dc/text.html http://mattmahoney.net/dc/text.html Matt doesn't do this on the Silesia corpus compression benchmark, even though it would make sense there as well: http://mattmahoney.net/dc/silesia.html http://mattmahoney.net/dc/silesia.html So a compressor of a few gigabyte would make sense if you would have a set of pictures of more then a few gigabyte. It's a bit similar to preprocessing text compression with a dictionary and adding the dictionary to the extractor to squeeze a bit more bytes.