Batch iso <=> cso converter.

[[Unknown]]

1. Sky's the limit. If you omit, it will auto detect based on number of cpu cores. I get approximately the same performance with 8 or 16 threads (I have 4 hyperthreaded cores.)

2. Correct. So if you don't mind waiting a while, "--use-zopfli" will give you best compression (it will use the other two as well by default.)

3. You also change METHOD here. So, if you want to only use 7-zip's deflate algorithm, use "--no-zlib --use-7zdeflate" (--use-7zdeflate is optional since it's enabled by default.) Yes, If you use "--no-zlib --no-7zdeflate", it will not compress at all, and create a larger file than the iso (due to cso format overhead.)

The reason to allow customizing the methods is because they take time. "zlib" runs four different trials of zlib compression (--fast disables this and uses only the standard method.) "7zdeflate" is a little slower than zlib but often gives better compression. zopfli is significantly slower than any of the others, but usually gives the best compression.

If you use zopfli only, you'll wait like 90% as long as using all of them, and end up with a slightly larger file. From my testing, every algorithm has sectors it can compress better than the others.

That being said, that's more interesting to a compression nerd like me. It all really only amounts to a 1% difference in filesize. I turned off zopfli by default because it was even slower than I expected it to be.

Anyway, --fast matches what other compressors use (although it might sometimes give slightly better compression), and is much faster, even on a single core, from my tests. So you can just use --fast and have it finish faster for free.

-[Unknown]