Update available
Changes
- added verbose mode
- added feature that allows you to enforce a different library to be loaded
- fixed issues related to imperfect stream patching
- fixed issues with old libraries with missing functions that cause xtool to crash on startup
- updated oodle codec
- updated reflate codec
- updated zstd codec
Notes
You can now enforce xtool to use a specific library rather than renaming the library for it to picked up by xtool.
Code:
[External compressor:xtool]
header = 0
packcmd = xtool.exe precomp { -moption} -c32mb -t100p --zstd=libzstd148.dll - - <stdin> <stdout>
unpackcmd = xtool.exe decode -t100p - - --zstd=libzstd148.dll <stdin> <stdout>
Available parameters for this are:
--zlib=filename, --lz4=filename, --lzo=filename, --zstd=filename, --oodle=filename
You can also use this to run multiple tests to see which zstd or oodle library gives the best output by running these commands on a small sample:
Code:
xtool.exe precomp -mkraken -c32mb -t100p --oodle=oo2core_7_win64.dll - - < %1 > %1.out1
xtool.exe precomp -mkraken -c32mb -t100p --oodle=oo2core_7_win64_2.dll - - < %1 > %1.out2
xtool.exe precomp -mkraken -c32mb -t100p --oodle=oo2core_8_win64.dll - - < %1 > %1.out3
xtool.exe precomp -mkraken -c32mb -t100p --oodle=oo2core_fifa22.dll - - < %1 > %1.out4
Then you can pick whichever produces the best results
___
Oodle code has been updated, if you having issues with stream detection you can use the parameter n1 (kraken,n1), where the old code will be used. it can also be n2 or n3, depending on how many times each stream should be redetected (for improved results but at a cost of speed)
If this doesn't help then use the side project.
___
Verbose mode was added to give information about input, as to what parameters are used for streams within the input and if you see a common trend then you directly use these parameters to speed up xtool and also give it the opportunity to patch streams if there are any invalid streams.
Verbose mode will enforce xtool to 1 thread thus it should not be used by default as it will slow now precompression.