Bench Test (LOLZ vs RAZOR vs MCM vs LZMA2)

[felice2011]

*** Enwik8 (95MB) : LOLZx64 ***

lolz_x64:mtt0:mt1:mtb32m:d64m:tt4:dtm1:dto0:dtd0:d tw1:gm00 (Comp\Time 2min:33sec) ( rc 25.26%) (Decomp\Time 2sec)

lolz_x64:dtb1:d128m:mtt1:mt6:mc1023 (Comp\Time 3min:37sec) (rc 25.26%) (Decomp\Time 3sec)

lolz_x64:mtt1:mt4:fba2:mc2:cm1 (Comp\Time 44sec) (rc 42.11%) (Decomp\Time 2sec)

lolz_x64:mtt0:mt1:mtb64:d128 (Comp\Time 3min:12s) (rc 25.26%) (Decomp\Time 2sec)

lolz_x64:mtt1:mt4:d200m (Comp\Time 2min:33sec) ( rc 25.26%) (Decomp\Time 2sec)

lolz_x64:mt4:d300m (Comp\Time 1min:36sec) (rc 26.32%) (Decomp\Time 1sec)

*** Generic *.PAK File (1000MB) : LOLZx64 ***

● First Step (XToolv0.9:ZLIB >= 176.95%)

● Second Step (SREP >= 148.43%)

● Final Step (LOLZ = 62.7%)

xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+lolz_x64: mtt1:mt4:d200m (Comp\Time 21min) ( rc 62.7%) (Decomp\Time 1min:34sec)

*** Generic Movies (*.mp4 + *.avi) (266MB) : LOLZx64 ***

lolz_x64:dtb1:d128m:mtt1:mt6:mc1023 (Comp\Time 4min:45sec) ( rc 96.24%) (Decomp\Time 7sec)

∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾ ∾∾∾∾∾∾∾

*** Enwik8 (95MB) : RAZOR ***

razor:256m (Comp\Time 3min:45sec) ( rc 21.05%) (Decomp\Time 2sec)

razor:512m (Comp\Time 3min:45sec) (rc 21.05%) (Decomp\Time 2sec)

razor:1023m (Comp\Time 4min:03sec) (rc 21.05%) (Decomp\Time 2sec)

*** Generic *.PAK File (1000MB) : RAZOR ***

● First Step (XToolv0.9:ZLIB >= 176.95%)

● Second Step (SREP >= 148.43%)

● Final Step (RAZOR = 68.4%)

xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+razor:512 m (Comp\Time 15min) ( rc 68.4%) (Decomp\Time 1min:26sec)

*** Generic Movies (*.mp4 + *.avi) (266MB) : RAZOR ***

razor:512m (Comp\Time 6min:03sec) ( rc 95.86%) (Decomp\Time 9sec)

∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾ ∾∾∾∾∾∾∾

*** Enwik8 (95MB) : MCM64 ***

mcm64:x10 (Comp\Time 1min50sec) ( rc 17.89%) (Decomp\Time 42sec)

mcm64:x11 (Comp\Time 55sec) (rc 17.89%) (Decomp\Time 47sec)

mcm64:x12 (Comp\Time 54sec) (rc 17.89%) (Decomp\Time 52sec)

*** Generic *.PAK File (1000MB) : MCM64 ***

● First Step (XToolv0.9:ZLIB >= 176.95%)

● Second Step (SREP >= 148.43%)

● Final Step (MCM64 = 68.4% / 68.2%)

xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+mcm64:x10 (Comp\Time 15min) ( rc 68.4%) -- No Tested

xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+mcm64:x11 (Comp\Time 17min) ( rc 68.2%) (Decomp\Time 17min:15sec)

*** Generic Movies (*.mp4 + *.avi) (266MB) : MCM64 ***

mcm64:x11 (Comp\Time 4min:08sec) ( rc 92.11%) (Decomp\Time 4min:48sec)

∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾ ∾∾∾∾∾∾∾

*** Enwik8 (95MB) : LZMA2 ***

Internal Set (-txz -an -mcrc=0 -m1=lzma2:d512m:fb=273:mf=bt4:mc=100000000:lc=4:lp= 0 -mmt=2 -mx9)

lzma2:d512m (Comp\Time 1min14sec) ( rc 24.21%) (Decomp\Time 2sec)

*** Generic *.PAK File (1000MB) : LZMA2 ***

Internal Set (-txz -an -mcrc=0 -m1=lzma2:d512m:fb=273:mf=bt4:mc=100000000:lc=4:lp= 0 -mmt=2 -mx9)

● First Step (XToolv0.9:ZLIB >= 176.95%)

● Second Step (SREP >= 148.43%)

● Final Step (LZMA2 = 69.07%)

xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+lzma2:d51 2m (Comp\Time 8min50sec) ( rc 69.07%) (Decomp\Time 1min:7sec)

*** Generic Movies (*.mp4 + *.avi) (266MB) : LZMA2 ***

Internal Set (-txz -an -mcrc=0 -m1=lzma2:d512m:fb=273:mf=bt4:mc=100000000:lc=4:lp= 0 -mmt=2 -mx9)

lzma2:d512m (Comp\Time 48sec) ( rc 96.24%) (Decomp\Time 7sec)

∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾ ∾∾∾∾∾∾∾

Example of Final Mask Tested On "The Conjuring House" :

srep64:m5f:a2:d1g+razor:256m (in 63MB / rc 19.35%) Or srep64:m5f:a2:d1g+mcm64:x11 (in 63MB / rc 17.74%) (Generic Files)

/$deflate=xprecomp:c32mb,t4:zlib+srep64:m5f:a2:d1g+lolz_x64:mtt1:mt4:d200m (Inflated Stream File)

/$void=srep64:m3f:a1:d512m+mcm64:x11 (Movies Files)

In Data (4.48GB) To Output in Single Archive (< 2.7GB) (Decomp\Time ∾5min / ∾10min) (Depending on Your System)

∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾∾ ∾∾∾∾∾∾∾

[JRD!]

lolz is a step above the rest, can you add the decompression time? That would be great, thanks for the report. Bye.

[K4miKaZe]

At least for me, lolz is faster in decompression than lzma but would be great to see a bench on this too. Thanks for taking your time and test it.
Im only using lolz as my final compressor, sure its slower than lzma but it has better ratio and also installs fast (at least for me).

[felice2011]

Added Decompression Time.

Do not underestimate the optimal compression threshold MCM, on almost all video formats, which uses a compression method similar to Fat Context Mixing (Mixer of contexts or orders) + (lzp), but unfortunately very slow in decompression.

But the developer seems to have had the idea thanks to another famous compressor, able to obtain the same compression threshold on video formats, halving the decompression time.

[PsYcHo_RaGE]

lzma is better in many ways but lolz still surpasses it when its about decompression.

lolz can be turtle or it can rabbit, can you try these settings for lolz

-dto0 -tt1 -dm00 -mc1023

[kokousic]

Is there any CPU specific benchmark showing which CPU best for LOLZ?