OK, I added to BALZ a dummy mixer. I think compression gain worth an extra processing time. Check out:
http://encode.ru/balz/index.htm
BTW, special thanks to Chris for his great ideas!
![]()
OK, I added to BALZ a dummy mixer. I think compression gain worth an extra processing time. Check out:
http://encode.ru/balz/index.htm
BTW, special thanks to Chris for his great ideas!
![]()
Thanks Ilia!
Mirror: Download
Just did a very quick test to see how fast it would be on my AMD Sempron 2400+, Windows XP SP2. Here's the result:
Test file: FP.log
balz ex
Compression Time: 216.91 Seconds
Decompression Time: 1.55 Seconds
balz e
Compression Time: 12.95 Seconds
Decompression Time: 1.69 Seconds
Last edited by LovePimple; 25th May 2008 at 17:51.
Thanks for testing!
Actually new BALZ became strong enough to beat LZMA:
butterfly4.bmp (18,048,054 bytes)
7-Zip 4.57, LZMA-Ultra: 6,838,321 bytes
BALZ v1.10, ex: 6,550,527 bytes
![]()
Some testing results for BALZ v1.10, ex:
sfc.7z: 12,030,330 bytes
A10.jpg: 837,771 bytes
acrord32.exe: 1,380,718 bytes
english.dic: 746,407 bytes
FlashMX.pdf: 3,726,322 bytes
fp.log: 554,646 bytes
mso97.dll: 1,809,664 bytes
ohs.doc: 807,154 bytes
rafale.bmp: 983,871 bytes
vcfiu.hlp: 640,483 bytes
world95.txt: 557,405 bytes
Total: 12,044,441 bytes
![]()
BALZ 1.10 = Super Fast PPMonstr opt. -02 !
Or a slower RAR?
It's a tough world out there, obviously more data is needed with regards to test sets though.
The main problem with large datasets is that nobody's able to check your results.
Hey encode ,do you plan some work in Quad?
Well, BALZ is kind of QUADv2. Some time ago I stopped working on QUAD, moving myself to more experiments... It should be noted that it is possible to increase compression ratio and speed of QUAD without changing the compatibility, however not too much. So, I'm hoping to finish working on BALZ making it sick compression software and release it at Source Forge as just my other projects - QUAD/LZPX(J)...
Work in progress!![]()
OK
Thanks!