Page 2 of 3 FirstFirst 123 LastLast
Results 31 to 60 of 80

Thread: Metacompressor.com benchmark updates

  1. #31
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by Darek View Post
    Could you suggest me what should I do to run this benchmark Properly?
    The problem is that Windows Server 2008 and also Windows Vista with default setting User Account Control (UAC) enabled do not allow file creation in root C:\ where the most paq executables try to write a temp file. Only a compile what do not write in the root C:\ shall run properly.

  2. #32
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by m^2 View Post
    Could you increase maximum command line length? I'm short of ~10 chars.
    I had increased the maximum command line length from 128 to 256 chars length but didn't test it.

  3. #33
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by m^2 View Post
    4. How do you measure times? Some results look strange, i.e. file2, FreeArc -mdelta+rep:800mb+rzm has 6s decompression CPU time. rzm - 71.5s. rep can help, possibly delta too, but....
    The time is a stopwatch in the test software what is started just before the command line is executed and stopped direct after when it's finished.

    The processor time is what Windows give back after the command line is finished.

    There can be differences because disk cache and in time free disk space can be further away from data at disk because the file uploads.

    Thanks for the suggestions 1-3 I keep them in mind for possible future improvements.

  4. #34
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Thanks.

    I still think that in case of delta+rep+rzm something went wrong.

  5. #35
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    I have added an option to filter results on archiver name.

  6. #36
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts

    Thumbs up

    Quote Originally Posted by Sportman View Post
    I have added an option to filter results on archiver name.
    Nice.

  7. #37
    Member biject.bwts's Avatar
    Join Date
    Jun 2008
    Location
    texas
    Posts
    449
    Thanks
    23
    Thanked 14 Times in 10 Posts

    Question USING DGPP G++ executables Windows XP

    Has anyone figured out how to get the programs compiled under
    G++ in DJGPP for Windows XP to run on the Metacompressor system?

    Also will it run a File.com batch file if all exectuables included?

    Thanks

  8. #38
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts
    No way. There's no support for DOS executables in Windows x64. Use MinGW.

  9. #39
    Member biject.bwts's Avatar
    Join Date
    Jun 2008
    Location
    texas
    Posts
    449
    Thanks
    23
    Thanked 14 Times in 10 Posts

    MinGW versus DjGPP

    Well after worrying about it bite the bullet and downloaded
    MINGW

    here is What version I am using

    C:\c\vista>g++ --version
    g++ (GCC) 3.4.5 (mingw-vista special r3)
    Copyright (C) 2004 Free Software Foundation, Inc.
    This is free software; see the source for copying conditions. There is NO
    warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

    I got one test to work on the meta compressor.

    The only trick besides downloading and installing
    is the PATH

    to use MinGW

    PATH=%PATHGJPP=MinGw%

    ** sorry the : D makes a Happy Face on Preview Post**

    to go back to DJGPP

    PATH=%PATH:MinGW=DGJPP%

    this is on an XP
    machine

    Dave

  10. #40
    Member biject.bwts's Avatar
    Join Date
    Jun 2008
    Location
    texas
    Posts
    449
    Thanks
    23
    Thanked 14 Times in 10 Posts
    Will here is the results for FILE 5 the random number file.

    arb25x is actually the MinGW version of arb255

    268,435,486 2693 00:44:51.0952505 1,953,792 2926 00:48:45.6115538 1,953,792 True arb25v test2 David A. Scott arb25x @infile @outfile unarb25x @infile @outfile

    It is designed for a 256 symbol I.I.D file its no wonder its one of the best
    on this list since it may be that sort of file.



    Thank You
    Sportsman

  11. #41
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    I added some JPG images to Metacompressor.com from the Infima Technologies LTD. sample dataset http://www.infima-compression.com/Ev...on/dataset.rar according their own test report 11/17/2008 http://www.infima-compression.com/Evaluation/report.pdf their JPACK best mode compress better and probably quicker then precomp and paq8o8zw. They use a Pentium 4, 3GHz CPU and PC1600 DDR memory as test server so their speed of 2MB/s (fast mode) and 500KB/s (best mode) shall probably be much quicker (2x-3x) at the Metacompressor test server.

    05312008151.jpg - 953,088 bytes (931KB)
    JPACK fast - 839,680 bytes (820KB / 2MB/s)
    precomp 0.3.8 - 760,161 bytes (742KB / 922KB/s)
    paq8o8zw - 716,045 bytes (699KB / 20KB/s)
    JPACK best - 679,936 bytes (664KB / 500KB/s)

    IMG_7627 1,766,664 bytes (1.68MB)
    JPACK fast - 1,551,892 bytes (1.48MB / 2MB/s)
    precomp 0.3.8 - 1,438,766 (1.37MB / 1.12MB/s)
    paq8o8zw - 1,409,197 bytes (1.34MB / 48KB/s)
    JPACK best - 1,279,263 bytes (1.22MB / 500KB/s)

    huge.jpg - 7,190,373 bytes (6.86MB)
    JPACK fast - 6,008,340 bytes (5.73MB / 2MB/s)
    precomp - 5,486,517 bytes (5.23MB / 907KB/s)
    paq8o8zw - 5,346,341 bytes (5.10 / 64KB/s)
    JPACK best - 4,875,878 bytes (4.65MB / 500KB/s)

    Also according their benchmark under Business Solutions Benchmark http://www.infima-compression.com/business_solutions/ they compress FlashMX.PDF better then paq8o8zw probably with their RichPack product.

    FlashMX.PDF 4,526,946 bytes
    paq8o8zw 3,558,014 bytes
    Infima 2,009,510 bytes

  12. #42
    Member Zonder's Avatar
    Join Date
    May 2008
    Location
    Home
    Posts
    55
    Thanks
    20
    Thanked 6 Times in 5 Posts
    Quote Originally Posted by Sportman View Post
    I added some JPG images to Metacompressor.com from the Infima Technologies LTD. sample dataset http://www.infima-compression.com...
    That site is a big scam

  13. #43
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    3,946
    Thanks
    343
    Thanked 326 Times in 129 Posts
    Quote Originally Posted by metacompressor.com
    Compression finished: 'file1' is compressed to 0 bytes in 0 seconds.
    Tried many times to add my new PPMX with no success...

    I tested site with both FireFox and Internet Explorer. Since previously metacompressor.com had some issues with FireFox. I even registered there...

  14. #44
    Member
    Join Date
    Oct 2007
    Location
    Germany, Hamburg
    Posts
    408
    Thanks
    0
    Thanked 5 Times in 5 Posts
    Quote Originally Posted by Sportman View Post

    Also according their benchmark under Business Solutions Benchmark http://www.infima-compression.com/business_solutions/ they compress FlashMX.PDF better then paq8o8zw probably with their RichPack product.

    FlashMX.PDF 4,526,946 bytes
    paq8o8zw 3,558,014 bytes
    Infima 2,009,510 bytes
    PrePaq8 is better

    FlashMX.pdf c8+c9 1,901,189 bytes
    Last edited by Simon Berger; 22nd January 2009 at 17:58.

  15. #45
    Programmer
    Join Date
    May 2008
    Location
    denmark
    Posts
    94
    Thanks
    0
    Thanked 2 Times in 2 Posts

    Fails to decompress

    Hi Sportman,

    Can't you take a look at qpress 0.31 ( http://www.quicklz.com/ )? It fails to decompress on metacompressor.com, with the error "Decompression failed, can be a parameter mistake".

    I'm eager to seeing the results since I added multi core support in version 0.30 (uses 2 cores as default when no flag is given).

  16. #46
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by Lasse Reinhold View Post
    Hi Sportman,

    Can't you take a look at qpress 0.31 ( http://www.quicklz.com/ )? It fails to decompress on metacompressor.com, with the error "Decompression failed, can be a parameter mistake".

    I'm eager to seeing the results since I added multi core support in version 0.30 (uses 2 cores as default when no flag is given).
    Qpress display a help screen, for some reason qpress do not like how the test application send parameters.

    I did some manual tests:

    Archiver Size Comp. Decomp.
    Qpress 0.20 838,711,660 bytes 34.336 sec. 53.663 sec.
    Qpress 0.31 741,167,922 bytes 32.495 sec. 51.527 sec.
    Qpress 0.31 T4 741,167,922 bytes 30.467 sec. 51.433 sec.

    Qpress 0.20 L3 752,575,769 bytes 52.400 sec. 50.467 sec.
    Qpress 0.31 L3 654,338,838 bytes 36.879 sec. 51.886 sec.
    Qpress 0.31 L3T4 654,338,838 bytes 31.387 sec. 49.577 sec.
    Last edited by Sportman; 26th January 2009 at 09:49.

  17. #47
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by Simon Berger View Post
    PrePaq8 is better

    FlashMX.pdf c8+c9 1,901,189 bytes
    Indeed I got 1,821,452 bytes with paq8o8pre -7.

  18. #48
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    Quote Originally Posted by Zonder View Post
    That site is a big scam
    I don't think so, they maybe violate copyrights by reusing source code a way their license do not allow and they install two extra applications (search/toolbar) with their video converter without a disable option in the installer as they did for the Yahoo toolbar but they are a real company http://www.myinfima.com/contact.aspx and http://www.matimop.org.il/newrdinf/company/c7747.htm with 5 employees and real software.

    Their software what's public available work fine:
    http://www.media-cell.com
    It's reviewed by CNet http://cnettv.cnet.com/2001-1_53-29090.html

    Also their Infima Compression Library SDK 2.0 is to download and work fine:
    http://www.download3k.com/Software-D...n-Library.html

    They did an improvement after this and now their JPACK SDK software is not to evaluate anymore without asking them by email.

    In their news they say http://www.infima-compression.com/marketing/tis.pdf their software is licensed to Top Image Systems http://www.topimagesystems.com who license their eFlow to Yahoo who use JPEG recompression to save 20% storage servers for Flickr http://www.flickr.com.

  19. #49
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    3,946
    Thanks
    343
    Thanked 326 Times in 129 Posts
    Infima is crap! It was banned by data compression community and me personally! Such software will not be labeled at this forum...

  20. #50
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Quote Originally Posted by encode View Post
    Infima is crap! It was banned by data compression community and me personally! Such software will not be labeled at this forum...
    Could you explain why?

  21. #51
    Member Zonder's Avatar
    Join Date
    May 2008
    Location
    Home
    Posts
    55
    Thanks
    20
    Thanked 6 Times in 5 Posts
    Quote Originally Posted by m^2 View Post
    Could you explain why?
    1) because they just use stolen compressors...
    2) bullshit/sci-fi/nonsense claims in compression achievements...
    3) and Infima sole purpose is to scam/suck money from possible investors. IMHO...
    Last edited by Zonder; 26th January 2009 at 15:09.

  22. #52
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Quote Originally Posted by Zonder View Post
    1) because they just use stolen compressors...
    2) bullshit/sci-fi/nonsense claims in compression achievements...
    3) and Infima sole purpose is to scam/suck money from possible investors. IMHO...
    Which codecs?

  23. #53
    Member Zonder's Avatar
    Join Date
    May 2008
    Location
    Home
    Posts
    55
    Thanks
    20
    Thanked 6 Times in 5 Posts
    Quote Originally Posted by m^2 View Post
    Which codecs?
    i don't remember, Infima is old scam.
    can offer this link - http://www.google.com/search?num=100&q=infima+fraud

    dunno if those old claims are still on that site, like:
    Xvid video files - lossless compression to ~60%, mp3 files - lossless compression to ~50%, pure nonsense
    Last edited by Zonder; 26th January 2009 at 15:41.

  24. #54
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,612
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Quote Originally Posted by Zonder View Post
    i don't remember, Infima is old scam.
    can offer this link - http://www.google.com/search?num=100&q=infima+fraud

    dunno if those old claims are still on that site, like:
    Xvid video files - lossless compression to ~60%, mp3 files - loosless compression to ~50%, pure nonsense
    Indeed nonsense.
    Thanks.

  25. #55
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    3,946
    Thanks
    343
    Thanked 326 Times in 129 Posts
    Quote Originally Posted by m^2 View Post
    Could you explain why?
    Yep, it's an old story. Infima stored many top compressors in a password-protected UHARC archives. After that, such program should have no respect and trust...

    Quote Originally Posted by http://www.c10n.info/archives/415
    Infima was just command-line versions of a large number of great compression algorithms bundled together (packed and encrypted). Monkey?s Audio Console Front End (v 4.01), Lossless Audio Compressor 0.4b, SBC Archiver 0.970, BMF 1.1, J.Class Optimizer for Windows. Version 1.00, ImageMagick 6.2.4 09/10/05 Q16, UPX 1.93 beta, gzip 1.2.4 Win32 (02 Dec 97), library and tools for JPEG images, DURILCA v.0.4b, Monstrous PPMII compressor based on PPMd var.J, DURILCA v.0.4b, Fast PPMII compressor for textual data, variant J, Feb 16 2006, 7-Zip 4.31, LAME version 4.0 (alpha 14, Sep 25 2005 10:04:14), PAQ 8g, photo PGM(P5) < => BMF file convertor, v.2.0, flasm 1.52 build Sep 30 2004, Word Replacing Transformation, rebuild of WRT, pdftk 1.12, rebuild of Optipng, fCoder Batch Converters, FFmpeg version CVS

  26. #56
    Member
    Join Date
    Oct 2007
    Location
    Germany, Hamburg
    Posts
    408
    Thanks
    0
    Thanked 5 Times in 5 Posts
    Yeah infima had a really bad start but the results at the moment are really good and believable. I don't like them and can understand that you also don't do.
    But it's funny to only say it's crap what they do. If they create the best compression they are the best, irrelevant what they did wrong and which licence they broke (on this fact).

  27. #57
    Member Zonder's Avatar
    Join Date
    May 2008
    Location
    Home
    Posts
    55
    Thanks
    20
    Thanked 6 Times in 5 Posts
    Quote Originally Posted by Simon Berger View Post
    Yeah infima had a really bad start but the results at the moment are really good and believable. I don't like them and can understand that you also don't do.
    But it's funny to only say it's crap what they do. If they create the best compression they are the best, irrelevant what they did wrong and which licence they broke (on this fact).
    they didnt create anything, read that google link i posted.

    Frauds done by same person(s):
    http://www.app-zilla.com/
    http://vio.thepiratebay.org/
    http://www.speedapps.com
    http://www.thepickapp.com/
    probably all those apps are stolen...

    they "invented" and patented "Adaptive Context Modeling" :
    http://www.freepatentsonline.com/y2007/0233477.html
    is that realy something new?
    Last edited by Zonder; 26th January 2009 at 18:31.

  28. #58
    Programmer
    Join Date
    May 2008
    Location
    denmark
    Posts
    94
    Thanks
    0
    Thanked 2 Times in 2 Posts
    Check out http://www.c10n.info/archives/415

    "For compression of already compressed files (mp3, jpeg etc?) they tried a neat (and dangerous) trick of modifying the original files before adding them to archive (so that their lossy techniques wont get caught during bit-wise comparisons of decompressed files with original files)."

  29. #59
    Member biject.bwts's Avatar
    Join Date
    Jun 2008
    Location
    texas
    Posts
    449
    Thanks
    23
    Thanked 14 Times in 10 Posts
    Quote Originally Posted by Lasse Reinhold View Post
    Check out http://www.c10n.info/archives/415

    "For compression of already compressed files (mp3, jpeg etc?) they tried a neat (and dangerous) trick of modifying the original files before adding them to archive (so that their lossy techniques wont get caught during bit-wise comparisons of decompressed files with original files)."
    Its sounds like they are a bunch of crooks. But If they would have used something like what I call my stability output for the final file a bit to bit compare would have failed. I suspect the next big fraud will modify the ouputs in a random way so that to catch them you actually have to do an instruction trace. Which use to be very easy on old machines like the UNIVAC. With today's machines it my be harder since to compare the code you may have to know the compiler they used so it might take a little more work but catching could be done. Or one still would have to examine the exe to undo any stability or other mods that occur just before the final output.
    There are an unlimited ways to mod the final output that would not change the compression by more than a byte so with out a little work it would be hard to catch.
    Dave

  30. #60
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    734
    Thanks
    62
    Thanked 251 Times in 176 Posts
    I have add Infima Jpeg Compression SDK 1.3 http://www.download3k.com/Install-In...ssion-SDK.html to Metacompressor and tested it at jpg test files.
    Last edited by Sportman; 26th January 2009 at 23:36.

Page 2 of 3 FirstFirst 123 LastLast

Similar Threads

  1. zpaq updates
    By Matt Mahoney in forum Data Compression
    Replies: 2522
    Last Post: 18th March 2019, 10:05
  2. Metacompressor.com first contest
    By Sportman in forum Data Compression
    Replies: 17
    Last Post: 15th October 2008, 00:50

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •