Page 1 of 10 123 ... LastLast
Results 1 to 30 of 502

Thread: Squeeze Chart

Hybrid View

Previous Post Previous Post   Next Post Next Post
  1. #1
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts

    Post Squeeze Chart

    Hey guys,

    I have brought online another update of my website.
    The main testsets and and the special treatment page is updated.

    On the main page I added BALZ 1.06, Slug 1.27 among with some others.
    I thought it might be interesting to see how Stuffit performs with all filters enabled
    and with just the two lossless filters engaged . So those results are also included.

    BALZ 1.06 made a huge jump upwards and also Slug 1.27 compresses better.

    Somehow offtopic but I've watched the new 'Knight Rider 2008' pilot movie
    (the american languaged one) and I found it really cool.. If only my car had
    a "NanoSkin" as well...

  2. #2
    Moderator

    Join Date
    May 2008
    Location
    Tristan da Cunha
    Posts
    2,034
    Thanks
    0
    Thanked 4 Times in 4 Posts
    Last edited by LovePimple; 15th May 2008 at 03:08.

  3. #3
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    4,010
    Thanks
    399
    Thanked 398 Times in 152 Posts
    Quote Originally Posted by Stephan Busch
    BALZ 1.06 made a huge jump upwards
    Cool! Thanks for testing! I'm pretty sure that new BALZ v1.07 will made another huge jump...

    Quote Originally Posted by Stephan Busch
    Somehow offtopic but I've watched the new 'Knight Rider 2008' pilot movie
    (the american languaged one) and I found it really cool.. If only my car had
    a "NanoSkin" as well...
    Will check it out ASAP...

  4. #4
    Moderator

    Join Date
    May 2008
    Location
    Tristan da Cunha
    Posts
    2,034
    Thanks
    0
    Thanked 4 Times in 4 Posts
    Quote Originally Posted by Stephan Busch View Post
    Somehow offtopic but I've watched the new 'Knight Rider 2008' pilot movie
    (the american languaged one) and I found it really cool.. If only my car had
    a "NanoSkin" as well...
    More info on the movie here.

    http://www.imdb.com/title/tt1114258/

    http://en.wikipedia.org/wiki/Knight_Rider_(2008_film)

    http://www.youtube.com/watch?v=j5Gs6f5pb5s

  5. #5
    Programmer
    Join Date
    Feb 2007
    Location
    Germany
    Posts
    420
    Thanks
    28
    Thanked 160 Times in 18 Posts
    Thanks a lot for the update, Stephan!

    Quote Originally Posted by Stephan Busch View Post
    I thought it might be interesting to see how Stuffit performs with all filters enabled
    All filters enabled means lossy, right? Obviously, it's performance on Squeezechart is good, but not really mind-blowing (7000s using 4 cores!?).

  6. #6
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Yep, Christian,

    the round about 7,000 sec were achieved using all 4 cores..
    With one core it would be as fast as LPAQ9e

    But 12.0.0.8 took about 8,000 sec so there's an improvement.

  7. #7
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,568
    Thanks
    775
    Thanked 687 Times in 372 Posts
    btw, rar uses 4 cores too. at least on metacompression site it's wall times are less than half of cpu times

  8. #8
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,568
    Thanks
    775
    Thanked 687 Times in 372 Posts
    Stephan, can you test new freearc version - http://www.haskell.org/bz/FreeArc-0....2008-05-15.exe ? i hope that it finally will perform better than 0.40 also i was surprised by too fast compression of 0.50. may be this is a bug in testing? 2500 seconds looks too good for me

    one more test i very interested is "4x4 5" mode. i expect it to be both faster and tighter than slug due to full utilization of 4 cores

  9. #9
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Tested WinARC 0.50, Slug 1.27b, QuickLZ 64 Bit (latest) and 4x4 in mode 5.

    Results are online now.


    Now I'm testing BALZ 1.07
    Last edited by Stephan Busch; 16th May 2008 at 18:41.

  10. #10
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    4,010
    Thanks
    399
    Thanked 398 Times in 152 Posts
    Quote Originally Posted by Stephan Busch View Post
    Now I'm testing BALZ 1.07
    Thanks! Can't wait...

  11. #11
    Tester
    Nania Francesco's Avatar
    Join Date
    May 2008
    Location
    Italy
    Posts
    1,565
    Thanks
    222
    Thanked 146 Times in 83 Posts
    Please Include PACKET v.0.02! Hi Stephan !

  12. #12
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts

    Talking

    Packet 0.02 and BALZ 1.07 results are online now.

  13. #13
    Member Alexander Rhatushnyak's Avatar
    Join Date
    Oct 2007
    Location
    Canada
    Posts
    248
    Thanks
    45
    Thanked 103 Times in 53 Posts
    Stephan, a couple of questions about lines
    MRP Minimum Rate Predictors v0.5 -I 12 Ichiro Matsuda ☺ 26.401,9 12.716.708
    ( big_building.pgm )

    and
    MRP Minimum Rate Predictors v0.5 -I 12 Ichiro Matsuda ☺ 16.804,2 12.634.480
    ( nikon8.tif )

    on page http://www.squeezechart.com/bitmap.html .

    1. Why didn't you try decompressing these compressed files? MRP-decompression is pretty fast.
    2. How long ago were these two results added?

    BTW, there's a link to http://img.photographyblog.com/revie...sample_images/
    (approximately at the middle of bitmap.html), this link is broken.
    Last edited by Alexander Rhatushnyak; 21st July 2010 at 05:30.

  14. #14
    Member Surfer's Avatar
    Join Date
    Mar 2009
    Location
    oren
    Posts
    203
    Thanks
    18
    Thanked 7 Times in 1 Post
    Quote Originally Posted by Alexander Rhatushnyak View Post
    BTW, there's a link to http://img.photographyblog.com/revie...sample_images/
    (approximately at the middle of bitmap.html), this link is broken.
    Suppose this http://www.photographyblog.com/revie...sample_images/

  15. #15
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Hi Osman,


    There are a bunch of HDR tools available such as MediaChance DynamicPhotoHDR, Photomatix.. - but they combine a set of images taken with different exposure value steps
    to fill shadows with details and raise the dynamic range. The same is done with cameras like the pentax k-x, fujifilm 'exr' denoted cameras and sony nex series.
    But the biggest disadvantage seems to be that those "rendered" HDR lose detail, are blurred.. I could not find any high resolution photo HDR in excellent quality at 100% zoom.

    __________________________________________________ __________________________________________________ ________________________________________________

    Hi Alexander

    These bitmap compression tests are fresh. They were done two weeks ago.
    This site is not ready now as there are still some files that need to be tested (sony48.tif, olympus48.tif, canon48.tif, sony24.tif, olympus24.tif, canon24.tif)
    Maybe I'll add decompression later.

    Thanks for reporting me this broken link. As Surfer supposed correctly, the link is http://www.photographyblog.com/revie...sample_images/

    Why GraLIC doesn't support uncompressed .TIFF directly? Converting to p.ex. .PGM/.PPM discards all metadata. Photographers of today are still missing
    a decent lossless image format which has tagging support like .TIFF and state-of-the-art lossless compression like JPEG2000, JPEG-LS or GraLIC.

  16. #16
    Member Alexander Rhatushnyak's Avatar
    Join Date
    Oct 2007
    Location
    Canada
    Posts
    248
    Thanks
    45
    Thanked 103 Times in 53 Posts
    Quote Originally Posted by Stephan Busch View Post
    Maybe I'll add decompression later.
    It is important to test decompression+comparison for bit-exactness
    (fc/b file1 file2 in DOS and Windows, diff file1 file2 in Linux).
    Did you ever bother comparing decompressed files for bit-exactness with originals while creating Squeeze Chart reports?

    The problem with MRP is that it will fail to decompress these files:
    big_building.pgm.mrp - 12.716.708
    sony8.pgm.mrp - 12.500.000 + x
    canon8.pgm.mrp - 12.500.000 + x
    nikon8.pgm.mrp - 12.634.480

    The reason is in the source file rc.c, and it can be easily corrected.
    Code:
        while(rc->range < RANGE_BOT){
    	putc(rc->low >> (RANGE_SIZE - 8), fp);
    	rc->code += 8;
    	if (rc->code > 1E8) {
    	    fprintf(stderr, "Too large!\n");
    	    exit(1);
    	}
    	rc->range = ((-rc->low) & (RANGE_BOT - 1)) << 8;
    	rc->low <<= 8;
        }
    100'000'000 bits equals to 12'500'000 bytes.
    Simply remove the if (rc->code > 1E check
    (you'll find it twice in rc.c)
    and compile the sources to correct this problem.
    If you need executables for Windows: http://www.imagecompression.info/gra...c/mrp05gcc.rar

    Quote Originally Posted by Stephan Busch
    Why GraLIC doesn't support uncompressed .TIFF directly?
    That's a lot of work.
    Are there any JPEG2000 implementations that support TIFF?
    Last edited by Alexander Rhatushnyak; 21st July 2010 at 21:20.

  17. #17
    Moderator

    Join Date
    May 2008
    Location
    Tristan da Cunha
    Posts
    2,034
    Thanks
    0
    Thanked 4 Times in 4 Posts
    Quote Originally Posted by Alexander Rhatushnyak View Post
    The problem with MRP is that it will fail to decompress these files:
    big_building.pgm.mrp - 12.716.708
    sony8.pgm.mrp - 12.500.000 + x
    canon8.pgm.mrp - 12.500.000 + x
    nikon8.pgm.mrp - 12.634.480

    The reason is in the source file rc.c, and it can be easily corrected.
    Code:
        while(rc->range < RANGE_BOT){
    	putc(rc->low >> (RANGE_SIZE - 8), fp);
    	rc->code += 8;
    	if (rc->code > 1E8) {
    	    fprintf(stderr, "Too large!\n");
    	    exit(1);
    	}
    	rc->range = ((-rc->low) & (RANGE_BOT - 1)) << 8;
    	rc->low <<= 8;
        }
    100'000'000 bits equals to 12'500'000 bytes.
    Simply remove the if (rc->code > 1E check
    (you'll find it twice in rc.c)
    and compile the sources to correct this problem.
    Corrected version attached...
    Attached Files Attached Files

  18. #18
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Dear Alexander,

    throughout the years I have done compression tests and compared the extracted data to the original using MD5 and SFV checksum verification -
    especially on all the recompressors. With MRP 0.5 I did checks on the old bitmap files and thought I don't have to test it again on the new files.
    So you are right - I should perform decompression tests on the bitmap page, too.

    There are some non-standard jpeg2000 implementations for tiff libraries, but not for the mostly used LIBTIFF,
    so neither free tools like XnView nor payware like Photoshop can create .TIFF using JPEG2000 compression.
    And converting a .TIFF to .JPEG2000 discards also all metadata since most software uses the JASPER library which
    does not support metadata.

    Whatever your plans for GraLIC are - adding .TIFF support might be a clever marketing idea,
    because lossless archiving of digital photographs (except JPEG) is still a hassle if you want all metadata and more than deflate compression.

    Dear LovePimple - thank you very much for the fixed version.

  19. #19
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Hi folks,

    I have updated the site.

  20. #20
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    4,010
    Thanks
    399
    Thanked 398 Times in 152 Posts
    Thanks a lot!

  21. #21
    Member biject.bwts's Avatar
    Join Date
    Jun 2008
    Location
    texas
    Posts
    449
    Thanks
    23
    Thanked 14 Times in 10 Posts
    Why isn't Yuta in your Squeeze Hall of Fame?

  22. #22
    Expert
    Matt Mahoney's Avatar
    Join Date
    May 2008
    Location
    Melbourne, Florida, USA
    Posts
    3,257
    Thanks
    307
    Thanked 796 Times in 488 Posts
    And about 10 photos of me? Shouldn't each person only be shown once?

  23. #23
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Who is Yuta?

    to Matt: each person is shown once for each invention/development/improvement.
    Some persons have developed more than one item - it would be unfair to ignore them.

  24. #24
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,918
    Thanks
    291
    Thanked 1,274 Times in 720 Posts

  25. #25
    The Founder encode's Avatar
    Join Date
    May 2006
    Location
    Moscow, Russia
    Posts
    4,010
    Thanks
    399
    Thanked 398 Times in 152 Posts
    Yuta - Yuta Mori the developer of DivSufSort!

  26. #26
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts

    Arrow new archiver rating method

    A happy new year to everyone.

    Since it has become boring for me to just rate archivers by compression strength,
    I want to introduce a new rating method that takes more items into account.

    This new rating method will only be applied to lets say the TOP 10 archivers
    and the results will be displayed on the index page of the Squeeze Chart.

    www.squeezechart.com/review.html

    Please let me know what you think about it.

    Regards,

    Stephan

  27. #27
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,568
    Thanks
    775
    Thanked 687 Times in 372 Posts
    seems that www.rarlabs.com had monopolized this market

    about freearc speeds - you just not tested it in other modes than slowest

    http://forum.ru-board.com/topic.cgi?...64&start=20#15 :

    Compressed 35,018 files, 6,636,984,411 => 3,606,580,870 bytes. Ratio 54.3%
    Compression time: cpu 96.19 secs, real 58.03 secs. Speed 114,371 kB/s
    Last edited by Bulat Ziganshin; 2nd January 2011 at 18:31.

  28. #28
    Member
    Join Date
    Oct 2010
    Location
    Germany
    Posts
    288
    Thanks
    10
    Thanked 33 Times in 21 Posts
    I'm not so happy about the appearance of SAC without notifying me.

  29. #29
    Tester
    Stephan Busch's Avatar
    Join Date
    May 2008
    Location
    Bremen, Germany
    Posts
    876
    Thanks
    474
    Thanked 175 Times in 85 Posts
    Sebastian: I can remove SAC if you want.

    Bulat: Testing on Squeeze Chart is mainly about testing best compression modes. If you want I'll try a faster mode.
    I love all five archivers and enjoy their strengths. I'm just trying to compare them in another way the pure chart does.
    Last edited by Stephan Busch; 2nd January 2011 at 21:24.

  30. #30
    Member
    Join Date
    Oct 2010
    Location
    Germany
    Posts
    288
    Thanks
    10
    Thanked 33 Times in 21 Posts
    Quote Originally Posted by Stephan Busch View Post
    Sebastian: I can remove SAC if you want.
    No it's not necessary, I only hope SAC extracts the test-file perfect. There are some known bugs in this early alpha when there is much noise.

Page 1 of 10 123 ... LastLast

Similar Threads

  1. Squeeze Chart 2007
    By Mike in forum Forum Archive
    Replies: 293
    Last Post: 22nd April 2008, 18:44
  2. Squeeze Chart 2006 - 02 April/13 May
    By encode in forum Forum Archive
    Replies: 9
    Last Post: 17th July 2006, 04:39

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
  •