Results 1 to 6 of 6

Thread: Unusual file, SUCKS OptimFROG.

  1. #1
    Member kampaster's Avatar
    Join Date
    Apr 2010
    Location
    ->
    Posts
    55
    Thanks
    4
    Thanked 6 Times in 6 Posts

    Unusual file, SUCKS OptimFROG.

    Hi, guys!
    Here audio the file from one game. It completely valid, it is possible to listen to it.
    But it isn't simply to squeeze!
    Who can tell that not so with this wav? All audio coders just faded out.
    Code:
    lzma  9 896 152 byte
    tta   6 051 576 byte
    sbc   5 885 770 byte
    TAK   5 821 212 byte
    La    5 647 731 byte
    FROG  5 618 768 byte
    Uharc 5 339 571 byte
    paq   4 203 909 byte paq8pxd_v7 -5 
    arc   3 656 490 byte (FreeArc:mm+ppmd)
    paq lost
    https://www.sendspace.com/file/1czbng file1.wav
    Last edited by kampaster; 22nd June 2014 at 16:21.

  2. #2
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,471
    Thanks
    26
    Thanked 120 Times in 94 Posts
    I would analyze it with http://mattmahoney.net/dc/fv.cpp and compare the results with analysis of other audio files.

  3. #3
    Member kampaster's Avatar
    Join Date
    Apr 2010
    Location
    ->
    Posts
    55
    Thanks
    4
    Thanked 6 Times in 6 Posts
    What's next?
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	fv.PNG 
Views:	149 
Size:	153.2 KB 
ID:	3005  

  4. #4
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,471
    Thanks
    26
    Thanked 120 Times in 94 Posts
    I think that the file is simultaneously amenable to delta filtering and repetition extraction. Probably there are more long range similarities than in usual audio file, thus audio codecs don't fare well as they code input data in small independent blocks.

    I would analyze with fv.cpp both the file you've stumbled upon and some usual audio files, both with delta-filter preprocessing and without and then post resulting bitmaps here for others to look at.

    You can also firstly preprocess the file with FreeArc's delta filter and then compress the resulting file with general purpose algorithms like lzma (or do it in one step using FreeArc's facilities).

  5. #5
    Member kampaster's Avatar
    Join Date
    Apr 2010
    Location
    ->
    Posts
    55
    Thanks
    4
    Thanked 6 Times in 6 Posts
    https://www.sendspace.com/file/1czbng file1.wav
    Code:
    file1.wav  12 354 092 byte
    lzma       9 896 152 byte
    delta+lzma 4 970 974 byte
    mm+lzma    4 253 187 byte
    _____
    https://www.sendspace.com/file/a1gkfq cut.wav
    Code Geass R2 OST 1 - 03 Madder Sky (cut to 1min)

    Code:
    cut.wav    12 361 192 byte 
    lzma       10 993 784 byte 
    delta+lzma 8 175 317 byte 
    mm+lzma    8 092 860 byte
    ____
    Code:
    cut.wav
    tta   5 045 890 byte 
    uharc 7 112 092 byte 
    arc   8 299 259 byte  (mm+ppmd)
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	file1.wav.PNG 
Views:	106 
Size:	153.2 KB 
ID:	3006   Click image for larger version. 

Name:	file1.wav_mm.PNG 
Views:	114 
Size:	215.1 KB 
ID:	3007   Click image for larger version. 

Name:	file1.wav_delta.PNG 
Views:	102 
Size:	226.0 KB 
ID:	3008   Click image for larger version. 

Name:	cut.wav.png 
Views:	96 
Size:	141.1 KB 
ID:	3009   Click image for larger version. 

Name:	cut.wav_mm.PNG 
Views:	107 
Size:	178.5 KB 
ID:	3010  

    Click image for larger version. 

Name:	cut.wav_delta.PNG 
Views:	101 
Size:	181.9 KB 
ID:	3011  

  6. #6
    Member
    Join Date
    Jun 2009
    Location
    Kraków, Poland
    Posts
    1,471
    Thanks
    26
    Thanked 120 Times in 94 Posts
    High compression ratio can be caused by low volume. Low volume means lower amplitudes and ultimately lower values in final WAVE file and in turn that means more repetitions.

    It would be worth checking how LZMA window size and PPMD model size (and maybe model order) influence the compression ratios on filtered and unfiltered versions of the sound files. PPMD should be run with model cutoff option enabled (as opposed to model flush/ restart on overflow). I expect that lowering the PPMD model size shouldn't hurt compression much in case of normal WAVE files. At least much less than in case of file with long range repetitions.

Similar Threads

  1. Replies: 5
    Last Post: 9th July 2012, 10:57
  2. OptimFROG is back!
    By m^2 in forum Data Compression
    Replies: 3
    Last Post: 16th February 2011, 00:23
  3. Can't extract file from ARC file.
    By Absurd in forum Data Compression
    Replies: 3
    Last Post: 26th January 2009, 21:11

Posting Permissions

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