Activity Stream

Filter
Sort By Time Show
Recent Recent Popular Popular Anytime Anytime Last 24 Hours Last 24 Hours Last 7 Days Last 7 Days Last 30 Days Last 30 Days All All Photos Photos Forum Forums
  • RichSelian's Avatar
    Today, 20:41
    RichSelian replied to a thread 2019-nCoV in The Off-Topic Lounge
    i brought hundreds of face masks (rather cheap in china now, about $0.2 per one) and a lot of alcohol spray home from carrefour today. all goods are supplied as normal, no looting. i thinks china is winning the coronavirus war. no more native people get infected for almost a week. all new cases come from immigration.
    31 replies | 1788 view(s)
  • kaitz's Avatar
    Today, 12:07
    kaitz replied to a thread Paq8pxd dict in Data Compression
    True, i did not realize it myself. Again was some long time ago set as to reduce memory usage. Also made quick test. Only matchmodels active: 174008558 250mb (buf) 172560820 1gb (buf)
    746 replies | 288261 view(s)
  • suryakandau@yahoo.co.id's Avatar
    Today, 03:15
    it is fork from cmix v7 because from ltcb site the fastest version of cmix is cmix v7....the result of xml file for cmixHP3 is: 5345280 bytes -> 267754 bytes in 2038.38 s. cross entropy: 0.401 enwik9 on running
    13 replies | 997 view(s)
  • Shelwien's Avatar
    Today, 02:27
    Shelwien replied to a thread Paq8pxd dict in Data Compression
    I found a suspicious thing: bufn.setsize(0x10000); if (level>=9) buf.setsize(0x10000000); //limit 256mb else buf.setsize(MEM()*8); Do I read it right and paq8pxd uses 256mb buffer for enwik9 here?
    746 replies | 288261 view(s)
  • Shelwien's Avatar
    Today, 02:18
    > But how many bits can this shaving the end save you really? ~128 bits at most, no? Well, if you're using a rangecoder with 128 bit precision, then it could be 128 bits I guess. Common rc overhead is like this: - pad at start to avoid extra condition on carry flush: 1 byte - decoder lookahead bytes flushed to avoid tail checks: 2-4 bytes depending on rc renorm threshold - EOF/filesize coding: 3-4 bytes But its also possible to find some use for the value in the final interval - filesize (can be used instead of EOF) or crc % range maybe. Its also possible to save any (variable) number of tail bytes, when its likely for the model to output all 0s or all 1s at the end (which happens for some data types, eg. executables tend to have long zero padding at the end). > Is coding it worth it? For slow coders (like paq or cmix or nncp) it makes sense to have entropy coder with lowest possible redundancy - its kinda dumb when you save 5 bytes by adding a whole new submodel, while your rangecoder keeps encoding a megabyte of zeroes to 400 bytes because of low precision: https://encode.su/threads/2515-mod_ppmd?p=51672&viewfull=1#post51672 But for speed-optimized coders it may be better to keep the overhead, because extra checks are not worth it. > After Arithmetic Coding is finished the complete file, > you end up between the final high and low bounds. > You can shave off a few bits at the end. Ex. 0.451-0.372 becomes 0.38. The volume of encoded information is usually not bit-aligned, so its usually impossible to completely avoid any redundancy. I have some low-overhead rangecoders here: https://encode.su/threads/3084-Simple-rangecoder-with-low-redundancy-on-short-strings?p=59643&viewfull=1#post59643
    165 replies | 8181 view(s)
  • Self_Recursive_Data's Avatar
    Today, 01:00
    Question. After Arithmetic Coding is finished the complete file, you end up between the final high and low bounds. You can shave off a few bits at the end. Ex. 0.451-0.372 becomes 0.38. But how many bits can this shaving the end save you really? ~128 bits at most, no? ​Is coding it worth it?
    165 replies | 8181 view(s)
  • Darek's Avatar
    Today, 00:30
    Darek replied to a thread Paq8pxd dict in Data Compression
    125'752'479 - enwik9_1423 -x15 by Paq8pxd_v74_AVX2 125'797'519 - enwik9_1423 -x15 by Paq8pxd_v78_AVX2 - slightly worse than paqpxd v74
    746 replies | 288261 view(s)
  • Shelwien's Avatar
    Yesterday, 03:57
    Shelwien replied to a thread cmixHP in Data Compression
    here, gcc 8.2. Seems like a patch of cmix_v7.
    13 replies | 997 view(s)
  • suryakandau@yahoo.co.id's Avatar
    Yesterday, 03:46
    cmixHP2 in this version i don't use lstm model.. the result is worse but for enwik9 benchmark it needs <10 gb RAM and the time is <= 100 hours 1000000000 bytes -> 149086410 bytes in 321779.79 s. this is the source and the binary compiled with dev c++. maybe someone could compiled it using mingw please ? thank you
    13 replies | 997 view(s)
  • Darek's Avatar
    26th March 2020, 22:28
    Darek replied to a thread Paq8pxd dict in Data Compression
    Other enwik8 scores: 16'316'789 - enwik8 -s8 by Paq8pxd_v77_AVX2 15'967'512 - enwik8 -s15 by Paq8pxd_v77_AVX2 16'255'214 - enwik8 -x8 by Paq8pxd_v77_AVX2 15'901'484 - enwik8 -x15 by Paq8pxd_v77_AVX2 15'856'824 - enwik8.drt -x15 by Paq8pxd_v77_AVX2 16'291'281 - enwik8 -s8 by Paq8pxd_v78_AVX2 15'941'450 - enwik8 -s15 by Paq8pxd_v78_AVX2 16'231'687 - enwik8 -x8 by Paq8pxd_v78_AVX2 15'877'659 - enwik8 -x15 by Paq8pxd_v78_AVX2 15'852'312 - enwik8.drt -x15 by Paq8pxd_v78_AVX2 - drt got smaller improvement than pure file however it still provides to best score ever for paq8pxd series! enwik9 estimate = 125'802'xxx - very close to paq8pxd v74!
    746 replies | 288261 view(s)
  • Romul's Avatar
    26th March 2020, 15:13
    Is your algorithm somehow related to the golden ratio? I have an algorithm that can theoretically compress any input stream. And this algorithm has some relation with the so-called god number (1,618...). https://en.wikipedia.org/wiki/Golden_ratio But, this requires verification. So far, everything is only in the form of equations. I got these equations about five years ago, and tried to better understand their properties. PS. I write through an online translator, so my text may not look very correct.
    24 replies | 1358 view(s)
  • lz77's Avatar
    26th March 2020, 13:20
    blzpack seems only LZ-type compressor (at least I see in enwik8.blz literals like 'mediawiki xmlns=' and '"http://www.'). blzpack -2 -b1024m enwik8 gives ~ the same ratio as my program (while this is a prototype program for debugging and improving my algorithm), but my program is much faster than blzpack -1... I want to get time/ratio similar to zstd v1.4.4 for win64: timer.exe zstd.exe -1 -f --no-progress enwik8 Both ratio & compression time will be the same but decompression time at this time 15% worse. :) I want to find a buyer(s) for my algorithms/sources. Maybe I will sell it as shareware...
    8 replies | 497 view(s)
  • Darek's Avatar
    26th March 2020, 12:04
    Darek replied to a thread Paq8pxd dict in Data Compression
    First enwik scores: 16'319'686 - enwik8 -s8 by Paq8pxd_v75_AVX2 16'314'392 - enwik8 -s8 by Paq8pxd_v76_AVX2 = -6'300 bytes 16'316'789 - enwik8 -s8 by Paq8pxd_v77_AVX2 = +2'400 bytes 16'291'281 - enwik8 -s8 by Paq8pxd_v78_AVX2 = -25'500 bytes -> good improvement!
    746 replies | 288261 view(s)
  • Jyrki Alakuijala's Avatar
    26th March 2020, 05:55
    I tried reaching out to his relatives. I didn't get a response back.
    305 replies | 315840 view(s)
  • Darek's Avatar
    26th March 2020, 02:19
    Darek replied to a thread Paq8pxd dict in Data Compression
    paq8pxd_v78 scores for 4 corpuses => another version with all 4 records for paq8pxd serie!
    746 replies | 288261 view(s)
  • birdie's Avatar
    26th March 2020, 01:10
    Any updates?
    305 replies | 315840 view(s)
  • Jyrki Alakuijala's Avatar
    25th March 2020, 21:09
    Single-threaded brunsli is 1 % weaker and 2.5x faster. Brunsli covers more use cases in the exotic jpeg land. Brunsli is single-threaded, lepton multi-threaded. Brunsli variation in jpeg xl may allow multi-threaded coding in 256x256 tiles.
    34 replies | 9827 view(s)
  • Shelwien's Avatar
    25th March 2020, 17:41
    You can see an example here: https://encode.su/threads/3128-brunsli-jpeg-recompressor?p=60448&viewfull=1#post60448 Lepton sometimes has faster decoding (due to MT) and better compression (not at the same time as MT).
    34 replies | 9827 view(s)
  • Jyrki Alakuijala's Avatar
    25th March 2020, 17:35
    gipfeli does something that is in between of Huffman and not doing it. It might qualify to your requirements (or not)...
    8 replies | 497 view(s)
  • Darek's Avatar
    25th March 2020, 15:14
    Darek replied to a thread Paq8pxd dict in Data Compression
    paq8pxd_v78 scores on my testset. In general no big changes. Some improvements for textual files. Some loses for bigger files.
    746 replies | 288261 view(s)
  • CompressMaster's Avatar
    25th March 2020, 15:14
    Here.
    8 replies | 497 view(s)
  • CompressMaster's Avatar
    25th March 2020, 15:11
    Here.
    8 replies | 497 view(s)
  • Darek's Avatar
    25th March 2020, 15:08
    Darek replied to a thread Paq8pxd dict in Data Compression
    enwik8/9 scores for paq8pxd_v76: 15'928'916 - enwik8 -x15 by Paq8pxd_v74_AVX2 125'752'479 - enwik9_1423 -x15 by Paq8pxd_v74_AVX2 15'912'509 - enwik8 -x15 by Paq8pxd_v75_AVX2 125'761'484 - enwik9_1423 -x15 by Paq8pxd_v75_AVX2 15'899'380 - enwik8 -x15 by Paq8pxd_v76_AVX2 125'974'773 - enwik9_1423 -x15 by Paq8pxd_v76_AVX2 - hmmm, there is an 0,17% loss to v75 version, 0.18% to v74 version. The v74 is still the best! paq8pxd v77 and v78 tests ongoing.
    746 replies | 288261 view(s)
  • well's Avatar
    25th March 2020, 11:11
    i have been testing phda9 $./phda9 C9 enwik9 out is it right? C9 - is it right option for enwik9?as you can see behind phda9 is not universal compressor, so i ask for right using instructions! $./phda9 C enwik9 out 91% Segmentation fault $ so C9 - is strictly for enwik9?
    87 replies | 30749 view(s)
  • pklat's Avatar
    25th March 2020, 11:03
    how does it compare to brunsli?
    34 replies | 9827 view(s)
  • necros's Avatar
    25th March 2020, 10:06
    Irfanview plugin would be great )
    34 replies | 9827 view(s)
  • lz77's Avatar
    25th March 2020, 09:19
    Pardon, I'm not an English speaker. I mean the program after LZ compression does not make Huffman/arithmetic/other compression. Where I can download blzpack.exe (or brieflz)?
    8 replies | 497 view(s)
  • compgt's Avatar
    24th March 2020, 20:48
    So, did you delete my sent messages again?! It's *my* Sent Folder, not yours!! You don't know anything...!! As i've told you before, this forum is not yours, even if you're the one paying for it. It's public. It has become public, it's existence maintained by the patronage of the forum members. I guess this "Shelwien" handle is now or still owned by British MI5, as before when i was 'M' and JB. But maybe you're not the first Shelwien.
    305 replies | 315840 view(s)
  • Shelwien's Avatar
    24th March 2020, 20:39
    I think its a really bad topic for your spam, just stop it.
    305 replies | 315840 view(s)
  • compgt's Avatar
    24th March 2020, 20:20
    All sent items gone in my Sent Folder, again! I messaged some of you here again, about Sami Runsas' brain tumor news in December 2012, because something bad happened to me too that month. Sami was a very good programmer. I just content myself now thinking that he wasn't being heckled or deterred by some computer or data storage companies' "mafia group" while writing his superior compression programs. What's happening to the world?! Manipulated by power groups for gain or profit.
    305 replies | 315840 view(s)
  • Jyrki Alakuijala's Avatar
    24th March 2020, 17:51
    What is a squeeze?
    8 replies | 497 view(s)
  • compgt's Avatar
    24th March 2020, 16:58
    My lzuf compressor compresses enwik8 to around 38MB maybe in 40s in my old dual-core machine in 2008, decompresses in 3s. I don't have it anymore but only lzuf2 and lzuf5. It is here in The Data Compression Guide's LZ77 page, plus the early LZT (or LZ-Tamayo) in 2008. The timings in LTCB for lzgt3a were somewhat different than mine: https://sites.google.com/site/datacompressionguide/lz77
    8 replies | 497 view(s)
  • jibz's Avatar
    24th March 2020, 14:53
    Timing depends on the CPU and RAM of course, but I am sure some of them can reach those goals, for instance BriefLZ gets: blzpack.exe -v -3 enwik8 enwik8.blz in 100000000 out 39508633 ratio 39% time 1.85 Oodle and lzturbo will likely do even better. Edit: though BriefLZ may not fit into pure LZ as it uses a universal code.
    8 replies | 497 view(s)
  • RAM's Avatar
    24th March 2020, 13:55
    Followers of Babushkin http://lurkmore.to/%D0%90%D0%BB%D0%B5%D0%BA%D1%81%D0%B5%D0%B9_%D0%91%D0%B0%D0%B1%D1%83%D1%88%D0%BA%D0%B8%D0%BD But seriously, I have a revolutionary algorithm. And it has nothing to do with the even steeper Lena.png algorithm. :D
    1 replies | 129 view(s)
  • lz77's Avatar
    24th March 2020, 13:49
    I did benchmark LZ only compressors (which do not squeeze after LZ-type compression) and looked at LZ4, Lizard, ulz, lzop, and quicklz. None of them can compress enwik8 on 1 core in <= 40Mb in real time (< 2 sec.) If you saw such LZ-type compressors, show me where I can download binaries for Windows, thanks. Where I can download snappy.exe?
    8 replies | 497 view(s)
  • kaitz's Avatar
    24th March 2020, 12:57
    ​https://tjournal.ru/stories/128216-moshenniki-3-0-kak-ne-popastsya-na-udochku-novogo-pokoleniya-prestupnikov-v-sfere-it
    1 replies | 129 view(s)
  • kaitz's Avatar
    24th March 2020, 12:26
    kaitz replied to a thread Paq8pxd dict in Data Compression
    paq8pxd_v78 - Change wordModel1,recordmodel l.pak,k.wad not fixed for now. ​ This change mostly will work only with internal wrt. drt processed files will not benefit from it. Most compression is on plain text files and comes from wordmodel. enwik8 -s8 should be 19kb smaller.
    746 replies | 288261 view(s)
  • LawCounsels's Avatar
    23rd March 2020, 21:01
    Yes very rare exceptions If input symbols ratio is 30 vs 70 ( or 50 vs 50 ) , then overall in general Output symbol ratios in general will be exact same 30 vs 70 ( or 50 vs 50 )
    15 replies | 737 view(s)
  • Jarek's Avatar
    23rd March 2020, 20:49
    N to N bit cipher is a bijection - for every input sequence there exists exactly one output sequence. Hence there exists e.g. input leading to output being all zeros, so YES - output frequencies can be very different from input. However, these are relatively rare cases, on average and typically ( https://en.wikipedia.org/wiki/Typical_set ) they have the same i.i.d. Pr(0)=Pr(1)=1/2 frequencies.
    15 replies | 737 view(s)
  • Jarek's Avatar
    23rd March 2020, 20:42
    Indeed Kraken is just LZ+Huffman/tANS - while it is a great general purpose lossless compressor, a specialized lossy texture compressor should easily beat it at this task, like http://gamma.cs.unc.edu/GST/
    8 replies | 843 view(s)
  • Scope's Avatar
    23rd March 2020, 20:17
    Xbox Series X’s BCPack Texture Compression Technique Reportedly Better than the PS5’s Kraken https://gamingbolt.com/xbox-series-xs-bcpack-texture-compression-technique-reportedly-better-than-the-ps5s-kraken
    8 replies | 843 view(s)
  • LawCounsels's Avatar
    23rd March 2020, 20:11
    Can there even be one simplest N bits to N bits Block Cipher .exe url link , where the output symbol frequencies does not EXACT preserve input symbol frequencies ?
    15 replies | 737 view(s)
  • Shelwien's Avatar
    23rd March 2020, 19:18
    Actually 1 bit is hard, since OS i/o is bytewise. In fact, 0 is easier, since we can assign it to any one specific file, which would have to be embedded in the decompressor. It would be even a perfectly valid compressor otherwise, like "filesize=0 ? output enwik8; otherwise decompress file data with cmix".
    42 replies | 3483 view(s)
  • SvenBent's Avatar
    23rd March 2020, 16:33
    Still not answered on a working decompressor. with no decompressor you results are pathetic. I can compress anything to 1 bit if decompression is not needed. Put up or shut up
    244 replies | 92339 view(s)
  • SvenBent's Avatar
    23rd March 2020, 16:32
    So it doesn't work?
    24 replies | 1358 view(s)
  • SvenBent's Avatar
    23rd March 2020, 16:17
    Amatuer. I can compress enwikk9 to 1 bit if i dont care about decompression. Once decompession is working then we can talk. You compression results are worthless without decompression
    42 replies | 3483 view(s)
  • Jarek's Avatar
    23rd March 2020, 13:33
    In N constrained bits you can hide N/rate bits of information, where rate is average Shannon entropy of these constraints. If you want to store N bits of information in N bits, then there is no place for constraints, you can only apply some 2^N -> 2^N bijection on them.
    15 replies | 737 view(s)
  • LawCounsels's Avatar
    23rd March 2020, 13:08
    Yes But NOT applicable when N bits input N bits output strictly
    15 replies | 737 view(s)
  • Jarek's Avatar
    23rd March 2020, 08:04
    Yes, standard ciphers should produce looking completely random sequence of bits: i.i.d. Pr(0)=Pr(1)=1/2, carrying 1 bit/bit (rate=1). But if needed, e.g. its output can be transformed into sequence of any chosen statistical model (using reversed entropy coding or Kuznetsov-Tsybakov) - not making it easier to decrypt, at cost of longer encoding sequence: the original length has to be divided by rate.
    15 replies | 737 view(s)
  • LawCounsels's Avatar
    23rd March 2020, 01:21
    Do you have URL s to such Block Cipher .exe ( simplest ones , N input N output ) to quick easy verify this so ( produces white noise binary output distributions) for any skewed input N bits It had seemed theoretic that Block Cipher could only maintain exact same binary output symbol frequencies as binary input symbol frequencies . Suggest one such white noise output symbols distribution simple Block Cipher scheme It had also seemed the original input's 2-bits-units symbols frequencies is NOT exact preserved, except only so far as falling within the general usual fluctuation chances dictated by the exact same output binary symbol distributions preserved
    15 replies | 737 view(s)
  • Piotr Tarsa's Avatar
    23rd March 2020, 00:41
    I think that the widely used encryption algorithms tend to make the output distribution near uniform, effectively prohibiting compression. They are also as unpredictable as possible, because predictable encryption algorithms are by definition weak. Therefore you cannot rely on any particular structure of encrypted file.
    15 replies | 737 view(s)
  • pacalovasjurijus's Avatar
    22nd March 2020, 18:50
    We can compress and compress information. I can compress any file to <=300000 bytes.
    42 replies | 3483 view(s)
  • LawCounsels's Avatar
    22nd March 2020, 17:03
    Supposing of 49 1s 51 0s input , 29 1s 21 0s in 1st 50 bits & 20 1s 30 0s in 2nd 50 bits ....will block cipher always produce white noise e.g. 24 1s 26 0s in 1st 50 bits & e.g. 25 1s 25 0s in 2nd 50 bits ? Does Block Cipher always produce exact same 49 vs 51 output symbols frequencies ? If so , would input's existing 2-bits-unit symbol frequencies ( ie 00s 01s 10s 11s) be ALSO exact same preserved in Cipher Block's output of the 49 vs 51s single bits symbol frequencies ?
    15 replies | 737 view(s)
  • suryakandau@yahoo.co.id's Avatar
    22nd March 2020, 16:17
    it is because the lstm model in cmix...so it can not beat phda9 now...if use another technique maybe it can faster...or open sourced of phda9 n improve the phda9 better because the baseline is phda9
    13 replies | 997 view(s)
  • Jarek's Avatar
    22nd March 2020, 13:58
    This is Kuznetsov-Tsybakov problem, you can see it as that we have some 'freedom bits' chosen to satisfy the constraint, and some complex bijection T: encoder searches for 'freedom bits' such that T(encoded bits|freedom bits) satisfies the constraints, then decoder just inverts T and discards 'freedom bits'. It would reach the theoretical limit, but requiring exponential computation cost ... however, approximations turn out quite good, see the paper.
    15 replies | 737 view(s)
  • LawCounsels's Avatar
    22nd March 2020, 13:40
    In simple few words how would encoder needs to encode ( knowing the defective cells ) so decoder can similar correct decode here ( not knowing the defective cells ) ?
    15 replies | 737 view(s)
  • Darek's Avatar
    22nd March 2020, 13:39
    Darek replied to a thread Paq8pxd dict in Data Compression
    some enwik scores gathered: 16'319'686 - enwik8 -s8 by Paq8pxd_v75_AVX2 15'976'838 - enwik8 -s15 by Paq8pxd_v75_AVX2 16'260'265 - enwik8 -x8 by Paq8pxd_v75_AVX2 15'912'509 - enwik8 -x15 by Paq8pxd_v75_AVX2 15'859'187 - enwik8.drt -x15 by Paq8pxd_v75_AVX2 125'761'484 - enwik9_1423 -x15 by Paq8pxd_v75_AVX2 126'074'749 estimated - enwik9_1423.drt -x15 by Paq8pxd_v75_AVX2 16'314'392 - enwik8 -s8 by Paq8pxd_v76_AVX2 - tested by Sportman 15'965'102 - enwik8 -s15 by Paq8pxd_v76_AVX2 - tested by Sportman 16'253'017 - enwik8 -x8 by Paq8pxd_v76_AVX2 15'899'380 - enwik8 -x15 by Paq8pxd_v76_AVX2 15'856'800 - enwik8.drt -x15 by Paq8pxd_v76_AVX2 16'316'789 - enwik8 -s8 by Paq8pxd_v77_AVX2 - tested by Sportman 15'967'512 - enwik8 -s15 by Paq8pxd_v77_AVX2- tested by Sportman 16'255'214 - enwik8 -x8 by Paq8pxd_v77_AVX2 15'901'484 - enwik8 -x15 by Paq8pxd_v77_AVX2 - tested by Kaitz 15'856'824 - enwik8.drt -x15 by Paq8pxd_v77_AVX2 125'65x'xxx estimated - enwik9_1423 -x15 by Paq8pxd_v76_AVX2
    746 replies | 288261 view(s)
  • Jarek's Avatar
    22nd March 2020, 13:11
    Sure, enforcing some distribution (constraints e.g. statistical) on encoding sequence is at cost of reduced information content - "rate" in example above saying how many bits/constrained bit can we hide there. Theoretical limit is just Shannon entropy: enforcing p probability of '1' in sequence of 0/1, we can store at most h(p) = -p lg(p) - (1-p)lg(1-p) bits/value. It is obtained with entropy coder, but requiring that both encoder and decoder know the statistical model (e.g. 'p' here, image to resemble for QR code). This requirement can be weakened to only encoder knowing the statistical model ('p' or image here), still approaching the same theoretical boundary (at higher computational cost for encoder) - generalization of Kuznetsov-Tsybakov problem in github and paper above.
    15 replies | 737 view(s)
  • Bulat Ziganshin's Avatar
    22nd March 2020, 12:58
    any sane cipher produces white nois with equal distribution of byte values
    15 replies | 737 view(s)
  • LawCounsels's Avatar
    22nd March 2020, 12:36
    It would appear input symbols of 49 1s 51 0s would not possible encrypt into 100 bits of 4 eg40 1s 60 0s ?
    15 replies | 737 view(s)
  • Jarek's Avatar
    22nd March 2020, 12:22
    Entropy coder (accurate: AC/ANS), if replacing encoder with decoder, can transform a looking random sequence of bits (e.g. encrypted data) into a sequence of symbols/bits of practically any chosen statistical model ... not making it easier to decrypt. Something like that can be done for steganography/watermarking ( https://en.wikipedia.org/wiki/Steganography ) applications, e.g. to encrypt data as QR code (0/1 sequence) resembling a chosen image (grayness of pixel determines probability of using '1' in code for this pixel): http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=7348695 Examples of such codes obtained with https://github.com/JarekDuda/PRcodes/ , rate is the number of hidden bits/pixel:
    15 replies | 737 view(s)
  • necros's Avatar
    22nd March 2020, 12:16
    OK genius where is binary to test?
    24 replies | 1358 view(s)
  • LawCounsels's Avatar
    22nd March 2020, 11:41
    Can Block Cipher ( n bits input n bits output ) ever produces different distributions of symbols from input distribution of symbols ?
    15 replies | 737 view(s)
  • Self_Recursive_Data's Avatar
    22nd March 2020, 01:26
    The neural predictor's predictions have probabilities, and the one you want to pick decides the next part of the decimal string to make during Arithmetic Coding. After the decimal string is made once arrives at the end of the file, it now transforms the decimal string into a bin file (binary) by changing it to a "number" in binary form. It's like Huffman Coding but does a near-perfect job of what Huffman Coding tries to do. My AC wasn't faulty in the end, it was just an input embedding issue, no real issue. For the first 100,000 bytes of enwiki8 Green gets it to 29,108 bytes Mine gets it to 28,572 bytes For the first 1,000,000 bytes of enwiki8 Green gets it to 253,800 bytes Mine gets it to 251,235 bytes For the first 10,000,000 bytes of enwiki8 Green gets it to 2,331,508 bytes Mine gets it to 2,319,337 bytes Note my parameters can still be tweaked to get that lower a bit.
    165 replies | 8181 view(s)
  • EyeCon's Avatar
    21st March 2020, 22:36
    EyeCon replied to a thread zpaq updates in Data Compression
    Hello, I have a question about zpaq. I have a use case where different people must add directories with the same name, but differing contents to the same archive. All files that get added are necessary (we are basically deleting nothing). zpaq would be ideal for this because of the great deduplication and compression. When different people add directories with the same name but different files, a normal extract only gives the latest set of added files because zpaq thinks the rest got deleted from the directory. That leads to my question: is there a way to use zpaq in an append-only mode (journaling which records no delete actions) which still has compression and deduplication? I know I can just use -all switch and deal with the files myself, but I was hoping for an easier way. I also want to thank you for the great tool. I use it daily, mainly because of the deduplication. Sorry for any mistakes, I'm not a native speaker.
    2528 replies | 1090321 view(s)
  • Darek's Avatar
    21st March 2020, 18:02
    Darek replied to a thread Paq8pxd dict in Data Compression
    Scores of 4 corpuses for paq8pxd v76 and v77. Despite my testset worse scores, for all 4 corpuses both paq8pxd v76 and v77 got the best scores and very good improvemets! For Silesia corpus there are 123KB less on paq8pxd v77 than paq8pxd v75!
    746 replies | 288261 view(s)
  • uhost's Avatar
    21st March 2020, 16:39
    which stage is taken for compression, is that binary ? hex ? decimal ? or string ? can you explain :_sad2:
    165 replies | 8181 view(s)
  • Shelwien's Avatar
    21st March 2020, 14:48
    Shelwien replied to a thread Brotli in Data Compression
    I added gcc exe to the post above without archive. Virustotal dislikes intelc version for some reason.
    228 replies | 78036 view(s)
  • well's Avatar
    21st March 2020, 14:12
    well replied to a thread cmixHP in Data Compression
    i'm running :)
    13 replies | 997 view(s)
  • Sportman's Avatar
    21st March 2020, 13:10
    Sportman replied to a thread cmixHP in Data Compression
    No, that takes 2 x 2 weeks. Probably: 122,047,657 bytes
    13 replies | 997 view(s)
More Activity