Results 1 to 30 of 30

Thread: Sloot Compression

  1. #1
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts

    Sloot Compression

    Does anyone know about this:

    http://jansloot.telcomsoft.nl/Source...m#.WTrPdcYlGlG

    Would this be potentially extensible to generic data?

  2. #2
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    733
    Thanked 659 Times in 354 Posts
    Hollywood uses this scheme for a years. In short, any movie can be condensed to one of 10 or so typical storylines. So, each year they just get one of patented storylines, and "extract" it using random set of actors. But this is already large industry, so if you will buy patent from Hollywood, it will hardly cost less than a 1mm bucks. Are your VC is ready for that? And they probably don't have Delphi API, so your VC will need to replace you with better developer. So think twice before going into that!

  3. The Following 3 Users Say Thank You to Bulat Ziganshin For This Useful Post:

    boxerab (19th June 2017),diskzip (9th June 2017),snowcat (11th June 2017)

  4. #3
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Bulat Ziganshin View Post
    Hollywood uses this scheme for a years. In short, any movie can be condensed to one of 10 or so typical storylines. So, each year they just get one of patented storylines, and "extract" it using random set of actors. But this is already large industry, so if you will buy patent from Hollywood, it will hardly cost less than a 1mm bucks. Are your VC is ready for that? And they probably don't have Delphi API, so your VC will need to replace you with better developer. So think twice before going into that!
    Haha!
    Last edited by diskzip; 10th June 2017 at 03:46.

  5. #4
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,134
    Thanks
    179
    Thanked 921 Times in 469 Posts
    Bulat is joking, because it was a fraud.
    Such people do appear even on this forum now and then - see http://www.faqs.org/faqs/compression...section-8.html
    Kinda hard to take this kind of stories seriously, once you know how it all works.

  6. The Following User Says Thank You to Shelwien For This Useful Post:

    diskzip (9th June 2017)

  7. #5
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Shelwien View Post
    Bulat is joking, because it was a fraud.
    Such people do appear even on this forum now and then - see http://www.faqs.org/faqs/compression...section-8.html
    Kinda hard to take this kind of stories seriously, once you know how it all works.
    I get it now.
    Last edited by diskzip; 10th June 2017 at 03:46.

  8. #6
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    772
    Thanks
    63
    Thanked 270 Times in 190 Posts
    Quote Originally Posted by Shelwien View Post
    Bulat is joking, because it was a fraud.
    In this case Jan Sloot and his SDCS invention was not a fraud.

    It could store (in 1999) 90 movies (DVD size) on a single 128Kb chip card and play back it (till 16 together with fast forward and backwards) from a small box (two sigar boxes size) with <400MB RAM/ROM and some fast CPU's in it connected to any monitor.

    Dutch ABN-AMRO Bank wanted to invest (signed June 23, 1999) in Dipro Research N.V. (June 5, 1999 renamed to FifthForce) 52,5mln Dutch Guilder in four trances (12,5mln, 10mln, 10mln, 20mln) so the company could do in 2001 or earlier an IPO at the Nasdaq for 3bln dollar.

    There where two problem every movie needed to be encoded and stored in that RAM/ROM so a consumer needed a box replacement once a while to decode the newest movies key code from the chip card.

    The other problem was that Jan Sloot was found death in his private back yard in Nieuwegein (The Netherlands) Sunday 11 July round 12 AM two days after he gave Tom Perkins a demonstration at Roel Pieper's private home in Bloemendaal (The Netherlands) and two days before Jan Sloot wanted to give the signed contracts and source code to the notary office, after that Jan Sloot shall receive the first 2,5mln Dutch Guilder (by selling 1% of the stocks). His room (normal full with all his papers) was total cleaned and nobody knows who did that, there where still some working demo devices but reverse engineering had no results in finding out the used algorithms.

  9. The Following User Says Thank You to Sportman For This Useful Post:

    diskzip (10th June 2017)

  10. #7
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    733
    Thanked 659 Times in 354 Posts
    yeah, Wikipedia cannot be wrong. Once i was too late to my work, so I quickly edited Gravity article and then flyed directly to my desk

  11. The Following 2 Users Say Thank You to Bulat Ziganshin For This Useful Post:

    diskzip (10th June 2017),snowcat (11th June 2017)

  12. #8
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    772
    Thanks
    63
    Thanked 270 Times in 190 Posts
    Everybody is free to think what he/she want but this inventing was real and working demo's where done also in US at big companies even without inventor.

    The first serious demo was March 4, 1999 at the head office of Philips in Amsterdam (The Netherlands) where Roel Pieper saw the demo and left Philips afterwards to join Jan Sloot's investors (mainly Marcel Boekhoorn) to market the inventing mainly in the US.

    There is still a document online compiled (by Marcel Boekhoorn) for this demo at Philips:
    https://web.archive.org/web/20060305..._info_memo.pdf

  13. The Following User Says Thank You to Sportman For This Useful Post:

    diskzip (10th June 2017)

  14. #9
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Sportman View Post
    Everybody is free to think what he/she want but this inventing was real and working demo's where done also in US at big companies even without inventor.

    The first serious demo was March 4, 1999 at the head office of Philips in Amsterdam (The Netherlands) where Roel Pieper saw the demo and left Philips afterwards to join Jan Sloot's investors (mainly Marcel Boekhoorn) to market the inventing mainly in the US.

    There is still a document online compiled (by Marcel Boekhoorn) for this demo at Philips:
    https://web.archive.org/web/20060305..._info_memo.pdf
    That all seems pretty legit to me.

    Definitely way more legitimate than any of that Oodle stuff!

    Let's hope the Oodle HQ doesn't get gunned down by the mafia before they are able to churn out an eval SDK!

  15. #10
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    733
    Thanked 659 Times in 354 Posts
    they can call mafia from GTA if you will continue to pursue them. Guess why noone tell bad word about them?

  16. The Following User Says Thank You to Bulat Ziganshin For This Useful Post:

    diskzip (10th June 2017)

  17. #11
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Bulat Ziganshin View Post
    yeah, Wikipedia cannot be wrong. Once i was too late to my work, so I quickly edited Gravity article and then flyed directly to my desk
    Wikipedia is absolutely evil as far as I'm concerned.

    They regularly censor articles, based on unpublished criteria. You cannot debate by rational argument. Even their own guidelines indicate that a bureaucratic process is involved (procedural correctness over factual accuracy).

    The entire encyclopedia is neither free, nor can it be edited by anyone. They have even driven their own editors to suicide.

    I have read elsewhere that they also regularly censor articles which go against the interests/viewpoints of the (former) British Empire.

    Last but not least...the Wikimedia Foundation has over 100 million in assets (I saw another source reporting this figure as 300 million, but I'll err on the side of caution). They still do public fundraising like crazy, making it sound like they're about to get shut down by their greedy web host any time, while they are the greedy ones. The foundation itself is super rich, and absolutely unaccountable for anything done by anyone involved in Wikipedia.

    I wonder if this entire project is just a thinly veiled brainwashing effort, one that holds the brain-washers free of any accountability, by virtue of the fact that ostensibly the whole system runs on donations and volunteer efforts alone, to all sights and sounds irrespective of the people making the big bucks behind it all?

    Food for thought.

  18. #12
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,134
    Thanks
    179
    Thanked 921 Times in 469 Posts
    > On this chipcard, the film’s 130 thousand frames (90 minutes x 60 seconds x 24 frames; one frame is 4 snippets for the red, blue, green, and audio components)
    > are not stored as such, but indexed by 130 thousand (frames) x 4 (components) x 24-bit numbers, costing about 1.5 Mb.

    There're obviously more than 2^24=16M different video frames, this approach would only work with a few specific video snippets.
    And of course, in a fair evaluation, the size of that frame database should be included, you can't talk about index like its the whole movie.
    Then, how many bits is really necessary to write an index of _any_ possible video frame is determined by real compression algorithms.

    As to the script idea - sure, in theory its possible to reverse-engineer a movie, build 3D models of all objects in it, and reconstruct a script
    that determines their movements. But its not a matter of "genius" or guessing the right formula.
    Even plain 3D-rendering software is very large and complicated, and requires many people to make.
    While this kind of "video codec" would require same software on decoding side - with realtime rendering capability even.
    And encoding side would be 10x more complex than that.
    Its clearly not feasible for a single person to make even in 2017.

  19. #13
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Shelwien View Post
    > On this chipcard, the film’s 130 thousand frames (90 minutes x 60 seconds x 24 frames; one frame is 4 snippets for the red, blue, green, and audio components)
    > are not stored as such, but indexed by 130 thousand (frames) x 4 (components) x 24-bit numbers, costing about 1.5 Mb.

    There're obviously more than 2^24=16M different video frames, this approach would only work with a few specific video snippets.
    And of course, in a fair evaluation, the size of that frame database should be included, you can't talk about index like its the whole movie.
    Then, how much bits is really necessary to write an index of _any_ possible video frame is determined by real compression algorithms.

    As to the script idea - sure, in theory its possible to reverse-engineer a movie, build 3D models of all objects in it, and reconstruct a script
    that determines their movements. But its not a matter of "genius" or guessing the right formula.
    Even plain 3D-rendering software is very large and complicated, and requires many people to make.
    While this kind of "video codec" would require same software on decoding side - with realtime rendering capability even.
    And encoding side would be 10x more complex than that.
    Its clearly not feasible for a single person to make even in 2017.
    How about the comment at the end of the blog post - the stuff with respect to analog components?

    That might be an interesting angle to pursue, as otherwise, clearly, digitally, we're not really thinking out of the box at all.

  20. #14
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by Bulat Ziganshin View Post
    they can call mafia from GTA if you will continue to pursue them. Guess why noone tell bad word about them?
    Don't shoot!

    (ducking)

  21. #15
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,134
    Thanks
    179
    Thanked 921 Times in 469 Posts
    Analog solutions are as the name says.
    Sometimes we can find a similarity in physical processes, so some financial models can be simulated with hydraulics, etc.
    https://en.wikipedia.org/wiki/MONIAC
    The benefit is that nature is massively parallel, so problems for which we can find a precise enough analog approximations can be solved faster than digital computers.
    In a way, quantum computing is the modern approach to that.

    However, at this point (since 198x or so), digital computers allow to solve this kind of problems better/cheaper.
    Mostly because only digital tech was evolving all this time. At some point analog tech could win again, I guess.

    Still, here its surely not the case, because even mechanically matching a 2D fotorobot/identikit to an image
    would be hard and annoying, all the more for 3D models.
    So, an analog model won't be any simpler than digital one, just in theory could beat it in speed due to natural parallelism.

    Also, there're plenty of hints (like that video index description) that it wasn't about analog computing.

  22. #16
    Member
    Join Date
    Nov 2014
    Location
    California
    Posts
    122
    Thanks
    36
    Thanked 33 Times in 24 Posts
    That all seems pretty legit to me.
    If it looks too good to be true ...
    If Sloot did indeed make the claim (as reported) to compress full length movies to 8 KB, this is simply not possible, even with a huge codec.

  23. #17
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,134
    Thanks
    179
    Thanked 921 Times in 469 Posts
    Its possible with a torrent, you just need internet connection :)

    Also, the idea to use similarity between different movies is valid.
    Surely we could compress movies much better, if there was a 100PB database of known movies to use as a reference.
    (Still not to 8kb or any other fixed size, of course).
    But that approach can't be used in practice because of copyright issues.

  24. #18
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    772
    Thanks
    63
    Thanked 270 Times in 190 Posts
    Quote Originally Posted by hexagone View Post
    If it looks too good to be true ...
    If Sloot did indeed make the claim (as reported) to compress full length movies to 8 KB, this is simply not possible, even with a huge codec.
    He did not, he transformed analog or digital input to numbers, then filled a matrix with this numbers and gave this matrix a unique number then he repeated that many times (4 or 5 times cascade) with the output of the last stage and stored or looked up the between values and output in a black box (his RAM/ROM), the last unique value (round 1Kb) was stored at the chip card.

    I guess his black box was a dictionary with all possible values he could split any input in plus a kind of Bloom filter where he could store only reference unique values and the next input book, audio, video stored did not damage the decoding of the input stored before. How longer the movies he wanted to store how more memory he needed, the amount of movies looked limited to the max unique combination of the key code.

    I just saw that Thomas Perkins https://en.wikipedia.org/wiki/Thomas..._(businessman) also wrote about Jan Sloot, Roel Pieper and the Firth Force in his book "Valley Boy" (starting at page 209):
    https://www.amazon.com/Valley-Boy-Ed.../dp/1592404030

  25. #19
    Member
    Join Date
    Nov 2014
    Location
    California
    Posts
    122
    Thanks
    36
    Thanked 33 Times in 24 Posts
    Quote Originally Posted by Sportman View Post
    He did not, he transformed analog or digital input to numbers, then filled a matrix with this numbers and gave this matrix a unique number then he repeated that many times (4 or 5 times cascade) with the output of the last stage and stored or looked up the between values and output in a black box (his RAM/ROM), the last unique value (round 1Kb) was stored at the chip card.

    I guess his black box was a dictionary with all possible values he could split any input in plus a kind of Bloom filter where he could store only reference unique values and the next input book, audio, video stored did not damage the decoding of the input stored before. How longer the movies he wanted to store how more memory he needed, the amount of movies looked limited to the max unique combination of the key code.

    I just saw that Thomas Perkins https://en.wikipedia.org/wiki/Thomas..._(businessman) also wrote about Jan Sloot, Roel Pieper and the Firth Force in his book "Valley Boy" (starting at page 209):
    https://www.amazon.com/Valley-Boy-Ed.../dp/1592404030
    My understanding is that he did. Other sources claimed 512 KB for a full length movie. For a 120 min movie, it means 73 B/sec or around 3 B/frame ... See the problem ?
    The analog/digital thing is a red herring IMO.
    Are you saying one different codec with a dictionary of vectors in needed per movie ? Otherwise, the dictionary has to be gigantic like Shelwien said.

  26. #20
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    772
    Thanks
    63
    Thanked 270 Times in 190 Posts
    Quote Originally Posted by hexagone View Post
    My understanding is that he did. Other sources claimed 512 KB for a full length movie. For a 120 min movie, it means 73 B/sec or around 3 B/frame ... See the problem ?
    Are you saying one different codec with a dictionary of vectors in needed per movie ? Otherwise, the dictionary has to be gigantic like Shelwien said.
    As far I know it was round 1KB end key for each movie. So 1KB key at a smartcard plugged in his magic box started 5 algorithms behind each other decoding the key to 5 RAM memories 12MB each in size (60MB total) in his older demo model for normal length movies or 74MB each in size (370MB total) in his newer demo model for longer length movies. There was also a reference ROM memory from unknown size where all the possible pixel values who could exist in a movie pixel where stored like color values. The five algorithms calculated with fast CPUs and 5 RAM memories (temp space) and one ROM memory the frames, rows, blocks, pixels and the output signal. The magic box was small and solid state there where no moving parts in.

  27. The Following User Says Thank You to Sportman For This Useful Post:

    diskzip (27th June 2017)

  28. #21
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    So I got the book "Valley Boy" by Tom Perkins where there's talk of Sloot.

    For your reading pleasure, I have scanned in the relevant chapter herein:

    Click image for larger version. 

Name:	page1.png 
Views:	198 
Size:	538.7 KB 
ID:	5403Click image for larger version. 

Name:	page2.jpg 
Views:	122 
Size:	180.8 KB 
ID:	5404Click image for larger version. 

Name:	page3.jpg 
Views:	109 
Size:	202.8 KB 
ID:	5405Click image for larger version. 

Name:	page4.jpg 
Views:	102 
Size:	194.3 KB 
ID:	5406Click image for larger version. 

Name:	page5.jpg 
Views:	98 
Size:	193.1 KB 
ID:	5407

    It seems to me the technology was entirely legitimate. I don't think Tom would be fooled easily.

    Anyone who doubts this accomplishment doesn't have faith in their own potential to achieve greatness.

    Although, I would love to know why and where the bad karma got compounded, of course.

    The death of the main character is well explained, but why the stuff after?

    It's a scary story...as real as life itself!

  29. The Following User Says Thank You to diskzip For This Useful Post:

    Sportman (14th October 2017)

  30. #22
    Member
    Join Date
    Nov 2014
    Location
    California
    Posts
    122
    Thanks
    36
    Thanked 33 Times in 24 Posts
    Quote Originally Posted by diskzip View Post
    So I got the book "Valley Boy" by Tom Perkins where there's talk of Sloot.

    Anyone who doubts this accomplishment doesn't have faith in their own potential to achieve greatness.
    Let me fix it for you


    Anyone who believes this accomplishment doesn't have faith in maths.

  31. #23
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by hexagone View Post
    Let me fix it for you
    Permission denied.

    You're not qualified.

  32. #24
    Member
    Join Date
    Nov 2014
    Location
    California
    Posts
    122
    Thanks
    36
    Thanked 33 Times in 24 Posts
    Quote Originally Posted by diskzip View Post
    You're not qualified.
    You don't know me.
    Personal attacks now ? It seems to follow your trend of bad behavior in this forum though.

  33. #25
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by hexagone View Post
    You don't know me.
    Personal attacks now ? It seems to follow your trend of bad behavior in this forum though.
    I did not mean to hurt your feelings at all. Take it easy!

    You're not qualified for two reasons of fact:

    1) I do not need to have faith in Math - I *know* Math is real. You usually have faith in things you cannot know, such as `God`.

    2) In tandem with the above, you missed the whole point of the essay - why the man died so suddenly and so `unexpectedly`.

    I'm assuming you've read the article, of course. Maybe not?

  34. #26
    Member
    Join Date
    Nov 2014
    Location
    California
    Posts
    122
    Thanks
    36
    Thanked 33 Times in 24 Posts
    Yes, I read the article and no, your comments make no sense:
    <QUOTE>It seems to me the technology was entirely legitimate</QUOTE>. That is just faith.
    BTW, instead of commenting about me again, look at the math.
    I am not interested in arguing with you, believe what you want.
    But the claims make no sense.

  35. #27
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    772
    Thanks
    63
    Thanked 270 Times in 190 Posts
    Quote Originally Posted by diskzip View Post
    I have scanned in the relevant chapter herein
    Thanks! Some interesting new things, some contradicting and some different or not so accurate as other info I received.

    Most interesting is the live demo, Sloot also did that in an other demo (recording live video camera in that case).
    It's the only proof (contradicting a later Pieper interview) that playback was also working for new video's with only a small new key code at the smart card.

  36. The Following User Says Thank You to Sportman For This Useful Post:

    diskzip (14th October 2017)

  37. #28
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Quote Originally Posted by hexagone View Post
    Yes, I read the article and no, your comments make no sense:
    <QUOTE>It seems to me the technology was entirely legitimate</QUOTE>. That is just faith.
    BTW, instead of commenting about me again, look at the math.
    I am not interested in arguing with you, believe what you want.
    But the claims make no sense.
    Just because you don't understand something doesn't make it nonsense. It makes it nonsense to *you* for sure.

    For someone who doesn't want to argue, you've been doing the exact opposite so far.

    Most of it, pretty funny. Like educating me that I don't know you, while you obviously don't know me either.

  38. #29
    Administrator Shelwien's Avatar
    Join Date
    May 2008
    Location
    Kharkov, Ukraine
    Posts
    3,134
    Thanks
    179
    Thanked 921 Times in 469 Posts
    I OCR'ed the pages.

    Sounds all the more like a fraud.
    "My role was also to put in near-term cash, and, more
    importantly, to assume some responsibility for raising the huge amounts
    that would be needed to carry Fifth Force to global prominence."
    Yeah, right.

    Also, there was a plausible theory about using a huge video collection as a reference,
    but this text contradicts it.
    There apparently was no visible storage at all in the hardware, so they were
    able to play 20min of random tv content _just_ from a 75kb "smartcard".

    What's fun too, is that it talks about video and audio like they're the same thing.

    I think Sloot was just chosen to act as a magical inventor for the fraud,
    but then he got too excited and died, and thus the whole project failed,
    as it became impossible to use Sloot as scapegoat later, like
    "I didn't know that it was fake, it seemed to work".
    Attached Files Attached Files

  39. The Following User Says Thank You to Shelwien For This Useful Post:

    Mike (14th October 2017)

  40. #30
    Member
    Join Date
    May 2017
    Location
    Australia
    Posts
    126
    Thanks
    92
    Thanked 31 Times in 20 Posts
    Thanks for OCR'ing the pages, albeit with recognition errors.

    Quote Originally Posted by Shelwien View Post
    Sounds all the more like a fraud.
    "My role was also to put in near-term cash, and, more
    importantly, to assume some responsibility for raising the huge amounts
    that would be needed to carry Fifth Force to global prominence."
    Yeah, right.
    Why does that surprise you? If you've ever been involved in a large project, you know how much cash (multiple millions) are needed to get it off the ground floor.

    Case in point from current times. Multiple millions of investment are behind those "viral" ads you see on today's media platforms, while most everybody thinks it "just caught on" somehow.

    If someone's thinking a good product is going to sell itself, (s)he's being naive.

Posting Permissions

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