From: thale@novell.com (Todd D. Hale) Newsgroups: alt.3d Subject: SIRDS/Stereogram/Stare-E-O FAQ Date: Mon, 4 Apr 1994 23:16:54 GMT Organization: Novell, Inc., Provo, UT, USA Lines: 897 Message-ID: NNTP-Posting-Host: thale.npd.provo.novell.com The following is a smattering of info I've collected about SIRDS. It is not totally organized, nor complete, but it answers a lot of FAQs out there. Please submit any updates/changes to me at thale@novell.com. The following sections are covered herein: Stereogram Terminology Stereogram FTP Sites Commercial SIRDS Software Stereogram Publications Sample TEXT Stereograms How Single Images Stereograms work Stereogram Algorithm C Code to Generate SIRTS Hope for the Hopeless Commercial Stereogram Producers ----------------------------------------------------------------------- ----- STEREOGRAM TERMINOLOGY ------------------------------------------ ----------------------------------------------------------------------- The definitions below are listed hierarchically (a SIRDS is a SIS is a Stereogram): Stereogram Any stereo picture using stereo pairs or multiple repeating patterns Stereo Pair Two pictures from slightly different viewpoints, as used in the old handheld viewers and viewmasters SIS (Single Image Stereogram), Autostereogram Includes SIRDS, SIRTS, Single Image Rendered Data Stereogram; extends stereo pair concept to "stereo megtuples" SIRDS (Single Image Random Dot Stereogram), RDS, AutoRDS, Random Dot Anaglyphs Repeating image is made up of random dots RLS (Random Line Stereogram) Original repeating image is made up of vertical lines of varying widths; very instructional, see Slinker and Burton paper SIRTS (Single Image Random Text Stereograms), ASCII stereograms Repeating image is made up of random text characters SITS? Non random text stereograms; see Sample Text Stereograms, done manually for sigs and such a SIRD (Single Image Random Dot) not much to view in a dot; used by newbies :-) Wall-eyed viewing Viewing stereograms by focusing past the actual image Cross-eyed viewing Viewing stereograms by focusing in front of the image Focusing to infinity Forcing your eyes' lines of sight into parallel, NOT necessary for wall-eyeing SIRDS! (only occurs when the length of the repeating pattern = distance between centers of eyes) ----------------------------------------------------------------------- ----- STEREOGRAM FTP SITES -------------------------------------------- ----------------------------------------------------------------------- Submissions from michael@eccles.anu.edu.au, E.Thompson@newcastle.ac.uk, kbaum@novell.com, ...: THE FTP SITE for Stereograms is KATZ: katz.anu.edu.au () /pub/stereograms 00Index.txt - This listing 00Readme.txt - An introduction to this site ls-lR - Full list of files under this directory SIRDS.FAQ - This file, created by Todd Thale (thale@novell.com) ascii/ - ASCII stereograms blue_red/ - Stereograms requiring Blue/Red glasses een/ - Single Image Stereograms created by EEN in both TIFF and BMP formats images/ - Single Image Stereograms (some GIF and PS) ls-lR - Directory listing of all stereogram files papers/ - Papers on single image stereograms programs/ - Programs for creating and viewing stereograms (IBM, X and Unix) mac/ - Programs and Images for Apple Macintosh computers acorn/ - Programs and Images for Acorn Archimedes range of computers WWW access for MOSAIC users: http://h2.ph.man.ac.uk/gareth/sirds.html From Gareth Richards (gareth@v2.ph.man.ac.uk, richards@cernvm.cern.ch, gareth@h2.ph.man.ac.uk) OTHERS (Most of this is available at katz, above): ftp.uwp.edu (131.210.1.4) /pub/msdos/picture.viewers/rdsdraw.zip EXCELLENT SIRDS painting/generating program. CHECK IT OUT! oak.oakland.edu /pub/msdos/graphics/mindim20.zip /pub/msdos/graphics/sirdvu11.zip /pub/msdos/graphics/perspect.zip wuarchive.wustl.edu (128.252.135.4) /pub/xpgs/xpgs /pub/MSDOS_UPLOADS/graphics/dynamic.exe /graphics/graphics/mirrors/sugrfx.acs.syr.edu/3d/randot/3drandot.zoo sunsite.unc.edu /pub/academic/computer-science/virtual-reality/3d/* ftp.cs.waikato.ac.nz /pub/papers/???.tar.Z gwaihir.dd.chalmers.se pub/een/SIS/*.tif pub/een/Dirk-SIS/*.tif <-- GET THESE (not on katz, yet) ----------------------------------------------------------------------- ----- COMMERCIAL SIRDS SOFTWARE --------------------------------------- ----------------------------------------------------------------------- STW_DEMO.EXE: the full package will allow RDS creation Approx 40$ N.E.Thing Enterprises P.O. Box 1827 Cambridge, MA 02139 A company called I/O Software, Inc. (909/483-5700 800/800-7970) has a program called STEREOLUSIONS. It works on both WINDOWS and Windows NT and allows people to create, render and even print SIRDS on any Windows compatible machine. (From William Saito, 3/07/94) KAI's POWER TOOLS for the MAC has a Photoshop add-on for SIS creation. Stareo-3D ?? SEND MORE INFO! ----------------------------------------------------------------------- ----- STEREOGRAM PUBLICATIONS ----------------------------------------- ----------------------------------------------------------------------- Search library databases with these keywords: Stereogram, Autostereogram, SIRDS (Single Image Random Dot Stereogram), RDS. Considered by most as the original work on Random Dot Stereograms: 'Principles of Cyclopean Perception' by Bela Julesz, from MIT press, circa 1972. >Charles Eicher (CEicher@Halcyon.com) An excellent source of information (sample RDS and source code) is contained in Andrew A.Kinsman - Random Dot Stereograms ISBN 0-9630142-1-8 $13.95 U.S. Published by Kinsman Physics, P.O. Box 22682, Rochester, N.Y. 14692-2682 >E.Thompson@ncl.ac.uk Journal of Imaging Science and Technology, May/June 1992, Vol 36, #3. Slinker & Burton. Their use of random line stereograms really helps to understand how stereograms work, as well as some of their limitations. New Scientist (9 October 1993) has a 4 page article on SIRDS. There is a pretty cool example too. >Simon Dear (sd@mrc-lmb.cam.ac.uk) David G. Stork and Chris Rocca. "Software for generating auto-random-dot stereograms", Behavior Research Methods, Instruments, and Computers, Vol. 21, No. 5, 1989, pp. 525-534. >Steve Tate (srt@cs.unt.edu) "Stereo Computer Graphics and Other True 3D Technologies" David F. McAllister, Ed. Princeton University Press, 1993 ISBN 0-691-08741-5 US$75.00 It has several nice color plates, with stereo "triads". The triads consist of a left, a right, then another left image. Use the left pair for viewing walleyed, or the right pair for viewing crosseyed. >Mike Weiblen (mew@digex.net) "Stereogram" (c) 1994, Cadence Books, San Francisco, CA "a newly edited version of CG STEREOGRAM and CG STEREOGRAM 2, published by Shogakukan Inc. In Tokyo, Japan" ISBN 0-929279-85-9 Cadence Books, A division of Viz Communications, Inc. P.O. Box 77010 I liked it. Much better in my opinion than the other Stereogram book I've seen ("Magic Eye"). This one includes much textual information, including the origins of stereograms, how to see them, precursors such as stereo pairs, and an article by Christopher W. Tyler, who invented the SIRDS. Best of all were the stereograms themselves. There are roughly 50 SIRDS, most of them full page (the book is softcover and about 8" square). The ones I've managed to see so far have been quite good, and also included are some of the very first ones. In the history department there are stereo pairs, stereo photographs, and even some stereo-pair paintings by Salvador Dali. All of the stereograms indicate whether they require wall-eyed or cross-eyed viewing (or either). Most are true 3d designs, not the "cutout" variety. The book is 93 pages and most of the plates are full-color. Well worth the money in my opinion. >Michael Moncur (mgm@xmission.com) _Magic Eye: A New Way of Looking at the World_ copyright (c) 1993 by N.E. Thing Enterprises. Andrews and McMeel, A Universal Press Syndicate Company, Kansas City ISBN: 0-8362-7006-1 First Printing, September 1993 ... Fifth Printing, January 1994 Introduction contains a history of the technique and phenomena. Viewing Techniques are explained. 25 pages of full-color STARE-E-O images. (Plus images inside the front and back covers.) "Answers" included. 32 pages, hardcover, 8.75x11.5 inches, horizontal format, with slipcover. US$12.95 ($16.95/Canada) I just bought this for US$7.77 at Media Play in Utah. It's currently on their best seller list. Well worth any price under $30! >Todd Hale, 3/25/94 Magic Eye II: Three Dimension Trip Vision Copyright 1992 by N.E. Thing Enterprises/Tenyo Co., Ltd. Korean Translation Copyright 1993 by Chungrim Publishing Co. All the text is in Korean, so I can't read it. But it has some pretty cool pictures. They are all SIRxS where x is various patterns/pictures. I paid US$20 for it. Interestingly, this title doesn't seem to be mentioned in my N.E. Thing catalog. >Mark Hudson (M_Hudson@delphi.com) SEND MORE REFERENCES! ----------------------------------------------------------------------- ----- SAMPLE TEXT Stereograms (Not random) ---------------------------- ----------------------------------------------------------------------- From: dthomas@bbx.basis.com (Dave Thomas) O O n n n n n n n n n n n n n n n n f f f f f f f f f f f f f e e e e e e e e e e e e e e e e a a a a a a a a a a a a a a a a a a a a a a a a a a a a a r r r r r r r r r r r r r r r r r r r r r r r r r r r r r g g g g g g g g g g g g g g g g g g g g g g g g g r r r r r r r r r r r r r r r r r r r e e e e e e e e e e e e e e e a a a a a a a a a a a a a t t t t t t t t t t t <<<<>>>><<<<>>>><<<<>>>><<<<>>>><<<<>>>><<<<>>>><<<<>>>><<<<>>>><<<<>>>> d d d d d d d d d d d e e e e e e e e e e e e e p p p p p p p p p p p p p p p t t t t t t t t t t t t t t t t t t t h h h h h h h h h h h h h h h h h h h h h h h h h From: jittlov@gumby.cs.caltech.edu (Mike Jittlov) . . . . . . . . @ . @ . @ . @ . > /^\ /^\ /^\ /^\ /^\ > #################################################################### > #################################################################### > / #### / ####\ / #### \ / #### \ /#### \ >/ ## \__/ ## \__/ ## \__/ ## \__/ ## \ > ____ ## ____ ## ____ ## ____ ## ____ ## ____ > / \## / \ ## / \ ## / \ ## / \ ##/ \ > | 2D |# | 2D |## | 2D | ## | 2D | ##| 2D | #| 2D | > | or |# | or |## | or | ## | or | ##| or | #| or | > | 3D |# | 3D |## | 3D | ## | 3D | ##| 3D | #| 3D | > | ?? |# | ?? |## | ?? | ## | ?? | ##| ?? | #| ?? | > | | | | | | | | | | | | > -------- -------- -------- -------- -------- -------- > \\\\\\\\ \\\\\\\\ \\\\\\\\ \\\\\\\\ \\\\\\\\ \\\\\\\\ > \\\\\\\\ \\\\\\\\ \\\n\r\\ \\f\o\d\ \\\\\\\\ \\\\\\\\ > \\\\\\ \\\\\\ \\\\\\ \\\\\\ \\\\\\ \\\\\\ > DR J _____________________________________________________________________________ [] [] [] | [] [] [] [] | [] [] [] [] | [] [] [] [] | [] [] [] [] | [] [] [] ============================================================================= * : * : * : * : * : * .' `: .' `. : .' `. : .' `. : .' `. :' `. Mike Jittlov Mike.Jittlov Mike Jittlov.Mike Jittlov Mike Jittlov Mike Jittlov The Wizard` The Wizard. The Wizard . The Wizard .The Wizard 'The Wizard * * of . * * of * * of * . * of * * of* * . o* * |\^/|peed . |\^/|Speed |\^/| Speed |\^/| Speed|\^/| Spee|\^/| . Spe|\^/| \V/ and . \V/ and .\V/ and \V/ and \V/. and \V/ . and\V/ /_\Time! . /_\ Time! ./_\ Time! /_\ Time! /_\. Time!/_\ . Time/_\ _/ \_ . _/ \_ _/ \_ _/.\_ _/ \_ _/ \_ . _/ \_ ============================================================================= WWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWW ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ One more from ... well ... YOU figure it out! _-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~ ! Adam Adam Adam Adam Adam Adam ! "Lawn ! !Spragg Spragg Spragg Spragg Spragg Spragg Spragg! service!" ! ! Adam Adam Adam Adam Adam Adam ! -Stup ! !Spragg Spragg Spragg Spragg Spragg Spragg Spragg! ! ! Adam Adam Adam Adam Adam Adam ! atspragg@ ! !Spragg Spragg Spragg Spragg Spragg Spragg Spragg! ucdavis.edu ! ! Adam Adam Adam Adam Adam Adam ! GO AGS! ! _-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~-_-~ ----------------------------------------------------------------------- ----- How Single Image Stereograms Work ------------------------------- ----------------------------------------------------------------------- Let's start with a simple repeating pattern. First we'll examine viewing stereograms wall-eyed (focusing beyond the picture). Adam Adam Adam <-- PERCEIVED IMAGE (3D) \ \ \ \ \ Adam Adam Adam Adam <-- ACTUAL PICTURE (Flat, 2D) \ \ \ \ <-- LINES OF SIGHT FROM EACH EYE \ \ (*) (*) <-- EYES As shown in the above graphic, the lines of sight go through different (but identical) words. The eyes must be directed to just the right level behind the picture to get the two "Adams" to line up. When they do line up, a single "Adam" appears at the level shown behind the picture. This is why it is easiest for some people to look at stereograms behind glass. You simply look at your reflection, which appears to be behind the glass at about the same level as the 3d stereogram image. More on this later. So, why is the second "Adam" in the above image "deeper" than the other two? Since the next two "Adams" that line up are FARTHER apart than the first two, your eyes must diverge a little more to get them to line up. As the eyes diverge, their lines of sight meet a little further away. This causes the next "Adam" to appear deeper in the image. Adam Adam | Adam <-- PERCEIVED IMAGE (3D) / \ / \ / \ Adam Adam Adam Adam <-- ACTUAL PICTURE (Flat, 2D) / \ / \ <-- LINES OF SIGHT FROM EACH EYE / \ (*) (*) <-- EYES The above graphic was crammed into 10 lines of text. If we moved the eyes back (to a more realistic distance), the whole 3d image would appear deeper, and not so spread out. Cross-eyed viewing follows the same principles, but brings the focal point up front of the picture and inverts the image. I'll use the same eyes and picture, but move the eyes back for room to illustrate. Adam Adam Adam Adam <-- ACTUAL PICTURE (Flat, 2D) \ | \| Adam <-- PERCEIVED IMAGE (3D) | \ | \ | \ <-- LINES OF SIGHT FROM EACH EYE | \ (*) (*) <-- EYES Now for the inverse part. Since the next two "Adams" are farther apart, you must cross your eyes MORE (converge instead of diverge your eyes) to get the "Adams" to overlap. This causes the next "Adam" in the image to appear closer than the first one. Adam Adam Adam Adam <-- ACTUAL PICTURE (Flat, 2D) \ / \ / Adam\ / Adam <-- PERCEIVED IMAGE (3D) Adam / \ / \ <-- LINES OF SIGHT FROM EACH EYE / \ (*) (*) <-- EYES Whew. Ascii graphics are a bit too tight to be very technical. By the way, each letter (or pixel) in the 3D image comes from two letters in the picture. Stereograms have blurry edges (right and left edges) because the line of sight from the one eye passes through the picture, but the line of sight from the other eye is off the picture. Examine the wall-eyed example below. Adam Adam Adam <-- PERCEIVED IMAGE (3D) \ \ \ \ \ \ Adam Adam Adam Adam <-- ACTUAL PICTURE (Flat, 2D) \ \ \ \ <-- LINES OF SIGHT \ \ (*) (*) <-- EYES Here, the line of sight for the right eye goes through first "Adam" in the picture, and the line of sight for the left eye is off the picture. (Everything to the left of the first "Adam" in the IMAGE is blurry.) HOW TO GENERATE STEREOGRAMS Random stereograms work the same way; they simple repeat random text or pixels, instead of words like the "Adam" used above. Below, the pattern, "ASDFGHJK", is repeated four times in each line. (I duplicated it for easier viewing.) ASDFGHJKASDFGHJKASDFGHJKASDFGHJK ASDFGHJKASDFGHJKASDFGHJKASDFGHJK ASDFGHJKASDFGHJKASDFGHJKASDFGHJK As you might guess, the image consists of THREE repetitions of the pattern, set back behind the picture (wall-eyed). To add some varying depth to the image, we can simple add a space in the middle; or better, we'll add a different letter, the letter Q. I've also dropped the K off the end to keep it suitable for framing. :-) ASDFGHJKASDFGHJKASDFGHJKASDFGHJK ASDFGHJKASDFGHJKASQDFGHJKASDFGHJ ASDFGHJKASDFGHJKASDFGHJKASDFGHJK Adding the Q has the same effect as adding the space between "Adams" in the previous example. Here it is again for comparison. Adam Adam Adam Adam Adam Adam Adam Adam Adam Adam Adam Ada Adam Adam Adam Adam Adam Adam These two stereograms demonstrate going down a level (deeper, for wall- eyed viewers) by adding a new letter where you want to go down. The image returns back to the original level because the added letter was not repeated down the line; it was essentially deleted again. Hence, deleting a letter causes the image to come forward one level. Lets look at this by deleting a letter before adding any. ASDFGHJKASDFGHJKASDFGHJKASDFGHJK ASDFGHJKASDFGJKASDFGHJKASDFGHJKA ASDFGHJKASDFGHJKASDFGHJKASDFGHJK I deleted an H from the middle row, and added an A on the end to make it flush again. This causes the image to come up a level where the H was deleted, then go back down where the H returns to the pattern. To fully control the ups and downs of stereogram generation, you simply keep track of what's repeating and add and delete letters when needed. For example, start with the pattern ASDFGHJK. If the input for the 3D stereogram image is 00000000000011111222221111111110000, where the numbers represent the levels of the 3D image, you simply delete a letter from the pattern when you go up a level (0 to 1, or 1 to 2), and add a letter to the pattern when you go down a level (2 to 1, or 1 to 0). If you want to go up multiple levels, e.g., from level 1 to 3, delete 2 letters from your pattern. Adding two letters to the pattern likewise causes a two-level drop in the image. Start each line of the image by dumping the whole pattern. Recall that in the examples above, four "Adams" in the picture creates three "Adams" in the image. By dumping the pattern to start, we are simple providing the first "Adam" to build the 3D image from. Now, we start building the 3D image. ASDFGHJK Since the first 12 levels are at level 0, we do no adding nor deleting; just dump 12 more letters from the pattern in order. ASDFGHJKASDFGHJKASDF ^image starts here We now have our first transition, from level 0 to 1. Which letter do we delete? Delete the next one in the pattern. ASDFGHJK ^the next letter in the pattern Since G is next, we delete it. Our new pattern is ASDFHJK, and H is the next letter to be used in the image. ASDFHJK ^the next letter to use in the image Forget about the G. Consider it gone forever. The input now indicates five letters at level 1. Since we made the transition to level one by deleting the G, we simply copy the next 5 letters from the pattern into the image. The next 5 letters in the pattern are HJKAS. (You've noticed by now that we wrap around when we get to the end of the pattern.) I'll lay the input next to our image to make it easier to see where we're at. 00000000000011111222221111111110000 ASDFGHJKASDFGHJKASDFHJKAS { }{ }{ } initial level 0 level pattern 1 To jump up to level 2 from level 1, we delete the next letter in the pattern, the letter D. The new pattern is shown below. ASFHJK ^the next letter to use We have 5 letters at level 2, so we copy the next 5 letters from the pattern, FHJKA, to the image. 00000000000011111222221111111110000 ASDFGHJKASDFGHJKASDFHJKASFHJKA { }{ }{ }{ } initial level 0 lev level pattern 1 2 We must now move down a level, from 2 to 1, by ADDING a letter to our pattern. Where do we add it? Add it as the next letter to use. Let's add an X (randomly chosen letters work better than previously used letters). Since A was the last letter used, S is the next to be used. ASFHJK ^add X here AXSFHJK ^next letter to use Since there are nine letters at level 1, we must copy nine letters from the pattern to the image (XSFHJKAXS). 00000000000011111222221111111110000 ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXS { }{ }{ }{ }{ } initial level 0 lev level level 1 pattern 1 2 Drop another level, from 1 to 0, by adding another letter to the pattern at the current position. AXSFHJK ^add random letter (M) here AXSMFHJK ^next letter to use Since there are four letters at level 0, we copy four letters from the pattern to the image (MFHJ). 00000000000011111222221111111110000 ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ { }{ }{ }{ }{ }{ } initial level 0 lev level level lev pattern 1 2 1 0 WE'RE DONE. Kinda anti-climatic, huh? Let's duplicate the image line to make it easier to view. ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ ASDFGHJKASDFGHJKASDFHJKASFHJKAXSFHJKAXSMFHJ For each new line in the image, start with a new random pattern, and follow the steps above. So, how are graphics stereograms made? The same way, using colored pixel values for input. 3drandot uses RED as the color deepest in the image, with GREEN in the middle, and BLUE as the highest color. The pseudocode below follows this algorithm. If you have any questions, post them in the newsgroup. Let me know if you are interested in C code that follows this algorithm. (The C code here in the FAQ follows a different algorithm, the "lookback" algorithm.) Enjoy!! thale@novell.com ----------------------------------------------------------------------- ----- STEREOGRAM ALGORITHM -------------------------------------------- ----------------------------------------------------------------------- The following is an algorithm I developed for SIRDS generation. It is functionally equivalent to the look back algorithms, without looking back. It can be used for graphic stereograms, as well as the text stereograms. It simply stores in PATTERN the values we would look back at. The input file is depth data, or a z-buffer (see example input below). The levels available in a picture are the maximum pattern size divided by two minus one. Maximum pattern size should be equivalent to half the distance between the eyes (for easy viewing), which is about 1 1/4 inches. So, for typical text stereograms on an 80 char display, that's about 10 characters for the maximum pattern size and 4 or 5 levels available. However, since 4 or 5 levels is unreasonable, we push the max pattern size to 16 (still less that the 20 character distance between the eyes). This gives us about 7 or 8 levels. If we do allow more than 7 levels when the max pattern size is 16, the repeating pattern (of size max pattern - current level) grows so small that four of them in a row (e.g., of size 7) would also create two double patterns in a row (of size 14), causing floaties (aliasing). I hope I'm making some sense. I'll try to spend some time rewriting this. In addition, I'll integrate answers to any questions that come as a result of this mess! PL : Previous Level value NL : Next Level value INITIAL LEVEL SETTING PL = 0 OPEN INPUT FILE FOR READING WHILE NOT EOF (Y DIMENSION: FOR EACH LINE) GENERATE NEW PATTERN SET PATTERN PTR TO BEGINNING OF PATTERN OUTPUT FULL PATTERN WHILE NOT EOL (X DIMENSION: FOR EACH POSITION) NL = NEXT INPUT LEVEL VALUE IF NL != PL (IF WE'RE CHANGING LEVELS) IF NL > PL (IF WE'RE MOVING UP) DELETE NEXT NL-PL BITS IN PATTERN FROM PTR FORWARD (IF GOING FROM LEVEL 0 TO 2, DELETE 2 BITS IN PATTERN) ELSE (IF MOVING DOWN, AWAY) INSERT PL-NL RANDOM BITS INTO PATTERN AT PTR POSITION (IF GOING FROM LEVEL 3 TO 2, INSERT 1 BIT INTO PATTERN) PL = NL (UPDATE PL) OUTPUT NEXT VALUE IN PATTERN ELSE OUTPUT NEXT VALUE IN PATTERN ADVANCE PATTERN PTR (CIRCULAR PATTERN) ADVANCE INPUT PTR (Redundancy exists to simplify readability.) Assuming spaces translate to level 0, input like this (thanks, Dave Thomas): 22222222222222222 11111111111111111111111 2222222222222222222 11111111111111111111111 22222222222222222222 11111111111111111111111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222222222222222222 11111 2222222222222222222 11111 22222222222222222 11111 Yields output like this: Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D mwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`, 7Z,o:.sw5[rHt&{:7Z,osw5[rHt&{:7Z,osw505[rHt&{:7Zosw505[rHt&{:7Zosw505[rR g"EzQy{lW_3C&B[8g"Ez{lW_3C&B[8g"Ez{lW_3S3C&B[8g"z{lW_3S3C&B[8g"z{lW_3S3y yiQ?fiVF!G{TAtrJyiQ?VF!G{TAtrJyiQ?VF!G{T]TAtrJyi?VF!G{T]TAtrJyi?VF!G{T]s 1`I|r?FLoB#Y8igID(]?"A0+;<>Y8ID(]?"A0+;<>Y8ID(]?"?"A0+;<>Y8ID(]?"?A0+;Y8ID(]?" eRl5hfJx.cD0tscheRl5Jx.cD0tscheRl5Jx.cDeD0tscheRl5Jx.cDeDtscheiRl5Jx.cDe v,r7f6%<)R-AaV^hv,r7%<)R-AaV^hv,r7%<)p)R-AaV^hv,r7%<)p)R-aV^hv>,r7%<)p)R VbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]q `^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v= This uses 16 as the Max pattern size. Notice on the third line of the output (the first line with level changes), the pattern is "7Z ... &{:". When we go from level 0 to level 2, we lose two characters (i.e., ":."). When we go from level 2 back to level 0, we insert two new characters (i.e., "05" after the "osw5"). Any questions? Or better yet, what did I leave out?? I do hope this helps someone!?! ----------------------------------------------------------------------- ----- C CODE TO GENERATE TEXT STEREOGRAMS ----------------------------- ----------------------------------------------------------------------- Also from Dave Thomas: Now here's a hack of a hack I found on the net. Basically it generates stereograms from ascii diagrams. Here's the program first (it uses the lookback algorithm, not the stored-pattern algorithm described above): ##define DEPTH0 16 ##define COLS 79 ##include main() { long now; int col; char m[COLS+1],s[COLS+1]; int i,e,c; /* seed random number generator */ time (&now); srand (now); /* print fusion X's */ for (col = 0; COLS - col >= DEPTH0; col += DEPTH0) { for (i = 0; i < DEPTH0-1; ++i) putchar (' '); putchar ('X'); } putchar ('\n'); /* process input image */ while (memset(m,'\0',COLS+1),fgets (m, COLS+1, stdin)) { e = 0; s[COLS] = 0; for(i=0; i= DEPTH0 && i < COLS && (c = m[i - DEPTH0]) >= '0' && c <= '9') { e = 1; s[i] = s[i-DEPTH0+c-'0']; ++i; } s[i++] = (e || i < DEPTH0) ? randasc() : s[i-DEPTH0]; } puts (s); } } randasc() { /* large set of random characters */ return ('!'+rand()%92); /* lowercase random letters */ /* return ('a'+rand()%26); */ } Try passing it this file as the standard input: 22222222222222222 11111111111111111111111 2222222222222222222 11111111111111111111111 22222222222222222222 11111111111111111111111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222 22222 11111 22222222222222222222 11111 2222222222222222222 11111 22222222222222222 11111 You'll get output like this: X X X X "i`<$$Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D(v"i`<$$Co6&`R(D dZ<$wamwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`,QFdZ<$wamwBy&>`, rHt&{:7Z,o:.sw5[rHt&{:7Z,osw5[rHt&{:7Z,osw505[rHt&{:7Zosw505[rHt&{:7Zosw505[rR 3C&B[8g"EzQy{lW_3C&B[8g"Ez{lW_3C&B[8g"Ez{lW_3S3C&B[8g"z{lW_3S3C&B[8g"z{lW_3S3y {TAtrJyiQ?fiVF!G{TAtrJyiQ?VF!G{TAtrJyiQ?VF!G{T]TAtrJyi?VF!G{T]TAtrJyi?VF!G{T]s ##Y8igID(]?"A0+;<>Y8ID(]?"A0+;<>Y8ID(]?"?"A0+;<>Y8ID(]?"?A0+;Y8ID(]?" D0tscheRl5hfJx.cD0tscheRl5Jx.cD0tscheRl5Jx.cDeD0tscheRl5Jx.cDeDtscheiRl5Jx.cDe -AaV^hv,r7f6%<)R-AaV^hv,r7%<)R-AaV^hv,r7%<)p)R-AaV^hv,r7%<)p)R-aV^hv>,r7%<)p)R IRW)ytVbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]qi(IRW)ytVbWqd^]q BX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v=[gBX(Aqq`^Y-z8v= ----------------------------------------------------------------------- ----- HOPE FOR THE HOPELESS :-) --------------------------------------- ----------------------------------------------------------------------- From: wchaga@vela.acs.oakland.edu (William C. Haga) Being one who has used wide-eyed vision on chain link fences ever since I was a kid, I was able to see the images in SIRDS right away. But I've had difficulty explaining the technique to friends. Today I had the latest Games magazine with me at my parents house. Games is running another contest using SIRDS, so there are three in the latest issue. This time I thought of the reflection idea. So I opened mom's china cabinet, put the magazine against the glass in the door, and told mom to keep looking at her own reflection in the glass until the image appeared. It took less than thirty seconds. When she saw the 3d train engines, I was subjected to a squeal of delight that I hadn't heard from her for a long time. "EEK! IT'S COMING OUT AT ME! IS THIS EVER NEAT!". Dad tried for about a minute but gave up. About an hour later, mom and I heard a shout. We went to the dining area, and there was dad with the magazine against the glass in the door. "Isn't that just the most amazing thing!", said he. Later they were making jokes about teaching old dogs new tricks. ----------------------------------------------------------------------- ----- COMMERCIAL Stereogram Producers -------------------------------------- ----------------------------------------------------------------------- Infix Technologies PO Box 381 Orem, UT 84057-0381 (801) 221-9233 contact: John M. Olsen, jolsen@nyx.cs.du.edu 18x24 Earth (mercator projection of the Earth's altitudes) 18x24 Salt Lake LDS Temple Centennial 18x24 Beethoven (300 DPI! Very smooth.) (others are coming soon, too.) Retail price for the 18x24 inch prints is $20 plus $3 s/h. Utah residents add 6.25% sales tax. Wholesale and distributor discounts are available. Quotes for custom work are also available. Cost and minimum order varies, based on content. Altered States 92 Turnmill St, Farringdon, London, EC1, U.K. +44 (0)71 490 2342 Paul Dale (P.A.Dale@bath.ac.uk) tel: +44 (0)225 826 215 Send a catolog request to: N.E. Thing Enterprises 19C Crosby Drive Bedford, MA 01730 >Info from: Neal T. Leverenz (at802@yfn.ysu.edu) Inner Dynamics, Inc. The distributor: Privileged Traveler 4914 Brook Road Lancaster, OH 43130 (614) 756-7406 Retail prices stated below plus $3 S&H. Ohio residents add 5.5% sales tax. GLOW IN The DARK Poster Series - $23 (size: 18" X 24") "Knight Vision" - suspended chess board with chess pieces above the board in daylight viewing - also an area in the center that has a Knight chess piece - random dot pattern glows and is viewable in the dark!! Premium Color Series - $16 (size: 18" X 24") "Gecko" - twin Gecko lizards "I Think Therefore I Am" - well known quote surrounded by stunning visuals "SoulMate" - hearts, spirals, and other symbols, for that special person "The Mighty Unicorn" - unicorn, mystical castle, wizard, and flying dragon "Excalibur" - legendary sword in the stone, castle, knights, etc. "Where's Wilbur?" - can you find him in the forrest? Optimum Series - $15 (size: 24" X 36", black and white) "Beyond Reality" - hearts, spirals, other cool shapes - extremely detailed "20/20 Third Sight" - an eye chart done in 3D "Illusions" - a labyrinth, try to find your way out! "Meditation" - contains an ancient mandala, a real stress buster "DreamWeaver" - unusual geometric shapes, helps induce lucid dreaming and dream remembrance "Icons" - the five symbols of life "Rainbows" - see color on a black and white poster (Not a 3D poster) >Sam Weaver (sew@cblpd.att.com) * Todd D. Hale * * * * thale@novell.com * * halet@bert.cs.byu.edu