HomeUpSearchMail
NEW

The results of my stereo questionnaire

Introduction

• pictures with lenticular lenses and
  
• wobble pictures, which are equivalent to the above, except one does not need the move the picture; the computer display is doing that;-)

The results

• In general only results are mentioned if more than one respondent responded with a specific result (except in the web publishing environment where not many people provided feedback)..
  
• I sent the message that introduced the questionnaire to five e-mail lists with in total some 2400 members mainly interested and knowing about stereo (tech-3d, photo-3d, anabuilder (mainly French people), number30 and stereofoto (a Dutch list)), I received 96 responses back. A response result of around 4%. Questions 10 to 11 (about helping beginners) were later added so these were only answered by the last 43 respondents
  
• Some 3% of the respondents were women.
• The distribution of respondents from several countries can be seen here:
  
  
  
  The Rest countries incorporated: Australia, Belgium, Canada, China, Czech Republic, Germany, Denmark, Ireland, Iran, Italy, Mexico, Portugal, Spain and Sweden.
  
• Some 10% of the respondents were not able to do any freeviewing (being parallel or crossed viewing).
• I did not include lenticular lenses and wobble pictures as freeviewing methods, but only 1% of the people mentioned lenticular lenses. So this omission does not skew the results, I think.
  
• Of the respondents who are able to do freeviewing; 40% started with crossed viewing, which developed towards a preference for crossed viewing at some 45%.
• It is interesting to see how respondents in different countries are able to use crossed viewing in the first instance and how this develops towards his/her preference for freeviewing :
  
  
  
  

  It looks like that over time the respondents get an increased preference for crossed viewing, but on average parallel viewing is still the most preferred (for ~55% of the respondents), although France (FR) and United Kingdom (UK) seem different
  It is though interesting to see that for instance in UK a slow initial take up of crossed viewing is visible, but the preference over time increases dramatically to around 60% of the UK respondents for crossed viewing. In The Netherlands (NL) a somewhat opposite effect is seen; the preferred freeviewing of parallel increases in popularity (to ~90% of the NL respondents).
  
• Distribution of preferring crossed or parallel viewing can be seen here:
  
  
  
  

  So parallel viewing can be experienced by most people (some 54%), while crossed viewing by some 47%. The difference between the two is too small, so that for publishing guidelines of stereo pairs both are needed.
  
• The reasons why the particular preferred freeviewing method is chosen, can be seen below:
  
  
  
  

  The distribution of the reasons are in some way obvious; Parallel viewing respondents experience that method as a method with the least eye strain and it looks that respondents who prefer parallel viewing have problems in using crossed viewing. Parallel viewing indeed gives a larger image, but crossed viewing can be done on larger pictures, which will result in a higher image resolution.
  
  Some other remarks were given on the difference between the freeviewing methods: parallel viewing provides a bigger image with a larger sense of depth, while crossed viewing can provide a higher resolution image (because a larger picture size can be used).
  
• Looking at all stereo viewing methods the preference is distributed as follows:
  
  
  
  

  The stereoscope is seen as the most preferred method (is indeed also the easiest when having printed pictures), followed closely by freeviewing. Followed by the anaglyph and projected method (they need resp. anaglyph and polarized glasses).
  The Rest consisted of: horizontal and vertical interleaving, lenticular lenses, mirrors and over/under method.
   
  An interesting (future?) method that was mentioned is the possibility of an LCD viewer (slide viewer type). Furthermore anaglyphs tend to be handy for composing the stereo pair (aligning, scaling and rotating the left and right picture and for positioning the stereo window [including minimizing ghosting]).
  
• The reasons why respondents choose for a certain viewing method can be seen in the below graph:
  
  
  
  

  Comfort is the most important (I combined comfort with least eye strain). It is perhaps interesting to note that quality is not yet seen as a very important reason. Could this be a sign of the fact that stereo pair viewing is not yet mature enough (see the steps of Maslow's hierarchy of needs)?
  
  Some additional remarks made: stereoscope/viewer/projection provide a clearer/brighter picture, full immersive experience would be great, color blind people have problems with anaglyphs.

Stimulating stereo viewing

• The following stumbling blocks for beginners were mentioned:
• there is no true user friendly viewing method
  
• lack of a simple and standard viewing method
• freeviewing (parallel/crossed) is low cost, but relatively difficult to learn (due to the autonomy of the untrained brain-eye combination concerning accommodation (focussing) and convergence of the eye)
• parallel freewiewing pairs on the web have most of the times wrong width (need to be smaller or equal to eye-seperation) and don't adjust to dots per inch of the diplay. Solution is implemented on this page.
  
• a lot of published stereo pairs causes eye strain and/or headaches (due technically badly produced stereo pairs)
  
• low quality of published stereo pairs (not much depth, ghosting, wrong size, need of plugin, not properly implemented jps format, etc.)
  
• the need of additional equipment: stereoscope, anaglyph glasses, shutter glasses and/or not knowing where to get them
  
• not easy to make yourself stereo pictures (analogue or digital). No handy point&shoot stereo cameras
• different guidelines for making stereo pictures.
  
• mounting/aligning pictures is difficult; so technically good stereo pairs are sometimes rare to find and difficult to make
  
• training is not easy to acquire without being in contact with a person or a stereo association
• lack of definitive information in one location
• motivation of beginners not high enough to overcome the above mentioned issues.
• How to help beginners:
  
• need to be patient and training;  repeating, making good and wrong pictures, experience, learn, play, etc. So have fun!
  
• ignore the problems, try to work around them;-)
• join an association and/or find someone who can help.
• present only good stereo pairs to beginners.
• get them hooked with good stereo presentations (well projected slides, IMAX 3D, slide viewer, etc.)
• show several picture-taking and viewing methods to beginners and let them become comfortable with at least one.
• auto stereoscopic viewing methods are preferred
• take your time to learn freeviewing; stop when eye strain or headache creeps up. But keep on practicing
• use some tools to learn freeviewing:
  
• crossed viewing:
  use a square hole in a piece of  paper, half the width of the left/right picture and hold that half way between the stereo pair and the eyes. Try to focus (or in principle converge!) on the paper hole and while keeping the eyes converged on the hole, focus on the pictures... This needs practice, because the brain-eye combination is not used to this.
  
• parallel viewing (method 3 of this link):
  hold one side of the a piece of paper between the left and right picture and the other side at the nose. And stare at the pictures. This needs practice, because the brain-eye combination is not used to this.
  For parallel viewing make the left and right picture width not larger than the eye separation.
• start viewing good quality anaglyphs (by making minimal ghosting).
  
• get a (adjustable) stereoscope (e.g. lorgnette type) for viewing stereo pairs.
• have software recognize and generate jps files with the correct Stereoscopic Descriptor.
  
• produce a simple digital viewer (like mp3 player but then for jps files;-). Who could do that? Let me know.
  
• don't reach to high; so do a stepped approach with regard to equipment (mono camera, using slide bar or swing over, point&shoot stereo camera, quality stereo camera, etc.)
• buy some equipment to make stereo pairs (you don't have to start with high quality!). Digitally aligning (mounting) stereo pairs is not that difficult when using e.g. AnaBuilder or SPM.
  
• make beginners aware of stereo composition and stereo base and this link
  
• provide books with regard to how to build your own equipment. This can be handy for some types of beginners

Software

• The software respondents use to edit and/or view the stereo pairs is broken down as follows:



Software like Photoshop (PS), Paint Shop Pro (PSP), etc. are used most of the time (in total some 33%). Then the two more dedicated stereo programs follow: StereoPhoto Maker (SPM) and AnaBuilder.
The Rest consists mainly of: Stereoscope Applet, Nvidia, Gimp and MyAlbum.

Some experience with this software can be seen at this site.
Some stereo video applications were also used like: StereoMovie Maker, Stereoscopic Player and Pinnacle Studio.

• To publish stereo pairs on the web, the most utilized methods are: stereo pairs as jpg/gif files in HTML and StereoPhoto Viewer Applet (SPVA), followed by the use of Stereobee-, JPSViewer-, Stereoscope-Applet, plugins and scripts.
  
  Some experience with this software can be seen at this site.
  

Disclaimer and Copyright
HomeUpSearchMail