Stereoscopic Displays and Virtual Reality Systems IX (2002)
Proceedings of the SPIE Volume 4660
Stereoscopic Standards Forum Summary
Tuesday, 22nd January 2002
The first Stereoscopic Standards Forum was held in 1998 with an aim to develop useful standards for the stereoscopic imaging field and aid the further development of stereoscopic imaging technologies. This year was the third Stereoscopic Standards Forum - there was a wide range of topics to cover and there was good input from the floor.
The first item of business was that the previous chairman of the standards committee, Mike Weissman, has unfortunately had to withdraw from the committee due to other work commitments. We take this opportunity to thank Mike for his previous work on the committee. Andrew Woods chaired this forum, with support from panel members John Rupkalvis (Stereoscope International, Burbank, CA) and Vivian Walworth (Rowland Institute, Cambridge, MA) who are also co-chairing the Stereoscopic Lexicon subcommittee.
Activity towards the Field-sequential stereoscopic video standard was summarized by Andrew Woods. Discussions have taken place with the chair of SMPTE's "I23" Television Image Technology standards committee over the year. It has been suggested that the "Field-sequential" standard should not be PAL/NTSC/SECAM specific (since these are analog legacy video standards) but generalized to include all interlaced video standards, which would then include the new digital broadcast video standards. Work on this topic is continuing.
Activity of the Stereoscopic Lexicon sub-committee was summarized by John Rupkalvis with input from Vivian Walworth. A mailing list has been created on the Yahoo Groups website <http://www.yahoogroups.com> and a basic protocol has been established for progressing lexicon definitions. Four background documents were identified as a reference point for the definition of words in the SD&A lexicon. They are:
(a) The glossary from "Three-dimensional Photography" by Herbert McKay (1953),
(b) "A glossary of terms used in stereoscopy" published in the International Stereoscopic Union (ISU) journal "Stereoscopy" in September 2000 and edited by Don Wratten,
(c) The Stereographics Developersí Handbook (1997), and
(d) The Manual of Photogrammetry (Fourth edition. Chester C. Slama, Charles Theurer, and Soren W. Henriksen, editors, Falls Church, Virginia: American Society for Photogrammetry and Remote Sensing, 1980).
John reiterated the need for a public lexicon - there is a wide range of terms, which are specific to the stereoscopic imaging field - but there are many terms that are either undefined or ill defined. An example was given of the term "disparity" - there are several definitions in common use. The intention is not necessarily to recommend a preferred usage, but to advise practitioners that there are several definitions in common usage and to reduce ambiguity in future publications.
At last year's stereoscopic standards forum, Dick Kessler (currently acting chairman of the Society for Information Display (SID) standards committee) raised the important issue that groups defining industry wide standards need to be careful of antitrust legislation. The issue is "if a select few sit down and agree upon a standard, the odds are the standard will hurt someone who is likely to be making a similar product which does not meet their standard. Nonetheless, standards are necessary." This text comes from an article published in SMPTE Journal in July 1991 (and republished in October 2000: "The Commitment of SMPTE to Standardization" by Alex E. Alden). The article goes onto quote the US Attorney General in 1922 as saying "I can see nothing illegal in the exercise of the activities mentioned, provided always that whatever is done is not used as a scheme or device to curtail production or enhance prices, and does not have the effect of suppressing competition." Standards organizations ordinarily use one of two mechanisms when setting standards to meet these requirements: the committee method or the canvass method. SMPTE use the committee method. The easiest way for us to ensure that the correct processes are followed, is to work with an existing standards setting organizations - of which there are many.
The idea of summarizing defacto standards on the website was suggested by the chair. This would have the advantage that it would provide a reference for the current status quo, and would allow some documentation to be made available where a full standards process was unlikely to be needed or processed. Three examples were given:
(a) The orientation of the polarizers in linear polarized 3D glasses. The general orientation used throughout the industry is the "V" orientation - where one eye is +45 degrees and the other eye is -45 degrees - although there are some other usages, e.g. some IMAX 3D theatres.
(b) The orientation of the polarizers in circular polarized 3D glasses. Most organizations use the same orientation of the circular polarization, however there is some variation regarding the orientation of the linear side of the circular polarizing filters.
(c) Connectors for Liquid Crystal Shutter (LCS) Glasses. Most LCS glasses use a 3.5mm stereo audio connector wired in a common way, however, again there is some variation.
In the case of (a) and (b) we will aim to describe the current defacto standards and find other documents which refer to a preferred orientation. In the case of (c), an industry player is currently conducting a survey of available hardware and we will seek to obtain a copy of that survey.
Some ideas that came from the floor during the forum:
Two mailing lists have been set up for the SD&A standards activities, they are:
Please consider subscribing and participating.
- Yosh Martinband (3ality, Israel) suggested that it would be useful to develop some guidelines for producing stereoscopic content.
- John Merritt jokingly suggested that there should be a standard, which states that the left perspective image should always be directed to the left eye and vice versa for the right eye.
- John Merritt also recommended that there should be a standard, which defines the maximum time sync offset, allowed between left and right streams. Lenny Lipton indicated that the SMPTE had done some work in this field and their recommendation was that there should be +/- 1/4 frame offset between the left and right streams. The reference for this work is: Jones, R. Clark; and Shurcliff, William A. 1954. "Equipment to measure and control synchronization errors in 3-D projection" in Journal of the SMPTE, 62:134-41.
- Lenny Lipton recommended that there should be a standard high-resolution
clock in PCs.
- Various suggestions came from the floor regarding stereoscopic HDTV over a
single channel and stereoscopic digital cinema.
- Theodore Goodman (University of California, Santa Cruz) advised that Motion JPEG 2000 was currently being defined and that it was an opportune time to define stereoscopic meta-data and stereoscopic region of interest for this
In closing the session, the audience was thanked for their participation and invited to participate in one of the standards sub-committees.