jellinghaus-robert@CS.YALE.EDU (Rob Jellinghaus) (08/13/90)
The "office of the future" concept in virtual reality has been widely
discussed--put on your goggles and bang you're in the office. But one
problem is going to have to be solved, and solved well, before the
virtual office exists: the technology will have to support virtual
documents that you can _read_.
Consider the problems in rendering a screen, or a page, of text. Not
only must the "eyephone" screens possess high enough resolution to let
people read virtual documents comfortably, for hours, but one must be
able to pick up virtual screens, move around them, etc., which means
text rendering must be very fast and very accurate.
Has anyone done any work on smooth, realistic rotation of text? Almost
all the 3D graphics stuff I've ever seen specializes in blitting
lots of polygons with good shading effects, which doesn't seem very
applicable to the special problems text presents.
Also, has anyone noticed the parallels between the discussions we've
been having here about virtual space and navigation therein, and the
work that's been done on hypertext information spaces? In both contexts
there is a lot of stuff in the world, and you need to be able to know
where you are and where you want to be. Maybe the two fields will inter-
breed at some proximate date.
--
Rob Jellinghaus | "Next time you see a lie being spread or a
jellinghaus-robert@CS.Yale.EDU | bad decision being made out of sheer ignor-
ROBERTJ@{yalecs,yalevm}.BITNET | ance, pause, and think of hypertext."
{everyone}!decvax!yale!robertj | -- K. Eric Drexler, _Engines of Creation_auric1@milton.u.washington.edu (Alan Stearns) (08/16/90)
In article <25797@cs.yale.edu> jellinghaus-robert@CS.YALE.EDU (Rob Jellinghaus) writes: >The "office of the future" concept in virtual reality has been widely >discussed--put on your goggles and bang you're in the office. But one >problem is going to have to be solved, and solved well, before the >virtual office exists: the technology will have to support virtual >documents that you can _read_. Don't worry too much about reading text in VR just yet. We haven't got to the point where the 2D representation of text I'm typing now is suitable for reading hours upon end. The virtual office is at least as far away as the paperless office, which is remotely feasible with today's technology but involves too many compromises to make it a widespread choice. If it isn't easier to use a computer for a given task, then don't. Text display will have to get as good as paper before we should make the switch. But even if we don't use a computer or VR for everything, they still have their place. How about, for the virtual office, we have partial-immersion VR. The VR display is projected on a pair of see-through goggles so we can see the real world around us, with a simperimposition of VR objects. The real world could be cut off by hooding the goggles, or the VR could be turned off so we could see only the real world. In a meeting with another office person, thousands of miles away, I see him superimposed in a chair in my office. I can read real-world documents on my desk, look at virtual objects he shows me, or read my computer screen (that he may be sending data to). We can also put on our blinders to shut out the real world, perhaps to traverse a data space or construct a presentation.
xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) (08/16/90)
jellinghaus-robert@CS.YALE.EDU (Rob Jellinghaus) writes: > >The "office of the future" concept in virtual reality has been widely >discussed--put on your goggles and bang you're in the office. But one >problem is going to have to be solved, and solved well, before the >virtual office exists: the technology will have to support virtual >documents that you can _read_. This brought to mind another human ability, nearly magic, that will be of _crucial_ experience for navigating virtual reality, and that seems to have been neglected as yet. This is the "cocktail party" phenomenon: in a crowd of fifty conversations, our ears can somehow pick out the one of interest to us, and focus on it to the exclusion of all others. Two items of research interest 1) how do we do it; i.e., what are the parts of the signal crucial to make this work, so that our virtual reality generator can be designed to present them to us; and 2) how do we indicate our new focus of attention to our virtual reality interface, so that we may use it for navigation, when at present it is all done "in our heads". The application is obvious: one can navigate a database along "fifty" simultaneous threads until the one of interest is isolated, then branch "fifty" times again to narrow the focus. This should be much faster than a visual interface to textual/verbal information, on a guess, if it can be replicated from the real world into the virtual reality world. Besides, isn't that how it's done in all the sexy cyberpunk stories? ;-) Kent, the man from xanth. <xanthian@Zorch.SF-Bay.ORG> <xanthian@well.sf.ca.us>
brucec%phoebus.phoebus.labs.tek.com@RELAY.CS.NET (Bruce Cohen;;50-662;LP=A;) (08/17/90)
In article <25797@cs.yale.edu> jellinghaus-robert@CS.YALE.EDU (Rob Jellinghaus) writes: > ... > > Has anyone done any work on smooth, realistic rotation of text? Almost > all the 3D graphics stuff I've ever seen specializes in blitting > lots of polygons with good shading effects, which doesn't seem very > applicable to the special problems text presents. > Yes, a lot of work has been done. This is essentially an anti-aliasing problem, with image-processing sorts of solutions. See, for instance: @Article{, author = "Crow, Frank C.", title = "The Use of Grayscale for Improved Raster Display of Vectors and Characters", journal = "Siggraph Proceedings", year = "1978", } % The copy of the article I have doesn't have any publishing information. @Article{, author = "Wieman, Carl F. R.", title = "Continuous Anti-Aliased Rotation and Zoom of Raster Images", journal = "Siggraph Proceedings", year = "1980", } @Article{, author = "Warnock, John E.", title = "THe Display of Characters Using Gray Level Sample Arrays", journal = "Xerox Technical Report CSL-80-6", year = "1980", } There are some more recent papers, but I haven't gotten them out of the boxes since the last move. I seem to remember one by Maureen Stone of Xerox PARC three or four years ago. > Also, has anyone noticed the parallels between the discussions we've > been having here about virtual space and navigation therein, and the > work that's been done on hypertext information spaces? In both contexts > there is a lot of stuff in the world, and you need to be able to know > where you are and where you want to be. Maybe the two fields will inter- > breed at some proximate date. > Yes, I noticed it because I recently spent a year building a hypertext system for software documentation, adn I had to spend quite a bit of time both researching the literature and thinking about navigation. The two problems are really one; IMHO hypermedia will become VR as the number and complexity of the input and output modalites increase. --------------------------------------------------------------------------- NOTE: USE THIS ADDRESS TO REPLY, REPLY-TO IN HEADER MAY BE BROKEN! Bruce Cohen, Computer Research Lab email: brucec@tekcrl.labs.tek.com Tektronix Laboratories, Tektronix, Inc. phone: (503)627-5241 M/S 50-662, P.O. Box 500, Beaverton, OR 97077
beshers@division.cs.columbia.edu (Clifford Beshers) (08/18/90)
In article <BRUCEC.90Aug16120816@phoebus.phoebus.labs.tek.com> brucec%phoebus.phoebus.labs.tek.com@RELAY.CS.NET (Bruce Cohen;;50-662;LP=A;) writes: > > Has anyone done any work on smooth, realistic rotation of text? Almost > all the 3D graphics stuff I've ever seen specializes in blitting > lots of polygons with good shading effects, which doesn't seem very > applicable to the special problems text presents. > Yes, a lot of work has been done. This is essentially an anti-aliasing problem, with image-processing sorts of solutions. See, for instance: That addresses the issue of high quality rendering of text, but if you want to read about a VR system that has text in 3D, look at the papers by Card, Mackinlay and Robertson from Xerox Parc, (not necessarily in that order). They have a system they call the "cognitive co-processor" with a 3D rooms flavor. See ACM User Interface and Software Technology (UIST) '89 and SIGGRAPH 90. There are blackboards on the walls with messages, etc., that pop up to a "head's up display", i.e. rotate from their 3D position into the plane of the screen. The SIGGRAPH paper was about how you build controllers for steering towards a particular object or piece of text that you would like to investigate. -- ----------------------------------------------- Clifford Beshers 450 Computer Science Department Columbia University New York, NY 10027 beshers@cs.columbia.edu
clw%tornado.Berkeley.EDU@ucbvax.Berkeley.EDU (A Ghost in the Machine) (08/21/90)
In article <1990Aug16.134553.29297@zorch.SF-Bay.ORG> xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) writes: > >This brought to mind another human ability, nearly magic, that will be >of _crucial_ experience for navigating virtual reality, and that seems >to have been neglected as yet. This is the "cocktail party" phenomenon: >in a crowd of fifty conversations, our ears can somehow pick out the >one of interest to us, and focus on it to the exclusion of all others. It is my understanding that this ability stems mostly from auditory preprocessing in the brain: the signals from both ears are compared (experiments show that relative amplitude, phase, and possibly timing are taken into account), and sounds are sorted according to location. The shape of the ear eliminates front-to-back symmetry (again, an experimental result). Thus it turns out that much of this ability is lost if the sensitivity balance between ears changes (i.e. partial loss in one ear), even if the actual hearing loss is quite small. People with such minor hearing damage do very poorly picking voices out of crowds or in noisy backgrounds, despite having reasonably acute hearing. An excellent example is to sit in a room of several conversations, and tape it with a single mike. Later, try to pick out what each person was saying from the monaural tape (not just the loudest, all of them). That's what it would be like without binaural hardware and a dedicated preprocessor. clw
mkant@a.gp.cs.cmu.edu (Mark Kantrowitz) (08/25/90)
In theory, 3D text should not be much harder than 2D text. Using an intelligent scan conversion algorithm, first generate the orthogonal projection of the outlines onto the plane of the "screen", then use the algorithm to produce bitmaps. Of course, while the Bitstream and Compugraphic algorithms could be modified to do this, neither would be fast enough unless supported in hardware. One could get good approximations by estimating visual point size, rasterizing at that size, and then doing a small projection. This would be much faster. Ideally, one would need an intelligent scan conversion algorithm which is independent of character orientation. Sadly, there is no such beast at this time. Naive scan conversion would fit the bill, but looks awful at resolutions of 100 dpi. Boost screen resolution to 1000dpi, and you'd have no problems. It also has the benefit of being the fastest method. --mark