From: cph@dmu.ac.uk (Chris Hand) Subject: Re: INFO: Homebrew at conferences (Meckler UK) Date: Fri, 28 May 1993 17:46:07 GMT Organization: De Montfort University, Leicester, UK (Hi Richard!) R.J. Burton (rjb@leicester.ac.uk) wrote: : > I have no doubt whatsoever that homebrew VR should be encouraged; : > its where many of today's VR companies started afterall. But who's : > going to do it? : > : > Congratulations Meckler USA, bad show Meckler London : : I was interested to read this item, as I didn't go to this years : Meckler Conference. However, I did go last year, and home-brew was : represented by a group of people called the VR Users group. I went to the VR93 exhibition in London and it was very small. Only a dozen or so stalls, and nothing at all in terms of "homebrew VR". Whether this is due to Meckler not taking the UK-VR scene seriously, or whether there is no UK-VR scene anyway is open to debate. One or two companies are doing well in what they do (ARRL, Division, W Industries), but we still have this problem with taking ourselves seriously here I think. Anyway, since the topic came up, here's a short write-up I did of the VR93 exhibition on the train back to Leicester, using my pen-based notepad (and biro ;-] ). Chris ------------------ Report on VR93 Exhibition. Olympia, London. 6th April 1993 Chris Hand (cph@dmu.ac.uk) Department of Computing Science De Montfort University, Leicester As part of the Virtual Reality International 93 conference (VR93), organisers Meckler arranged a parallel exhibition showcasing VR- related products. The exhibition was much smaller than the conference's delegate list would have suggested. Also, in a VR community still confused over the fate of VPL, gloves were conspicuous by their total absence -- many demos instead made use of "free" joysticks based on Polhemus electro- magnetic 6-DOF tracking devices. Division were present, with a demonstration of their new dVS 2.0 system. The demo allowed two users to don a Virtual Research Flight Helmet and navigate a virtual office. The display provided by the VR helmet was disappointing in that the RGB triads of the LCD panels were very clearly visible. Despite headphones mounted in the unit, all that could be heard most of the time was low-level white noise. On a nearby stand, Transformation Software (Thame, Oxfordshire) were demonstrating ModelGen, a software application running on a Silicon Graphics machine. Of greater interest was the Fake Space Labs BOOM system being used to provide stereoscopic images -- with a spatial resolution of 1280 x 1024 with 64 grey levels, this display is surely state-of-the-art, making the LCD-based displays look extremely crude in comparison. The BOOM system comprises two CRT's in a metal box with handles at either side for manipulation. Rather than mounting the displays on the user's head, when viewing is required the eyes are placed in front of the unit, held on a counter-balanced boom. Hand tracking was provided by a polhemus 6 DOF tracking device, glued to a normal mouse (!) with a fabric loop over the top allowing it to be attached to the hand. The quality of the display was very impressive, though some aspects of the software highlighted user interface issues. For example, a "menu" of graphical objects was available for selection (by grasping the object); the lighting model used in rendering was constant shading and no shadows were cast by the virtual hand. In addition the user (or at least I) had no idea what the size of such a "menu" should be in relation to a virtual hand, especially since the objects displayed included a "sun", a cloud and an aeroplane. Consequently it was very difficult to judge depth and actually grasp these objects. Back firmly in the physical world, Bio Control Systems (Palo Alto, California USA) were showing their BioMuse bio-signal processing computer. Capable of picking up and digitising electrical signals from muscle activity and even from brainwaves, this unit looks like extending the boundaries of what can be done using computers, or at least ways of doing things we do already. An enthusiastic exhibition- goer was creating a musical performance using MIDI instruments linked to the BioMuse via a Macintosh running the MAX software package. Sensors strapped around the upper forearm provided electrical signals when the hand was clenched or fingers moved. An additional strap around the forehead (with extensions onto the cheekbones) was able to detect eye movement through pick-up of eye muscle activity. With the left arm's output mapped to amplitude and the right arm controlling pitch and MIDI note on/off, the user was able to create dynamically varying (and more than a little avant-garde) music. I was assured by BioControl president Hugh Lusted that it's actually possible to decipher the muscular signals to such an extent that tracking of individual finger position would be possible. The clever part seems to be involved in extracting the useful information from the electro-myogram (EMG) signals; the company currently has three patents pending related to the BioMuse system. The only other noteworthy exhibits were related to products that would not initially appear to be impressive, but on further reflection may prove to be significant. Shooting Star Technologies has to have the cheapest head-tracker available: for around US$ 1,500 their system physically tracks head location using a mechanical multi-axis boom with potentiometers in the joints. As comical or crude as this may seem, the unit does work well and is fairly smooth. Obviously the user is tethered to the unit, although this may not always be a disadvantage. The tracker seems capable of following the head anywhere the user can put it without moving out of the chair or standing up. The only reservation I had was that there was a noticeable mass on the top of my head, causing a distracting inertial effect whenever I stopped moving. The second surprise came from InfoDisp (Bordon, Hants), who are sell- ing the "Private Eye" display. I had assumed this device had disap- peared because it has had practically no mention in the literature since its launch (perhaps with the exception of Randy Pausch's famous "Virtual Reality on $5 a Day" paper, in which he built a HMD using two of the devices). The unit uses a vibrating mirror and a column of red LED's to create an image of 720 x 280 pixels (capable of emulating CGA on a PC). The image is very clear, except for a few horizontal artefacts, and includes a focus adjustment control. Again, I only had one reservation and in this case it was that the image is bright red, hardly the easiest colour on the eye. I was informed that a colour version would be produced, although this was likely to take 3 to 4 years since the yield on blue LED's is still too low to make integrated LED arrays feasible. InfoDisp are currently working on using two Private Eyes to create a stereo HMD similar to that created by Randy Pausch. The current Private Eye is being sold by InfoDisp as a development kit with an interface for an IBM PC, at a price of L700. Despite one or two useful displays, this exhibition was very small and in no way represented the state of the global VR industry. I can only assume that the effects of the recession are still being felt, although I suspect that the "other" Meckler VR93 exhibition and conference to be held in San Jose, California at the end of May will elicit a more enthusiastic response from exhibitors. ---ENDS---8x--- -- Chris Hand, Lecturer Internet mail: cph@dmu.ac.uk Dept of Computing Science, Voice: +44 533 551551 x8476 De Montfort University, The Gateway, Fax: +44 533 541891 FIDOnet: 2:440/32.50 LEICESTER, UK LE1 9BH >> Linux: *free* unix for IBM PCs! <<