From: John Wann Subject: Re: SCI: Some notes on J.Wann: What's wrong with your HMD Olaf(Kelle) wrote: >I think it is not necessary to install different depth planes >depending on the visual task. With one fixed value you have >the ability to look in a range from 3m to infinity without any >binocular stress. Accommodation problems are also eliminated. >For the most tasks this will suffice. Olaf is quite right and I have no arguments with his thorough evaluation, its just good to see others taking these issues seriously. My only reservation is that the 3m-infinity range is fine for visualisation tasks, but the least useful for interaction tasks. If you want to manipulate objects (e.g. virtual surgery) then you want to work in close visual space (e.g. 50cm) where the binocular cues will still conflict with the 3m image plane. Of course you could just give the surgeon long arms so his hand could reach out to beyond 3m, but as stereo-acuity drops off markedly in this zone, depth judgements would not benefit greatly from a stereo display and you might as well go back to a monoscopic (bi-ocular) display. I think the most important "zones" to recreate are near (manipulation) space and far (3m +) space, I still think to get the best of both worlds you need to adjust virtual image depth. We will have the facility to do try this in Dec. thanks to collaborative support from Bruce Bassett of Virtual Research. John Wann Edinburgh Very Repetitive Group Ps The CyberEdge article isn't out yet, Olaf had a pre-print of the text, but I believe it will be published soon.