From: throopw@sheol.org (Wayne Throop)
Newsgroups: rec.arts.sf.tv.babylon5.moderated
Subject: Re: C&C starfield (Q&A)
Distribution: world
Organization: sheol
Message-ID: <842198339@sheol.org>
Date: Sun, 08 Sep 1996 15:58:59 GMT
References: <841870882@sheol.org> <322F5BEF.5335@rconnect.com> <842032220@sheol.org> <3231F6D6.CBB@rconnect.com>
:: throopw@sheol.org (Wayne Throop)
:: The spin center still has to be at the center of the window.
: Chris Abbey <cabbey@rconnect.com>
: When the spin center and the LOS are centered in the window there is a
: limit to the radii which can be seen through the window, my guess is
: about 1/2 the diagonal measure of the window will be the longest
: visible. This length is shorter that the 15' possible radius of a 30'
: disk. This is the drastic reduction of which I spoke, further, that
: if the rotational center was moved to a point, say 10 foot from the
: center of the window, you would be able to get arcs with radii between
: 10ft +/- 0.5*width of window in direction moved.
Ah, I think I follow. At least; I *think* I do.
The above looks to me like you want to trade off the size of a disk that
remains stationary during all shots, vs one that you keep very close to
the window, and offset depending on the current POV placement.
Consider this diagram:
Depicted are two POV points within a "C&C floorspace, showing their
angles of view out the window. Also shown are two disks. Now, for an
offset POV, it looks like you can reduce the needed size of the disk,
if you offset the spin axis; yet you map all the needed star points
and trajectories right onto the larger disk.
Two points. First, if you look very close, you'll see that offsetting
the disk seems to make its needed size, at a fixed distance, larger; it
makes it worse. That is, if the offset disk were at the same distance
as the centred one, it'd need to be the same size, and as you move it
closer (or further), it is actually larger than a scaled, centered disk
at the same distance, in order to cover the needed POVs.
But second, it turns out, that's only so with the centered disk
distant from the window. The distance from POV to window forms
a boundary case; as the offset disk nears the window, it will have
an advantage over the centered disk, in that the centered disk will
become unrealistic (as its total viewpoint angle gets small), and
the offset disk retains the reaism of a centered disk farther away.
That is, offsetting the disk allows it to be brought closer with
equal realism WRT a disk at "infinite" distance. This would work
as long as we can change the disk offset at any change of POV.
Is that what was meant?
Just looking at the diagram, it looks like this trick might about halve
the needed disk size from 4 or 5 times the window width, to 2 or 3
times. Just at a quick estimateby eye: I haven't put numbers to it.
--
Wayne Throop throopw@sheol.org http://sheol.org/throopw
throopw@cisco.com
Depicted are two POV points within a "C&C floorspace, showing their
angles of view out the window. Also shown are two disks. Now, for an
offset POV, it looks like you can reduce the needed size of the disk,
if you offset the spin axis; yet you map all the needed star points
and trajectories right onto the larger disk.
Two points. First, if you look very close, you'll see that offsetting
the disk seems to make its needed size, at a fixed distance, larger; it
makes it worse. That is, if the offset disk were at the same distance
as the centred one, it'd need to be the same size, and as you move it
closer (or further), it is actually larger than a scaled, centered disk
at the same distance, in order to cover the needed POVs.
But second, it turns out, that's only so with the centered disk
distant from the window. The distance from POV to window forms
a boundary case; as the offset disk nears the window, it will have
an advantage over the centered disk, in that the centered disk will
become unrealistic (as its total viewpoint angle gets small), and
the offset disk retains the reaism of a centered disk farther away.
That is, offsetting the disk allows it to be brought closer with
equal realism WRT a disk at "infinite" distance. This would work
as long as we can change the disk offset at any change of POV.
Is that what was meant?
Just looking at the diagram, it looks like this trick might about halve
the needed disk size from 4 or 5 times the window width, to 2 or 3
times. Just at a quick estimateby eye: I haven't put numbers to it.
--
Wayne Throop throopw@sheol.org http://sheol.org/throopw
throopw@cisco.com
