man/man1/vwrays.1

.\" RCSid "$Id: vwrays.1,v 1.9 2012/06/14 05:19:05 greg Exp $"
.TH VWRAYS 1 1/15/99 RADIANCE
.SH NAME
vwrays - compute rays for a given picture or view
.SH SYNOPSIS
.B vwrays
.B "[ -i -u -f{a|f|d} -c rept | -d ]"
{
.B "view opts .."
|
.B picture
.B [zbuf]
}
.SH DESCRIPTION
.I Vwrays
takes a picture or view specification and computes the ray origin and
direction corresponding to each pixel in the image.
This information may then be passed to
.I rtrace(1)
to perform other calculations.
If a given pixel has no corresponding ray (because it is outside the
legal view boundaries), then six zero values are sent instead.
.PP
The
.I \-i
option may be used to specify desired pixel positions on the standard
input rather than generating all the pixels for a given view.
If the
.I \-u
option is also given, output will be unbuffered.
.PP
The
.I \-f
option may be used to set the record format to something other than the
default ASCII.
Using raw float or double records for example can reduce the time
requirements of transferring and interpreting information in
.I rtrace.
.PP
The
.I \-c
option repeats each pixel the given number of times (default is 1).
This is most useful when sending rays to
.I rcontrib(1)
with the same
.I \-c
setting, providing a much faster way to average pixels over image sets.
The
.I \-pj
option should be used to jitter sample postions in most cases.
.PP
View options may be any combination of standard view parameters described
in the
.I rpict(1)
manual page, including input from a view file with the
.I \-vf
option.
Additionally, the target X and Y dimensions may be specified with
.I \-x
and
.I \-y
options, and the pixel aspect ratio may be given with
.I \-pa.
The default dimensions are 512x512, with a pixel aspect ratio of 1.0.
Just as in
.I rpict,
the X or the Y dimension will be reduced if necessary
to best match the specified pixel
aspect ratio, unless this ratio is set to zero.
The
.I \-pj
option may be used to jitter samples.
The default value of 0 turns off ray jittering.
.PP
If the
.I \-d
option is given, then
.I vwrays
just prints the computed image dimensions, which are based on the view
aspect and the pixel aspect ratio just described.
The
.I \-ld
switch will also be printed, with
.I \-ld+
if the view file has an aft clipping plane, and
.I \-ld-
otherwise.
This is useful for passing options to the
.I rtrace
command line.
(See below.)
.PP
If the view contains an aft clipping plane
.I (-va
option), then the magnitudes of the ray directions will
equal the maximum distance for each pixel, which will be interpreted
correctly by
.I rtrace
with the
.I \-ld+
option.
Note that this option should not be given unless there is an aft
clipping plane, since the ray direction vectors will be normalized
otherwise, which would produce a uniform clipping distance of 1.
.PP
If a picture is given on the command line rather than a set of view options,
then the view and image dimensions are taken from the picture file, and
the reported ray origins and directions will match the center of each
pixel in the picture (plus optional jitter).
.PP
If a depth buffer file is given as well, then
.I vwrays
computes the intersection point of each pixel ray (equal to the ray origin
plus the depth times the ray direction), and reports this instead of the
ray origin.
The reported ray direction will also be reversed.
The interpretation of this data is an image of origins and directions
for light rays leaving the scene surfaces to strike each pixel.
.SH EXAMPLES
To compute the ray intersection points and returned directions corresponding
to a picture and its depth buffer:
.IP "" .2i
vwrays scene_v2.hdr scene_v2.zbf > scene_v2.pts
.PP
To determine what the dimensions of a given view would be:
.IP "" .2i
vwrays \-d \-vf myview.vf \-x 2048 \-y 2048
.PP
To generate a RADIANCE picture using
.I rtrace
instead of
.I rpict:
.IP "" .2i
vwrays \-ff \-vf view1.vf \-x 1024 \-y 1024 |
rtrace `vwrays \-d \-vf view1.vf \-x 1024 \-y 1024` \-ffc scene.oct > view1.hdr
.SH AUTHOR
Greg Ward Larson
.SH ACKNOWLEDGMENT
This work was supported by Silicon Graphics, Inc.
.SH BUGS
Although
.I vwrays
can reproduce any pixel ordering (i.e., any image orientation) when given
a rendered picture, it will only produce standard scanline-ordered rays when 
given a set of view parameters.
.SH "SEE ALSO"
rcalc(1), rpict(1), rcontrib(1), rtrace(1)
Revision:	1.10
Committed:	Thu Jun 14 22:42:21 2012 UTC (12 years, 10 months ago) by greg
Branch:	MAIN
CVS Tags:	rad5R2, rad4R2P2, rad5R0, rad5R1, rad4R2, rad4R2P1
Changes since 1.9:	+3 -3 lines
Log Message:	Reworked rtcontrib into rcontrib program
#	User	Rev	Content
1	greg	1.10	.\" RCSid "$Id: vwrays.1,v 1.9 2012/06/14 05:19:05 greg Exp $"
2	greg	1.1	.TH VWRAYS 1 1/15/99 RADIANCE
3			.SH NAME
4			vwrays - compute rays for a given picture or view
5			.SH SYNOPSIS
6			.B vwrays
7	greg	1.9	.B "[ -i -u -f{a\|f\|d} -c rept \| -d ]"
8	greg	1.1	{
9			.B "view opts .."
10			\|
11			.B picture
12			.B [zbuf]
13			}
14			.SH DESCRIPTION
15			.I Vwrays
16			takes a picture or view specification and computes the ray origin and
17			direction corresponding to each pixel in the image.
18			This information may then be passed to
19			.I rtrace(1)
20			to perform other calculations.
21			If a given pixel has no corresponding ray (because it is outside the
22			legal view boundaries), then six zero values are sent instead.
23			.PP
24			The
25	greg	1.4	.I \-i
26	greg	1.1	option may be used to specify desired pixel positions on the standard
27			input rather than generating all the pixels for a given view.
28	greg	1.8	If the
29			.I \-u
30			option is also given, output will be unbuffered.
31	greg	1.1	.PP
32			The
33	greg	1.4	.I \-f
34	greg	1.1	option may be used to set the record format to something other than the
35			default ASCII.
36			Using raw float or double records for example can reduce the time
37			requirements of transferring and interpreting information in
38			.I rtrace.
39			.PP
40	greg	1.9	The
41			.I \-c
42			option repeats each pixel the given number of times (default is 1).
43			This is most useful when sending rays to
44	greg	1.10	.I rcontrib(1)
45	greg	1.9	with the same
46			.I \-c
47			setting, providing a much faster way to average pixels over image sets.
48			The
49			.I \-pj
50			option should be used to jitter sample postions in most cases.
51			.PP
52	greg	1.1	View options may be any combination of standard view parameters described
53			in the
54			.I rpict(1)
55			manual page, including input from a view file with the
56			.I \-vf
57			option.
58			Additionally, the target X and Y dimensions may be specified with
59	greg	1.4	.I \-x
60	greg	1.1	and
61	greg	1.4	.I \-y
62	greg	1.1	options, and the pixel aspect ratio may be given with
63	greg	1.4	.I \-pa.
64	greg	1.1	The default dimensions are 512x512, with a pixel aspect ratio of 1.0.
65			Just as in
66			.I rpict,
67			the X or the Y dimension will be reduced if necessary
68			to best match the specified pixel
69			aspect ratio, unless this ratio is set to zero.
70	greg	1.4	The
71			.I \-pj
72			option may be used to jitter samples.
73			The default value of 0 turns off ray jittering.
74	greg	1.1	.PP
75			If the
76	greg	1.4	.I \-d
77	greg	1.1	option is given, then
78			.I vwrays
79			just prints the computed image dimensions, which are based on the view
80			aspect and the pixel aspect ratio just described.
81			The
82	greg	1.4	.I \-ld
83	greg	1.1	switch will also be printed, with
84	greg	1.4	.I \-ld+
85	greg	1.1	if the view file has an aft clipping plane, and
86	greg	1.4	.I \-ld-
87	greg	1.1	otherwise.
88			This is useful for passing options to the
89			.I rtrace
90			command line.
91			(See below.)
92			.PP
93			If the view contains an aft clipping plane
94			.I (-va
95			option), then the magnitudes of the ray directions will
96			equal the maximum distance for each pixel, which will be interpreted
97			correctly by
98			.I rtrace
99			with the
100	greg	1.4	.I \-ld+
101	greg	1.1	option.
102			Note that this option should not be given unless there is an aft
103			clipping plane, since the ray direction vectors will be normalized
104			otherwise, which would produce a uniform clipping distance of 1.
105			.PP
106			If a picture is given on the command line rather than a set of view options,
107			then the view and image dimensions are taken from the picture file, and
108	greg	1.7	the reported ray origins and directions will match the center of each
109			pixel in the picture (plus optional jitter).
110	greg	1.1	.PP
111			If a depth buffer file is given as well, then
112			.I vwrays
113			computes the intersection point of each pixel ray (equal to the ray origin
114			plus the depth times the ray direction), and reports this instead of the
115			ray origin.
116			The reported ray direction will also be reversed.
117			The interpretation of this data is an image of origins and directions
118			for light rays leaving the scene surfaces to strike each pixel.
119			.SH EXAMPLES
120			To compute the ray intersection points and returned directions corresponding
121			to a picture and its depth buffer:
122			.IP "" .2i
123	greg	1.6	vwrays scene_v2.hdr scene_v2.zbf > scene_v2.pts
124	greg	1.1	.PP
125			To determine what the dimensions of a given view would be:
126			.IP "" .2i
127	greg	1.5	vwrays \-d \-vf myview.vf \-x 2048 \-y 2048
128	greg	1.1	.PP
129			To generate a RADIANCE picture using
130			.I rtrace
131			instead of
132			.I rpict:
133			.IP "" .2i
134	greg	1.5	vwrays \-ff \-vf view1.vf \-x 1024 \-y 1024 \|
135	greg	1.6	rtrace `vwrays \-d \-vf view1.vf \-x 1024 \-y 1024` \-ffc scene.oct > view1.hdr
136	greg	1.1	.SH AUTHOR
137			Greg Ward Larson
138			.SH ACKNOWLEDGMENT
139			This work was supported by Silicon Graphics, Inc.
140			.SH BUGS
141			Although
142			.I vwrays
143			can reproduce any pixel ordering (i.e., any image orientation) when given
144			a rendered picture, it will only produce standard scanline-ordered rays when
145			given a set of view parameters.
146			.SH "SEE ALSO"
147	greg	1.10	rcalc(1), rpict(1), rcontrib(1), rtrace(1)