** Proprietary **
Hi Everyone,
As some of my last posts indicate, I am using ranimate to set up a
series of animations. Everything seems to be working, however I am
having a hard time getting the exposures to work out. Because my
interior space is mostly windows, my exposure varies in order to get the
renderings to look normal. If I set the exposure to one value, then
some frames look great while others are way overexposed. It seems that
to correct this problem, a file with exposure settings can be
developed.
However, I will have to go frame by frame and try and figure out what
exposure settings might work in order to develop this file. And if I
make 150 or so images per animation like I am planning, this will be a
very tedious task to accomplish with rview. I guess I can interpolate
and guesstimate, but perhaps there is a better way of doing this?
Anyone run into this issue before?
Mark de la Fuente
[email protected]
Hi Mark,
Check out phisto (http://radsite.lbl.gov/radiance/man_html/phisto_1.htm).
This will help, as will using the -h parameter of pcond (there's an example on the phisto manpage).
Cheers.
···
On Jan 30, 2004, at 6:04 PM, Mark de la Fuente wrote:
edious task to accomplish with rview. I guess I can interpolate
and guesstimate, but perhaps there is a better way of doing this?
Anyone run into this issue before?
=================
Rob Guglielmetti
www.rumblestrip.org
Rob Guglielmetti wrote:
edious task to accomplish with rview. I guess I can interpolate
and guesstimate, but perhaps there is a better way of doing this?
Anyone run into this issue before?
Hi Mark,
Check out phisto (http://radsite.lbl.gov/radiance/man_html/phisto_1.htm).
This will help, as will using the -h parameter of pcond (there's an example on the phisto manpage).
Cheers.
=================
Rob Guglielmetti
www.rumblestrip.org
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
phisto seems to be one useful tool in the process, but the interesting part may be the semi-automatic, yet smooth, expose "drift". It's probably hard to do fully automatic, as a "good" exposure depends on the part of the image you are interested in. I had glanced a bit at the problem in May 2001 (the http://www.pab-opto.de/pers/animation/hist1.mpg animation shows the histogram overlaid on the image), but haven't found a generally nice method.
Personally I was surprised that my digital Nikon handles automatic exposures quite well, apparently a lot of engineering has gone into an intelligent guess as to what is imporant in an image. Similar algorithms plus (spline-based ?) interpolation may be very useful for automatic exposure settings in Radiance animations.
cheers
Peter
···
On Jan 30, 2004, at 6:04 PM, Mark de la Fuente wrote:
--
pab-opto, Freiburg, Germany, http://www.pab-opto.de
However, I will have to go frame by frame and try and figure out what
exposure settings might work in order to develop this file. And if I
make 150 or so images per animation like I am planning, this will be a
very tedious task to accomplish with rview.
It's "rvu" now ... 
I guess I can interpolate
and guesstimate, but perhaps there is a better way of doing this?
Anyone run into this issue before?
If you just want to see everything on your pictures pass them
through pcond(1) using the -h (human visual response) switch.
You can't compare two of them afterwards, of course.
If you want to create compareable pictures you could pcomb(1) a
selection of your images (i. e. 16 scatterd throughout your
animation) and find an average value for the exposure on this
big image. Use this value for all the smaller images.
I could send you an image of one of this combined renderings.
I have no website to make it available to the group, though.
Thomas
···
On Fri, Jan 30, 2004 at 05:04:00PM -0600, Mark de la Fuente wrote:
