Hello...

1 - I use PoV 3.1 to model various astronomical phenomena (animations
to be soon on the WEB...), and general physics phenomena for education.
I recently modelled (artist's view only, as only a rough sketch
of the physical processes is available !)
the central parts of an active nucleus galaxy (the thing with a black
hole in the center), made it rotate, and so on. Nice, but it rotates as
a solid
object, which is wrong. Does anybody know of a way of "warping"
continuously the "spiral" texture to make it rotate faster in the
central
parts than in the outer parts ? I searched a lot in the (excellent)
in-line doc, with no success...  Maybe I missed something...

2 - I am going to try to write an external program computing a B-spline
approximation to a set of points ( two sets in fact, one for the camera,

the other for the "flying object" ) The best way I know so far is using
PCG2.21 (Patrick's Curve Generator), which takes as input the clock
variable
and a spline.dat ASCII file containing the knots of the path, and
outputs
an ASCII file wtih a set of #declare. The whole thing is invoqued via a
Pre_Frame_Command in the .ini file. This is the structure which is
planned for the code I am about to write. But the PGC program does not
compute the local curvature and twist of the 3D spline, which are needed
to
compute the "correct" attitude of the plane, and cannot compute two
distinct
spline approximations on two data sets for the same clock value.
The "spline" add-on from Chris Colefax is not well suited for multiple
knot
splines, as you must give all the control points to be used for all the
knots;
or did I miss something ?
I tried AnimSys, too, but the "plane" attitude is not computed from the
fly path shape, just given a priori.
Before engaging further into  development, I would like to know
if anybody heard about something addressing this problem...
It should have been solved by all the people working in the flight
simulation
area...

3 - I used (before switching to the last release of PoV for Windows) the

dispersion patch. I was able to render all kinds of gems, to demonstrate

refraction into complex objects during physics courses. Great, but I was

not able to obtain a spectrum-like patch of light on a screen following
a glass ball, although I got nice effects in the ball itself.
Is that normal, or just a matter of wrong code ?
I tried to disperse light through a prism, and get a spectrum on a
screen :
no result...

Many thanks to the fabulous POV team !

Gilles AdamObservatoire de Lyon
69561 Saint Genis Laval cedex
France
ga### [at] obsuniv-lyon1fr

Post a reply to this message

Attachments:
Download 'us-ascii' (1 KB)

I can help out with the 3rd thing anyway. Use UVPov Alpha 5, which has
been recently updated; it's a custom version of POV-Ray by Nathan Kopp. It
can cast light through objects and produce a spectral dispersion. URL is:

 http://nathan.kopp.com/patched.htm

There are certain things you may need to know about using it that may be
found in the newsgroups povray.unofficial.patches and povray.general if
you search for uvpov or photons. Or just ask any questions there if having
a problem.


Gilles ADAM wrote:
> 
> Hello...
> 
> 1 - I use PoV 3.1 to model various astronomical phenomena (animations
> to be soon on the WEB...), and general physics phenomena for education.
> I recently modelled (artist's view only, as only a rough sketch
> of the physical processes is available !)
> the central parts of an active nucleus galaxy (the thing with a black
> hole in the center), made it rotate, and so on. Nice, but it rotates as
> a solid
> object, which is wrong. Does anybody know of a way of "warping"
> continuously the "spiral" texture to make it rotate faster in the
> central
> parts than in the outer parts ? I searched a lot in the (excellent)
> in-line doc, with no success...  Maybe I missed something...
> 
> 2 - I am going to try to write an external program computing a B-spline
> approximation to a set of points ( two sets in fact, one for the camera,
> 
> the other for the "flying object" ) The best way I know so far is using
> PCG2.21 (Patrick's Curve Generator), which takes as input the clock
> variable
> and a spline.dat ASCII file containing the knots of the path, and
> outputs
> an ASCII file wtih a set of #declare. The whole thing is invoqued via a
> Pre_Frame_Command in the .ini file. This is the structure which is
> planned for the code I am about to write. But the PGC program does not
> compute the local curvature and twist of the 3D spline, which are needed
> to
> compute the "correct" attitude of the plane, and cannot compute two
> distinct
> spline approximations on two data sets for the same clock value.
> The "spline" add-on from Chris Colefax is not well suited for multiple
> knot
> splines, as you must give all the control points to be used for all the
> knots;
> or did I miss something ?
> I tried AnimSys, too, but the "plane" attitude is not computed from the
> fly path shape, just given a priori.
> Before engaging further into  development, I would like to know
> if anybody heard about something addressing this problem...
> It should have been solved by all the people working in the flight
> simulation
> area...
> 
> 3 - I used (before switching to the last release of PoV for Windows) the
> 
> dispersion patch. I was able to render all kinds of gems, to demonstrate
> 
> refraction into complex objects during physics courses. Great, but I was
> 
> not able to obtain a spectrum-like patch of light on a screen following
> a glass ball, although I got nice effects in the ball itself.
> Is that normal, or just a matter of wrong code ?
> I tried to disperse light through a prism, and get a spectrum on a
> screen :
> no result...
> 
> Many thanks to the fabulous POV team !
> 
> Gilles AdamObservatoire de Lyon
> 69561 Saint Genis Laval cedex
> France
> ga### [at] obsuniv-lyon1fr

-- 
 omniVERSE: beyond the universe
  http://members.aol.com/inversez/homepage.htm
 mailto://inversez@aol.com?Subject=PoV-News

Post a reply to this message

I'm going to post a "slipping" spiral mpg animation at the
binaries.animations group and the script used to make it at
text.scene-files too (several minutes from now). It's a obvious cheat to
get the spiral to slip past itself at 3 distinct radii. Maybe completely
useless or maybe a way to fake it depending on what exactly is needed. The
main textures could be anything really.


Bob Hughes wrote:
> 
> I can help out with the 3rd thing anyway. Use UVPov Alpha 5, which has
> been recently updated; it's a custom version of POV-Ray by Nathan Kopp. It
> can cast light through objects and produce a spectral dispersion. URL is:
> 
>  http://nathan.kopp.com/patched.htm
> 
> There are certain things you may need to know about using it that may be
> found in the newsgroups povray.unofficial.patches and povray.general if
> you search for uvpov or photons. Or just ask any questions there if having
> a problem.
> 
> Gilles ADAM wrote:
> >
> > Hello...
> >
> > 1 - I use PoV 3.1 to model various astronomical phenomena (animations
> > to be soon on the WEB...), and general physics phenomena for education.
> > I recently modelled (artist's view only, as only a rough sketch
> > of the physical processes is available !)
> > the central parts of an active nucleus galaxy (the thing with a black
> > hole in the center), made it rotate, and so on. Nice, but it rotates as
> > a solid
> > object, which is wrong. Does anybody know of a way of "warping"
> > continuously the "spiral" texture to make it rotate faster in the
> > central
> > parts than in the outer parts ? I searched a lot in the (excellent)
> > in-line doc, with no success...  Maybe I missed something...
> >
> > 2 - I am going to try to write an external program computing a B-spline
> > approximation to a set of points ( two sets in fact, one for the camera,
> >
> > the other for the "flying object" ) The best way I know so far is using
> > PCG2.21 (Patrick's Curve Generator), which takes as input the clock
> > variable
> > and a spline.dat ASCII file containing the knots of the path, and
> > outputs
> > an ASCII file wtih a set of #declare. The whole thing is invoqued via a
> > Pre_Frame_Command in the .ini file. This is the structure which is
> > planned for the code I am about to write. But the PGC program does not
> > compute the local curvature and twist of the 3D spline, which are needed
> > to
> > compute the "correct" attitude of the plane, and cannot compute two
> > distinct
> > spline approximations on two data sets for the same clock value.
> > The "spline" add-on from Chris Colefax is not well suited for multiple
> > knot
> > splines, as you must give all the control points to be used for all the
> > knots;
> > or did I miss something ?
> > I tried AnimSys, too, but the "plane" attitude is not computed from the
> > fly path shape, just given a priori.
> > Before engaging further into  development, I would like to know
> > if anybody heard about something addressing this problem...
> > It should have been solved by all the people working in the flight
> > simulation
> > area...
> >
> > 3 - I used (before switching to the last release of PoV for Windows) the
> >
> > dispersion patch. I was able to render all kinds of gems, to demonstrate
> >
> > refraction into complex objects during physics courses. Great, but I was
> >
> > not able to obtain a spectrum-like patch of light on a screen following
> > a glass ball, although I got nice effects in the ball itself.
> > Is that normal, or just a matter of wrong code ?
> > I tried to disperse light through a prism, and get a spectrum on a
> > screen :
> > no result...
> >
> > Many thanks to the fabulous POV team !
> >
> > Gilles AdamObservatoire de Lyon
> > 69561 Saint Genis Laval cedex
> > France
> > ga### [at] obsuniv-lyon1fr
> 
> --
>  omniVERSE: beyond the universe
>   http://members.aol.com/inversez/homepage.htm
>  mailto://inversez@aol.com?Subject=PoV-News

-- 
 omniVERSE: beyond the universe
  http://members.aol.com/inversez/homepage.htm
 mailto://inversez@aol.com?Subject=PoV-News

Post a reply to this message

The way NASA animates galaxies is to animate each star independantly.
Perhaps you could do something similar by using 10 or 20 concentric rings
instead of millions of points.

steve

Gilles ADAM wrote:

> Hello...
>
> 1 - I use PoV 3.1 to model various astronomical phenomena (animations
> to be soon on the WEB...), and general physics phenomena for education.
> I recently modelled (artist's view only, as only a rough sketch
> of the physical processes is available !)
> the central parts of an active nucleus galaxy (the thing with a black
> hole in the center), made it rotate, and so on. Nice, but it rotates as
> a solid
> object, which is wrong. Does anybody know of a way of "warping"
> continuously the "spiral" texture to make it rotate faster in the
> central
> parts than in the outer parts ? I searched a lot in the (excellent)
> in-line doc, with no success...  Maybe I missed something...
>
> 2 - I am going to try to write an external program computing a B-spline
> approximation to a set of points ( two sets in fact, one for the camera,
>
> the other for the "flying object" ) The best way I know so far is using
> PCG2.21 (Patrick's Curve Generator), which takes as input the clock
> variable
> and a spline.dat ASCII file containing the knots of the path, and
> outputs
> an ASCII file wtih a set of #declare. The whole thing is invoqued via a
> Pre_Frame_Command in the .ini file. This is the structure which is
> planned for the code I am about to write. But the PGC program does not
> compute the local curvature and twist of the 3D spline, which are needed
> to
> compute the "correct" attitude of the plane, and cannot compute two
> distinct
> spline approximations on two data sets for the same clock value.
> The "spline" add-on from Chris Colefax is not well suited for multiple
> knot
> splines, as you must give all the control points to be used for all the
> knots;
> or did I miss something ?
> I tried AnimSys, too, but the "plane" attitude is not computed from the
> fly path shape, just given a priori.
> Before engaging further into  development, I would like to know
> if anybody heard about something addressing this problem...
> It should have been solved by all the people working in the flight
> simulation
> area...
>
> 3 - I used (before switching to the last release of PoV for Windows) the
>
> dispersion patch. I was able to render all kinds of gems, to demonstrate
>
> refraction into complex objects during physics courses. Great, but I was
>
> not able to obtain a spectrum-like patch of light on a screen following
> a glass ball, although I got nice effects in the ball itself.
> Is that normal, or just a matter of wrong code ?
> I tried to disperse light through a prism, and get a spectrum on a
> screen :
> no result...
>
> Many thanks to the fabulous POV team !
>
> Gilles AdamObservatoire de Lyon
> 69561 Saint Genis Laval cedex
> France
> ga### [at] obsuniv-lyon1fr

Post a reply to this message

For the knot problem, have you looked at Knot 3.7? It is a 3D knot design
software that also animates. I have it on the Mac, don't know if it's on
other platforms. It's a fascinating and very well done application. I can
send you some samples if you'd like.

[Image]
steve

Gilles ADAM wrote:

> Hello...
>
> 1 - I use PoV 3.1 to model various astronomical phenomena (animations
> to be soon on the WEB...), and general physics phenomena for education.
> I recently modelled (artist's view only, as only a rough sketch
> of the physical processes is available !)
> the central parts of an active nucleus galaxy (the thing with a black
> hole in the center), made it rotate, and so on. Nice, but it rotates as
> a solid
> object, which is wrong. Does anybody know of a way of "warping"
> continuously the "spiral" texture to make it rotate faster in the
> central
> parts than in the outer parts ? I searched a lot in the (excellent)
> in-line doc, with no success...  Maybe I missed something...
>
> 2 - I am going to try to write an external program computing a B-spline
> approximation to a set of points ( two sets in fact, one for the camera,
>
> the other for the "flying object" ) The best way I know so far is using
> PCG2.21 (Patrick's Curve Generator), which takes as input the clock
> variable
> and a spline.dat ASCII file containing the knots of the path, and
> outputs
> an ASCII file wtih a set of #declare. The whole thing is invoqued via a
> Pre_Frame_Command in the .ini file. This is the structure which is
> planned for the code I am about to write. But the PGC program does not
> compute the local curvature and twist of the 3D spline, which are needed
> to
> compute the "correct" attitude of the plane, and cannot compute two
> distinct
> spline approximations on two data sets for the same clock value.
> The "spline" add-on from Chris Colefax is not well suited for multiple
> knot
> splines, as you must give all the control points to be used for all the
> knots;
> or did I miss something ?
> I tried AnimSys, too, but the "plane" attitude is not computed from the
> fly path shape, just given a priori.
> Before engaging further into  development, I would like to know
> if anybody heard about something addressing this problem...
> It should have been solved by all the people working in the flight
> simulation
> area...
>
> 3 - I used (before switching to the last release of PoV for Windows) the
>
> dispersion patch. I was able to render all kinds of gems, to demonstrate
>
> refraction into complex objects during physics courses. Great, but I was
>
> not able to obtain a spectrum-like patch of light on a screen following
> a glass ball, although I got nice effects in the ball itself.
> Is that normal, or just a matter of wrong code ?
> I tried to disperse light through a prism, and get a spectrum on a
> screen :
> no result...
>
> Many thanks to the fabulous POV team !
>
> Gilles AdamObservatoire de Lyon
> 69561 Saint Genis Laval cedex
> France
> ga### [at] obsuniv-lyon1fr

Post a reply to this message

Attachments:
Download 'us-ascii' (4 KB) Download '/system%20drive/temporary%20items/nsmail8.jpg' (7 KB)

Preview of image '/system%20drive/temporary%20items/nsmail8.jpg'

Bob wrote:
> 
> I'm going to post a "slipping" spiral mpg animation at the
> binaries.animations group and the script used to make it at
> text.scene-files too (several minutes from now). It's a obvious cheat to
> get the spiral to slip past itself at 3 distinct radii. Maybe completely
> useless or maybe a way to fake it depending on what exactly is needed. The
> main textures could be anything really.
> 
> Bob Hughes wrote:
> >
> > I can help out with the 3rd thing anyway. Use UVPov Alpha 5, which has
> > been recently updated; it's a custom version of POV-Ray by Nathan Kopp. It
> > can cast light through objects and produce a spectral dispersion. URL is:
> >
> >  http://nathan.kopp.com/patched.htm
> >
> > There are certain things you may need to know about using it that may be
> > found in the newsgroups povray.unofficial.patches and povray.general if
> > you search for uvpov or photons. Or just ask any questions there if having
> > a problem.
> >

Sorry for beeing so long to reply : a stupid news client problem,
here...
I tried the slipping spiral, and definitely learned some useful things !
But visually, it does fit, as it introduces a visible break in the
spiral arms.
It may not be too difficult to introduce a "variable spiral warp" as a
texture
modifier, but this is beyond my knowledge as a C programmer; putting my
fingers
into POV Ray code, o My God !
I downloaded UV-POV, and it sure will cure my rainbow problem. I did not
realized
that it was able to "replace" the dispersion patch as well...
Gilles
For the spline problem, see the reply to Steve

Post a reply to this message

Steve wrote:
> 
> The way NASA animates galaxies is to animate each star independantly.
> Perhaps you could do something similar by using 10 or 20 concentric rings
> instead of millions of points.
> 
> steve
>
Hello...
Sorry for beeing so late : we encountered a news client problem, here...
what you suggest (animating separately a lot of particles)
can definitely be done, but I should then animate small blobs of 'smoke"
embedded into a foggy medium, starting maybe with a spiral distribution.
But I am afraid of the rendering time. I am -as anybody after the
checkered floor
phase- trying to render
the "best naturally looking high resolution earth from space", and came
to
something I like, with a rendering time in the 30 minutes range on my
K6-233,
if I remember well... Then I decided to add the atmosphere, with
Rayleigh
diffusion, and the time expanded immediately to 2h30. So, I am cautious
about playing with media/interior/scattering and so on...
That is why my "black hole environment" is a texture, not a 3D object, I
confess...
But maybe I am wrong. Someday I shall try that for a galaxy, with a few
thousands
of individual stars; just for fun; just for a few minutes, maybe...

Gilles

Post a reply to this message

Steve wrote:
> 
> For the knot problem, have you looked at Knot 3.7? It is a 3D knot
> design software that also animates. I have it on the Mac, don't know
> if it's on other platforms. It's a fascinating and very well done
> application. I can send you some samples if you'd like.
> 
> steve

Hi

Sorry for the late reply, delayed by a news client problem here...

I do not know KNOTS 3.7, and I am going to search the WEB for that.
Looks
interesting, for sure.
In the mean time, I developped a POV macro, from the basic equations of
cubic
splines. I shall put int on this server as soon as it works fully.
As of July 1st, it computes correctly the flight path, the speed vector,
but fails to roll the plane the correct angle. So the F14 is doing some
strange uncontrolled
aerobatics. I think I am going to  give up taking into
account the plane speed, and compute the roll only from the curvature
radius...
Or maybe not, according to my lazyness ! Not simple, car the "speed" is
not easy to define, when dynamical effects are to be taken into
account... I shall cheat...
O yes, I am going to cheat...
Back to knots 3.7, on this server there is an interesting torspline.inc
macro
file which do something in the same spirit : "dressing" a spline with
torus sections smoothly connected. Very nice job from Ron Parker.

Gilles

Post a reply to this message

On Thu, 01 Jul 1999 18:21:41 +0200, Gilles Adam <ga### [at] obsuniv-lyon1fr>
wrote:

[snip]
>That is why my "black hole environment" is a texture, not a 3D object, I
>confess...
>But maybe I am wrong. Someday I shall try that for a galaxy, with a few
>thousands
>of individual stars; just for fun; just for a few minutes, maybe...
>
Hmmm... black hole enviornment eh? Animated? Interesting. Source
code/pretty pictures please?
Cheers,

Cliff Bowman
Why not pay my 3D Dr Who site a visit at http://www.who3d.cwc.net/
PS change ".duffcom" to ".net" if replying via e-mail

Post a reply to this message

Cliff Bowman wrote:
> Hmmm... black hole enviornment eh? Animated? Interesting. Source
> code/pretty pictures please?

Well... I destroyed the existing animation, thinking that
it was a matter of days, no more, to compute it again 
with, at last, a nice plane flying above.
This was three months ago...
But I do have a "nice picture", which will be on the
net as soon as I find time for that; I would say a few
days. This is what we call, in french, a "promesse de gascon". Promesse
is, of course, "promise", and a gascon is a man from Gascogne, in the
South of France, where people are sayed to be proud, quick-tempered, and
ready to make promises they will eventually not be able to keep. I am
not gascon, but...
The code will be on the net, too, of course.
Gilles

Post a reply to this message