I have some scenes with complex lighting models that render terribly slowly, I
would like to find alternative lighting options for these scenes.

Does HDRI lend itself to this sort of thing?

Fake Lighing, any suggestions?

I need reflection, refraction and media effects.  Am I stuck with traditional
lighting and week-long renders?

I saw some experiments with progressive re-rendering of Radiosity images, a
while back.  Any suggestions, ideas?

Regards,

A.D.B.

Post a reply to this message

"Anthony D. Baye" <Sha### [at] spamnomorehotmailcom> wrote:
> I have some scenes with complex lighting models that render terribly slowly, I
> would like to find alternative lighting options for these scenes.
>
> Does HDRI lend itself to this sort of thing?
>
> Fake Lighing, any suggestions?
>
> I need reflection, refraction and media effects.  Am I stuck with traditional
> lighting and week-long renders?
>
> I saw some experiments with progressive re-rendering of Radiosity images, a
> while back.  Any suggestions, ideas?
>
> Regards,
>
> A.D.B.

Can you describe the scene in more detail? Are the light internal or external to
the visible parts of the scene? How do the complex elements (you mention
reflection, refraction and media effects) interact with each other, the scene
elements and the lighting?

There may be ways of faking the scene that would give similar results to a
"real" render, but it really depends on the scene itself.

Cheers,
Edouard.

Post a reply to this message

They say "A picture says a thousand words": So I give you five pictures...

http://speedy.sdsmt.edu/~1305761/images/Art/FountainSquare3.jpg
http://speedy.sdsmt.edu/~1305761/images/FountainCloseup
http://speedy.sdsmt.edu/~1305761/images/Elementaljpg
http://speedy.sdsmt.edu/~1305761/images/lotus01.jpg
http://speedy.sdsmt.edu/~1305761/images/Circle3.jpg

The first image can either be during the day, or at night.  The daylight model
is obviously the simpler one.  The night-time model requires fifteen small
flames in the base of the fountain, illuminating the water from below, and
street lamps around the perimiter, either gas, or oil.

I had an idea for the fountain square that involved many little hanging lamps,
suspended from ropes, or perhaps stone lamp-posts/bollards, I'm still working
out the surroundings.

The third image is something I rendered a LONG time ago, years ago, the
lighting, coupled with radiosity and normals on the walls made it take what
seemed a long time to render (I can't remember which computer I was using at the
time, but it was not my current model)

Image four is the lamp I intend to use in the fountain base.  That alone took
about a week to render.

And Image 5 is a test I did with another scene I'm working on,  The lighting
levels are not nearly sufficient, even though it's supposed to be dark.

Regards,

A.D.B.

"Edouard" <pov### [at] edouardinfo> wrote:
> "Anthony D. Baye" <Sha### [at] spamnomorehotmailcom> wrote:
> > I have some scenes with complex lighting models that render terribly slowly, I
> > would like to find alternative lighting options for these scenes.
> >
> > Does HDRI lend itself to this sort of thing?
> >
> > Fake Lighing, any suggestions?
> >
> > I need reflection, refraction and media effects.  Am I stuck with traditional
> > lighting and week-long renders?
> >
> > I saw some experiments with progressive re-rendering of Radiosity images, a
> > while back.  Any suggestions, ideas?
> >
> > Regards,
> >
> > A.D.B.
>
> Can you describe the scene in more detail? Are the light internal or external to
> the visible parts of the scene? How do the complex elements (you mention
> reflection, refraction and media effects) interact with each other, the scene
> elements and the lighting?
>
> There may be ways of faking the scene that would give similar results to a
> "real" render, but it really depends on the scene itself.
>
> Cheers,
> Edouard.

Post a reply to this message

Forgive me, those addresses should be:

http://speedy.sdsmt.edu/~1305761/images/Art/FountainSquare3.jpg
http://speedy.sdsmt.edu/~1305761/images/FountainCloseup.jpg
http://speedy.sdsmt.edu/~1305761/images/Elemental.jpg
http://speedy.sdsmt.edu/~1305761/images/lotus01.png
http://speedy.sdsmt.edu/~1305761/images/Circle3.png

The last one is about a hundred times as bright in firefox as it is in Windows
picture and fax viewer.

A.D.B.

Post a reply to this message

> They say "A picture says a thousand words": So I give you five pictures...
> 
> http://speedy.sdsmt.edu/~1305761/images/Art/FountainSquare3.jpg
> http://speedy.sdsmt.edu/~1305761/images/FountainCloseup
> http://speedy.sdsmt.edu/~1305761/images/Elementaljpg
> http://speedy.sdsmt.edu/~1305761/images/lotus01.jpg
> http://speedy.sdsmt.edu/~1305761/images/Circle3.jpg
> 
> The first image can either be during the day, or at night.  The daylight model
> is obviously the simpler one.  The night-time model requires fifteen small
> flames in the base of the fountain, illuminating the water from below, and
> street lamps around the perimiter, either gas, or oil.
Looks nice.
> 
> I had an idea for the fountain square that involved many little hanging lamps,
> suspended from ropes, or perhaps stone lamp-posts/bollards, I'm still working
> out the surroundings.
As the lamps are visible in the scene itself, you need them. As there 
are many of them, you can probably use regular point lights, at least 
for most.
Keep those simple.
> 
> The third image is something I rendered a LONG time ago, years ago, the
> lighting, coupled with radiosity and normals on the walls made it take what
> seemed a long time to render (I can't remember which computer I was using at the
> time, but it was not my current model)
> 
> Image four is the lamp I intend to use in the fountain base.  That alone took
> about a week to render.
How large will it be in the final scene? How visible will it be?
It's possible that a largely simplified version would be OK.
As it is now, it use media, complexe geometry, transparency, and maybe 
photons.
Some simplifications:
The center part could be replaced by a simple lathe or sor object with 
two spiral patterns, one scaled <-1,1,1>, white with high ambient and 
transparent.
If you still use media, only use emissive inside the core.
> 
> And Image 5 is a test I did with another scene I'm working on,  The lighting
> levels are not nearly sufficient, even though it's supposed to be dark.
In that one, adding a few faint lights could improve things.
> 
> Regards,
> 
> A.D.B.
> 
The judicious use of spotlight or light_group can possibly improve 
performance.


Alain

Post a reply to this message

Anthony D. Baye wrote:
> The third image is something I rendered a LONG time ago, years ago, the
> lighting, coupled with radiosity and normals on the walls made it take what
> seemed a long time to render (I can't remember which computer I was using at the
> time, but it was not my current model)
> 

If that is the view you intend for the full scene, the normals on the
wall are pretty minor. You could get away with using a second layer of
texture to create that rough pattern without all the added time that the
normals add.

> Image four is the lamp I intend to use in the fountain base.  That alone took
> about a week to render.
> 

Could I get that one in desktop resolution?

You don't mention how you created the glowing structure. If you used
sphere sweeps with a very high sphere count, that could add a good deal
of time to a render. As Alain suggested you could replace it with a sor,
or you might be able to get good performance from an isosurface.

> And Image 5 is a test I did with another scene I'm working on,  The lighting
> levels are not nearly sufficient, even though it's supposed to be dark.

Looking at the image, I am not even sure where the light is coming from.
It looks like the candles and center flames are contributing a lot, from
the highlights on the columns. However, the way the floor seems so
evenly lit doesn't match. Lots of scattering media, perhaps?

If you are using scattering media to darken the room, you might try just
changing the fade_power and fade_distance of the light sources.

Post a reply to this message

"Anthony D. Baye" <Sha### [at] spamnomorehotmailcom> wrote:
> I have some scenes with complex lighting models that render terribly slowly, I
> would like to find alternative lighting options for these scenes.
>
> Does HDRI lend itself to this sort of thing?
>
> Fake Lighing, any suggestions?
>
> I need reflection, refraction and media effects.  Am I stuck with traditional
> lighting and week-long renders?
>
> I saw some experiments with progressive re-rendering of Radiosity images, a
> while back.  Any suggestions, ideas?
>
> Regards,
>
> A.D.B.

If you are looking for overall scene lighting and are familiar with HDR light
probes, you can try my lightdome macro:
http://news.povray.org/povray.binaries.scene-files/thread/%
3Cweb.46488bca4e05d400c150d4c10@news.povray.org%3E/

As it use .hdr format light probes, you need either megaPOV or the 3.7 betas (it
hasn't been tested extensively in 3.7, but it does seem to work as intended)

-tgq

Post a reply to this message

> > Image four is the lamp I intend to use in the fountain base.  That alone took
> > about a week to render.
> >
>
> Could I get that one in desktop resolution?
>

    I don't have it at a larger resolution.  Like I said, it took about a week
to render.  I'll see what happens, perhaps I can optimize some stuff.

> > And Image 5 is a test I did with another scene I'm working on,  The lighting
> > levels are not nearly sufficient, even though it's supposed to be dark.
>
> Looking at the image, I am not even sure where the light is coming from.
> It looks like the candles and center flames are contributing a lot, from
> the highlights on the columns. However, the way the floor seems so
> evenly lit doesn't match. Lots of scattering media, perhaps?
>
> If you are using scattering media to darken the room, you might try just
> changing the fade_power and fade_distance of the light sources.

All of the media objects are contained.

Regards,

A.D.B.

Post a reply to this message

>>> Image four is the lamp I intend to use in the fountain base.  That alone took
>>> about a week to render.
>>>
>> Could I get that one in desktop resolution?
>>
> 
>     I don't have it at a larger resolution.  Like I said, it took about a week
> to render.  I'll see what happens, perhaps I can optimize some stuff.
> 
> Regards,
> 
> A.D.B.
> 
> 
Can you provide the source for it.

After a second look, it looks like there are many small cylinders 
forming the central shape, grouped into a merge. It looks like they are 
transparent, have an ior, reflection and highlights.
Maybe you can use the mesh() built in function for an isosurface. You'd 
need to convert to a polar coordinate to have it whap in the desired 
shape. This is a case where an isosurface could be much faster than some 
CSG object.

If you share the source as it is now, others will be able to take a look 
and possibly come up with some clever alternate approach.


Alain

Post a reply to this message

Anthony D. Baye wrote:
>>> Image four is the lamp I intend to use in the fountain base.  That alone took
>>> about a week to render.
>>>
>> Could I get that one in desktop resolution?
>>
> 
>     I don't have it at a larger resolution.  Like I said, it took about a week
> to render.  I'll see what happens, perhaps I can optimize some stuff.
>

That was meant as a complement, not a request.

>>> And Image 5 is a test I did with another scene I'm working on,  The lighting
>>> levels are not nearly sufficient, even though it's supposed to be dark.
>> Looking at the image, I am not even sure where the light is coming from.
>> It looks like the candles and center flames are contributing a lot, from
>> the highlights on the columns. However, the way the floor seems so
>> evenly lit doesn't match. Lots of scattering media, perhaps?
>>
>> If you are using scattering media to darken the room, you might try just
>> changing the fade_power and fade_distance of the light sources.
> 
> All of the media objects are contained.

Was just my guess.

If I come up with any other ideas on what could be slowing you down I
will post them.

Post a reply to this message