"Hartmut Wagener" <har### [at] arcormailde> wrote in message
news:38bc0183@news.povray.org...
| Yes, i know the problem of the headlights, but did not want to wait another
| 2 or 3 hours. And i was realy happy that media has worked. I have rendered
| some images (in lower resolution) that realy diapointed. The picture was
| black, or it was totaly white. Spotlights with a falloff-value look very
| disapointing, has anyone suggestions how to put a falloff-spotlight in media
| without disapointing "pixels" at the border?

A lot of 'samples' perhaps?

| PS : For a person not talking english all the day, the names of the
| media-parameters are hard to understand, and the documentation says not
| enough. I will try to buy a dictionary to fully understand what every
| parameter means ...

I don't think confidence, variance, ratio really ever need to be used anyway
(unless you want messy results on purpose).  The other keywords are
understandable I would think.

Bob

Post a reply to this message

S far as the headlights cut them off and pull them in so their shape is
restricted bu the cut around the lens...

Hartmut Wagener wrote:

> here is the picture again. Hopefully it is now uploaded ...
>
> Hartmut
>
> Comments ? ...
>
>  [Image]

Post a reply to this message

I like the prism effect by the front wheel, caused by the purple wind
shield...
Very interesting...

Hartmut Wagener wrote:

> here is the picture again. Hopefully it is now uploaded ...
>
> Hartmut
>
> Comments ? ...
>
>  [Image]

Post a reply to this message

Hartmut Wagener wrote:
> 
> What about the parameters of media that i have not used (confidence, rate,
> variation)?

> ... (?)) and a jitter of 1.
> 
	Method 3 isn't faster in and of itself, but it produces
better results. To get the same results with method 1, you
need to push the samples and confidence waaayyy high and the
variance waaayyy low... and then render time goes through
the roof.

	BTW:
* variance and confidence don't mean anything with method 3,
they only change something with method 1, and sometimes
method 2 (if min and max samples are different).
* ratio only changes something for spotlights, but you can
usually ignore it.

		Jerome
-- 

* Doctor Jekyll had something * mailto:ber### [at] inamecom
* to Hyde...                  * http://www.enst.fr/~jberger
*******************************

Post a reply to this message

Thanks, that saves lot of compution time ... And of course most of the
parameters can be understood. Anything about falloff-spotlights? It took
long for me to see that the falloff was disturbing the light-cones. Perhaps
triying a little bit with tightness will bring an answer ...

Hartmut

Bob Hughes schrieb in Nachricht <38bc6f4d@news.povray.org>...

>I don't think confidence, variance, ratio really ever need to be used
anyway
>(unless you want messy results on purpose).  The other keywords are
>understandable I would think.
>
>Bob

Post a reply to this message

It could be that the 'samples' and/or 'intervals' need to be increased.  When
the intensity is lower the granularity might be showing up more and that's what
you are seeing then.  If the media is smooth everywhere but certain places, like
shadow edges, then almost certainly it's those parameters which need to be
raised.  It doesn't have anything to do with the lights themselves, except the
more you have then the more sorting out that has to be done, but that's nothing
to do with 'falloff' values.
All I knew before was that 'ratio' in the media is meant to make changes to the
light edges, if it does indeed affect the falling off of light too then I never
knew that before.  And if so it might need to be set lower (default is 0.9).  It
does say in the help that it's lit to unlit ratio.  I just got it into my mind
that meant edge of the light cone (or cylinder).

Bob

"Hartmut Wagener" <har### [at] arcormailde> wrote in message
news:38bd98f1$1@news.povray.org...
| Thanks, that saves lot of compution time ... And of course most of the
| parameters can be understood. Anything about falloff-spotlights? It took
| long for me to see that the falloff was disturbing the light-cones. Perhaps
| triying a little bit with tightness will bring an answer ...
|
| Hartmut
|
| Bob Hughes schrieb in Nachricht <38bc6f4d@news.povray.org>...
|
| >I don't think confidence, variance, ratio really ever need to be used
| anyway
| >(unless you want messy results on purpose).  The other keywords are
| >understandable I would think.
| >
| >Bob
|
|
|
|

Post a reply to this message

I should point out that I keep reading your "fall-off" to be the keyword
'falloff' in the 'light_source'.  You of course might also be talking about the
edges of the spotlight cones.  Same thing anyway, 'ratio' is intended to change
how the lit and unlit media is sampled.  As the help file states, it is biased
toward lit media so I'm guessing lower than 0.9 will move toward a unlit bias.
And if a ratio of 0 to 1 is the intended range then 0.5 should be an equality.

Bob

"Bob Hughes" <omn### [at] hotmailcom?subject=PoV-News:> wrote in message
news:38bd9ebc@news.povray.org...
| It could be that the 'samples' and/or 'intervals' need to be increased.  When
| the intensity is lower the granularity might be showing up more and that's
what
| you are seeing then.  If the media is smooth everywhere but certain places,
like
| shadow edges, then almost certainly it's those parameters which need to be
| raised.  It doesn't have anything to do with the lights themselves, except the
| more you have then the more sorting out that has to be done, but that's
nothing
| to do with 'falloff' values.
| All I knew before was that 'ratio' in the media is meant to make changes to
the
| light edges, if it does indeed affect the falling off of light too then I
never
| knew that before.  And if so it might need to be set lower (default is 0.9).
It
| does say in the help that it's lit to unlit ratio.  I just got it into my mind
| that meant edge of the light cone (or cylinder).
|
| Bob
|
| "Hartmut Wagener" <har### [at] arcormailde> wrote in message
| news:38bd98f1$1@news.povray.org...
| | Thanks, that saves lot of compution time ... And of course most of the
| | parameters can be understood. Anything about falloff-spotlights? It took
| | long for me to see that the falloff was disturbing the light-cones. Perhaps
| | triying a little bit with tightness will bring an answer ...
| |
| | Hartmut
| |
| | Bob Hughes schrieb in Nachricht <38bc6f4d@news.povray.org>...
| |
| | >I don't think confidence, variance, ratio really ever need to be used
| | anyway
| | >(unless you want messy results on purpose).  The other keywords are
| | >understandable I would think.
| | >
| | >Bob
| |
| |
| |
| |
|
|

Post a reply to this message

Thanks,

i hope that helps me understanding media and spotlights (i use spotlights
very rarely).

I think i will test a little bit media without my converted models, perhaps
the rendering is faster then ...

Hartmut

Bob Hughes schrieb in Nachricht <38bda02d@news.povray.org>...
>I should point out that I keep reading your "fall-off" to be the keyword
>'falloff' in the 'light_source'.  You of course might also be talking about
the

...
>"Bob Hughes" <omn### [at] hotmailcom?subject=PoV-News:> wrote in message
>news:38bd9ebc@news.povray.org...

...

>| |
>| |
>| |
>| |
>|
>|
>
>

Post a reply to this message