 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degroot org> wrote:
> On 3-2-2014 15:44, Paolo Gibellini wrote:
> > Both the carpets seem realistic and fluffy.
> > Very nice works!
>
> Thanks! Different techniques of course, but without LanuHum's work I
> would not have thought of this possibility while it was so very obvious ;-)
>
> Thomas
Thomas, very good works!
But washing would be better if I applied a pigment image of high resolution and
I
reduced random particles a little.
Now there is no time for experiment :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>Thomas de Groot on date 03/02/2014 16.04 wrote:
> This version is better as the proportions of the original carpet have
> now been respected.
>
> Not really an issue, but an even better version would be obtained with
> an uv_mapped image_map on the height_field. However distortions from the
> folds of the height_field are hardly or not visible in the present image.
>
> Thomas
A flying carpet will soon be sold at Gancaloon's court...
;-)
Paolo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Cool stuff LanuHum, Thomas.
Thomas de Groot <tho### [at] degrootorg> wrote:
> height_field + image_map + trace() + eval_pigment() + 500k cylinders
> placed randomly.
>
> Much slower of course: about 16 minutes, of which 15.5 minutes for
> parsing. Writing results to file would make subsequent renders much
> faster of course.
POV-Ray is notoriously slow when it comes to accessing macros from #include
files. Try placing the eval_pigment macro into your main scene file; the parse
time should be significantly reduced by doing so :)
Sam
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 3-2-2014 20:48, Samuel Benge wrote:
> POV-Ray is notoriously slow when it comes to accessing macros from #include
> files. Try placing the eval_pigment macro into your main scene file; the parse
> time should be significantly reduced by doing so :)
Absolutely. About twice as fast.
Here is a close up of the carpet (using cones by the way).
Thomas
Post a reply to this message
Attachments:
Download 'eval_pigmenttest.png' (1078 KB)
Preview of image 'eval_pigmenttest.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
>
> Absolutely. About twice as fast.
>
> Here is a close up of the carpet (using cones by the way).
>
> Thomas
Very good! Excellently! :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
> On 3-2-2014 20:48, Samuel Benge wrote:
> > POV-Ray is notoriously slow when it comes to accessing macros from #include
> > files. Try placing the eval_pigment macro into your main scene file; the parse
> > time should be significantly reduced by doing so :)
>
> Absolutely. About twice as fast.
>
> Here is a close up of the carpet (using cones by the way).
Nice :)
You might be able to speed up the parse time even more. Since eval_pigment()
evaluates the RGB channels of a pigment, you might get a speed increase if you
make a function that samples only one channel, adding color after the fact. A
linear spline could be used in lieu of a color_map...
Sam
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 4-2-2014 17:50, LanuHum wrote:
> Very good! Excellently! :)
Thanks!
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 4-2-2014 19:27, Samuel Benge wrote:
> Nice :)
>
> You might be able to speed up the parse time even more. Since eval_pigment()
> evaluates the RGB channels of a pigment, you might get a speed increase if you
> make a function that samples only one channel, adding color after the fact. A
> linear spline could be used in lieu of a color_map...
I am afraid that this is beyond my understanding :-)
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
> On 4-2-2014 19:27, Samuel Benge wrote:
> > Nice :)
> >
> > You might be able to speed up the parse time even more. Since eval_pigment()
> > evaluates the RGB channels of a pigment, you might get a speed increase if you
> > make a function that samples only one channel, adding color after the fact. A
> > linear spline could be used in lieu of a color_map...
>
> I am afraid that this is beyond my understanding :-)
>
> Thomas
For me it is secret too :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 3-2-2014 16:04, Thomas de Groot wrote:
> Not really an issue, but an even better version would be obtained with
> an uv_mapped image_map on the height_field.
...however, as was concluded in the thread:
http://news.povray.org/povray.newusers/thread/%3Cweb.4f943e8d8677299d5d50475e0%40news.povray.org%3E/?ttop=390889&toff=50
this is not possible to do. A pity.
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
