POV-Ray : Newsgroups : povray.binaries.images : trace() accuracy problem Server Time
2 Nov 2024 18:51:14 EDT (-0400)
  trace() accuracy problem (Message 1 to 4 of 4)  
From: Jörg 'Yadgar' Bleimann
Subject: trace() accuracy problem
Date: 21 Feb 2010 22:01:42
Message: <4b81f396@news.povray.org>
High!

Meanwhile, I managed to get surface views from my isosurface desert 
planet ("Ghurghusht")... the four pictures attached here show the same 
landscape from 17,000, 1,000, 200 and 80 metres above ground - with the 
last one, I got the render warning "Camera is inside a non-hollow object".

Ghurghusht has a radius of about 5117 units and is placed 570,000 units 
from the origin... could it be that trace() is that inaccurate? Is there 
any way to get rid of it?

See you in Khyberspace!

Yadgar


Post a reply to this message


Attachments:
Download '2010-02-22 ghurghusht, northern shore of glatzer lacus from 17 kms altitude,take 1.jpg' (37 KB) Download '2010-02-22 ghurghusht, glatzer lacus from 1 km altitude, take 1.jpg' (29 KB) Download '2010-02-22 ghurghusht, glatzer lacus from 200 metres altitude above jawandinsula, take 1.jpg' (28 KB) Download '2010-02-22 ghurghusht, glatzer lacus from 80 metres altitude above jawandinsula, take 1.jpg' (7 KB)

Preview of image '2010-02-22 ghurghusht, northern shore of glatzer lacus from 17 kms altitude,take 1.jpg'
2010-02-22 ghurghusht, northern shore of glatzer lacus from 17 kms altitude,take 1.jpg

Preview of image '2010-02-22 ghurghusht, glatzer lacus from 1 km altitude, take 1.jpg'
2010-02-22 ghurghusht, glatzer lacus from 1 km altitude, take 1.jpg

Preview of image '2010-02-22 ghurghusht, glatzer lacus from 200 metres altitude above jawandinsula, take 1.jpg'
2010-02-22 ghurghusht, glatzer lacus from 200 metres altitude above jawandinsula, take 1.jpg

Preview of image '2010-02-22 ghurghusht, glatzer lacus from 80 metres altitude above jawandinsula, take 1.jpg'
2010-02-22 ghurghusht, glatzer lacus from 80 metres altitude above jawandinsula, take 1.jpg


 

From: Jörg 'Yadgar' Bleimann
Subject: Re: trace() accuracy problem
Date: 21 Feb 2010 22:03:44
Message: <4b81f410$1@news.povray.org>
High!

I forgot to add that 1 POV unit equals 1 kilometre!

See you in Khyberspace!

Yadgar


Post a reply to this message

From: Alain
Subject: Re: trace() accuracy problem
Date: 22 Feb 2010 13:34:43
Message: <4b82ce43@news.povray.org>

> High!
>
> Meanwhile, I managed to get surface views from my isosurface desert
> planet ("Ghurghusht")... the four pictures attached here show the same
> landscape from 17,000, 1,000, 200 and 80 metres above ground - with the
> last one, I got the render warning "Camera is inside a non-hollow object".
>
> Ghurghusht has a radius of about 5117 units and is placed 570,000 units
> from the origin... could it be that trace() is that inaccurate? Is there
> any way to get rid of it?
>
> See you in Khyberspace!
>
> Yadgar
>

Are you tracing against the actual surface, or the water's surface?

For the last image, you may be between the water and the ground. Try to 
make your water transparent and hollow and see what it gives.


Alain


Post a reply to this message

From: Jörg 'Yadgar' Bleimann
Subject: Re: trace() accuracy problem
Date: 22 Feb 2010 17:06:50
Message: <4b82fffa@news.povray.org>
Alain wrote:

> Are you tracing against the actual surface, or the water's surface?
> 
> For the last image, you may be between the water and the ground. Try to 
> make your water transparent and hollow and see what it gives.

I tried that, but it didn't work... but then, I separated the original 
union of isosurface and "water sphere" and traced just against the 
isosurface - and then it worked! Nevertheless, thank you for your hint 
(and be prepared to find your name one day in some of my other planetary 
systems)!

Then I also applied double_illuminate to get rid of the strange 
shadows... below is the result: Lake Glatzer as seen from Jawand Island 
at "pedestrian perspective", i. e. 1.7 metres above ground, looking 
approximately northward (five degrees to the east)...

And now the next, probably much more difficult problem: currently I only 
use a granite pattern for the terrain isosurface (plus some large-scale 
bozo for global height variation), so it generally looks like smooth 
sand dunes or loess hills - which would be perfect if I intended to 
model the whole planet after some parts of northern Afghanistan, but I'd 
rather go for more rugged mountains, as in fact they tower up to 30 
kilometres (!) high in many parts.

I think applying some small-scale noise could do the trick - but how can 
I restrict that noise to medium and high altitudes only while leaving 
the lower parts smooth?

See you in Khyberspace!

Yadgar

Now playing: Anne Boleyn "The Day thou gavest Lord hath ended" (Rick 
Wakeman)


Post a reply to this message


Attachments:
Download '2010-02-22 ghurghusht, glatzer lacus from jawand insula, pedestrian view,take 1.jpg' (24 KB)

Preview of image '2010-02-22 ghurghusht, glatzer lacus from jawand insula, pedestrian view,take 1.jpg'
2010-02-22 ghurghusht, glatzer lacus from jawand insula, pedestrian view,take 1.jpg


 

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.