POV-Ray : Newsgroups : povray.binaries.images : Gancaloon: Old City street problem Server Time
30 Jul 2024 06:28:52 EDT (-0400)
  Gancaloon: Old City street problem (Message 11 to 15 of 15)  
<<< Previous 10 Messages Goto Initial 10 Messages
From: s day
Subject: Re: Gancaloon: Old City street problem
Date: 27 Mar 2013 13:30:02
Message: <web.51532c5865932ef94c34011d0@news.povray.org>
Thomas de Groot <tho### [at] degrootorg> wrote:
> The advantage of the urbanism macro is that even at close up the scenes
> are looking good, like in this image.
>
> However...
>
> Thomas

That is a great view, fantastic to have been generated by a macro. You need to
add some dust/dirt to the floors to add more realism, should be easy to do ;-)


Post a reply to this message

From: Kenneth
Subject: Re: Gancaloon: Old City street problem
Date: 27 Mar 2013 18:25:01
Message: <web.5153703b65932ef9c2d977c20@news.povray.org>
Thomas de Groot <tho### [at] degrootorg> wrote:
> On 26-3-2013 9:59, Thomas de Groot wrote:

>
> My solution is to create a street surface related to the highest side of
> the street, extending enough over the other side to enable the trace()
> function to place the houses there at the correct level.
>

These are really nice views of your city!

I assume that you *do* want sloping streets(?) They look very natural like that,
IMO. So if I understand the problem correctly, the sides of the buildings that
are lower down on the slope are being 'buried', while the sides with the steps
and the door are looking too high (that is, the steps etc. are ending before
they reach the ground.)

Although I don't know the method you are using to trace() the buildings onto the
ground (if indeed that *is* the method you're using), it seems to me that the
trace for each building should be done at all four corners--four traces, in
order to get the *highest* of the four. Then, that elevation would be the one to
place the building at. And along with that idea, the buildings/steps etc. would
need to be built so that they *always* extend farther into -y. Most of the time,
that extra building depth would not be seen--it would be underneath the
ground--but in situations such as in your example scene, it would show up and
look natural. (Yes, there would be *lots* of steps, depending on the ground
slope--but that might look quite nice.) I'm thinking that the 'extra' building
depth (and steps) should be deep enough in -y to compensate for, say, a 45-deg.
ground slope(!) as a maximum. Or maybe 35-40 deg.


Post a reply to this message

From: Thomas de Groot
Subject: Re: Gancaloon: Old City street problem
Date: 28 Mar 2013 04:09:47
Message: <5153facb$1@news.povray.org>
On 27-3-2013 23:22, Kenneth wrote:
> These are really nice views of your city!

Thanks Kenneth.

>
> I assume that you *do* want sloping streets(?) They look very natural like that,
> IMO. So if I understand the problem correctly, the sides of the buildings that
> are lower down on the slope are being 'buried', while the sides with the steps
> and the door are looking too high (that is, the steps etc. are ending before
> they reach the ground.)

The 'only' problem is the /sidewise/ slope of the street surface, if I 
make myself clear. So, in any case, that surface has to be corrected, 
otherwise carts will turn over :-)

>
> Although I don't know the method you are using to trace() the buildings onto the
> ground (if indeed that *is* the method you're using), it seems to me that the
> trace for each building should be done at all four corners--four traces, in
> order to get the *highest* of the four. Then, that elevation would be the one to
> place the building at.

I only trace the centre point of the buildings, which obviously is too 
few a test to use on highly uneven surfaces. Your suggestion is good, 
however an extra test would be needed: Depending on which /side/ of the 
street the building is to be situated (high or low end) the /lowest/ or 
the /highest/ test should be selected. Now, I assume that I can know 
this by testing the height of the nearest street boundary to the 
building... Hmm. I have to test this to see how that works.

> And along with that idea, the buildings/steps etc. would
> need to be built so that they *always* extend farther into -y.

They already do ;-)

> Most of the time,
> that extra building depth would not be seen--it would be underneath the
> ground--but in situations such as in your example scene, it would show up and
> look natural. (Yes, there would be *lots* of steps, depending on the ground
> slope--but that might look quite nice.) I'm thinking that the 'extra' building
> depth (and steps) should be deep enough in -y to compensate for, say, a 45-deg.
> ground slope(!) as a maximum. Or maybe 35-40 deg.

Indeed. Still, as I said above, the street surface should not slope too 
much laterally in order to be useful. There still is the need for a 
/built/ street surface...

Thomas


Post a reply to this message

From: Thomas de Groot
Subject: Re: Gancaloon: Old City street problem
Date: 28 Mar 2013 04:12:18
Message: <5153fb62$1@news.povray.org>
On 27-3-2013 18:28, s.day wrote:
> That is a great view, fantastic to have been generated by a macro. You need to
> add some dust/dirt to the floors to add more realism, should be easy to do ;-)

Yes, it is even looking better than I expected. In addition to dust, I 
need also to diversify a bit more the house types. Also easy to do and 
only needing a little adaptation of the macro.

Thomas


Post a reply to this message

From: Christian Froeschlin
Subject: Re: Gancaloon: Old City street problem
Date: 28 Mar 2013 05:35:28
Message: <51540ee0@news.povray.org>
s.day wrote:

> That is a great view, fantastic to have been generated by a macro. 

yes indeed. I think it needs an export to some game
engine terrain system that you can walk around in it ;)


Post a reply to this message

<<< Previous 10 Messages Goto Initial 10 Messages

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