Thomas de Groot wrote:
> ... and so we are condemned to fly over the Ringworld and never be able to
> land!  <sniff>  :-)
> 
> Indeed, like Mike Williams wrote, one can plug the hole. I think that for
> any close up view, a combination of the larger view with a detailed scene is
> needed. The boundaries would be hardly visible (with some care).
> 
> Thomas
> 

I changed the value of EPSILON and recompiled.  It doesn't have any 
effect.  I'll try playing with it a bit to see what I can discover.

David

Post a reply to this message

Thomas de Groot wrote:
> ... and so we are condemned to fly over the Ringworld and never be able to
> land!  <sniff>  :-)
> 
> Indeed, like Mike Williams wrote, one can plug the hole. I think that for
> any close up view, a combination of the larger view with a detailed scene is
> needed. The boundaries would be hardly visible (with some care).
> 
> Thomas
> 
> 

Ok, I got it.  The problem is in the source file cones.cpp.  It declares 
a constant called Cone_Tolerance to be 1.0e-6.  When I change this to
1.0e-9, I can move the camera down to 0.1 without problems.  I see the 
same artifacts at 0.01. I may be able to reduce it further.

I can provide an unofficial Windows executable to anyone who wants to 
try it.

David Buck

Post a reply to this message

Whoops, did it again. Posted something interesting on a Friday, didn't touch
the net until Monday and missed it all. :)

I suspected it would be something like that as long ago as 2000, when I
originally built a Ringworld. I never noticed these particular artifacts
because I never put the camera close to the ground - I was 2000 miles above
the ring, looking along the rim mountains. I had a lot of difficulty with
the textures and object positions that I couldn't blame on stupidity, and
ended up just making do.

David Buck <dav### [at] simberoncom> wrote:
> I can provide an unofficial Windows executable to anyone who wants to
> try it.
That'd be brilliant - I've been making Fist-Of-God mountain this weekend and
the clouds are definitely too high up! I was going to compensate by making
the mountain 20x bigger, but it would be subtly wrong.

By email or weblink? Either's good for me.

I'm kind of surprised that nobody's ever noticed this issue before...

Bill

Post a reply to this message

Bill Pragnell wrote:
> David Buck <dav### [at] simberoncom> wrote:
> 
>>I can provide an unofficial Windows executable to anyone who wants to
>>try it.
> 
> That'd be brilliant - I've been making Fist-Of-God mountain this weekend and
> the clouds are definitely too high up! I was going to compensate by making
> the mountain 20x bigger, but it would be subtly wrong.
> 
> By email or weblink? Either's good for me.
> 
> I'm kind of surprised that nobody's ever noticed this issue before...
> 
> Bill
> 

You can get it from the following link:

   http://www.simberon.com/povringworld.zip

Unzip the file and drop it into the bin directory of the POV-Ray 
directory in Program Files.  Be sure to backup the old version already 
in that directory.  This version is based on 3.6.0.

Let me know if this works for you.

David Buck

Post a reply to this message

I downloaded the file OK, but the program won't run. Instead, it whinges
thusly: "Editor DLL initialisation failed [LoadLibrary failed, code is
00000485]".

I've tried this in 3.5 and 3.6.1. I can't find a windows version of 3.6.0
anywhere (typical!). I suppose I could always call it from a command
prompt, but I'm not immediately sure how to do this...

Any advice? I appear to need a 3.6.0 release, if such a thing exists.

Bill

Post a reply to this message

Bill Pragnell wrote:
> I downloaded the file OK, but the program won't run. Instead, it whinges
> thusly: "Editor DLL initialisation failed [LoadLibrary failed, code is
> 00000485]".
> 
> I've tried this in 3.5 and 3.6.1. I can't find a windows version of 3.6.0
> anywhere (typical!). I suppose I could always call it from a command
> prompt, but I'm not immediately sure how to do this...
> 
> Any advice? I appear to need a 3.6.0 release, if such a thing exists.
> 
> Bill
> 

The 3.6.0 source code is the latest I can access, so I can't make the 
change on 3.6.1.  I've asked Chris Cason to make the change in the 
latest stream so it will appear in the 3.7 betas.  He'll have to decide 
whether or not to make the change.

David Buck

Post a reply to this message

David Buck <dav### [at] simberoncom> wrote:
> The 3.6.0 source code is the latest I can access, so I can't make the
> change on 3.6.1.  I've asked Chris Cason to make the change in the
> latest stream so it will appear in the 3.7 betas.  He'll have to decide
> whether or not to make the change.

Everything's alright! I found a copy of 3.6.0 on one of the mirror sites,
plugged in the new pvengine, fired it up and it works a treat. I'll post
some results tomorrow (when it's finished rendering!).

It seems to be running more slowly that the previous version - is there a
time cost for this alteration or is it something I've done? If there's a
demand for it maybe the tolerance could go into global_settings or
something...

Thanks anyway David! That's end-user support above and beyond the call of
duty, I'd say. Microsoft, take note... :)

Bill

Post a reply to this message

Bill Pragnell wrote:
> David Buck <dav### [at] simberoncom> wrote:
> 
>>The 3.6.0 source code is the latest I can access, so I can't make the
>>change on 3.6.1.  I've asked Chris Cason to make the change in the
>>latest stream so it will appear in the 3.7 betas.  He'll have to decide
>>whether or not to make the change.
> 
> 
> Everything's alright! I found a copy of 3.6.0 on one of the mirror sites,
> plugged in the new pvengine, fired it up and it works a treat. I'll post
> some results tomorrow (when it's finished rendering!).
> 
> It seems to be running more slowly that the previous version - is there a
> time cost for this alteration or is it something I've done? If there's a
> demand for it maybe the tolerance could go into global_settings or
> something...
> 
> Thanks anyway David! That's end-user support above and beyond the call of
> duty, I'd say. Microsoft, take note... :)
> 
> Bill

Well, I'm not part of the POV-Ray development team any more and haven't 
been for over 12 years, so it's not really end-user support - consider 
it peer-to-peer support.  In any event, it's my pleasure.  I always like 
seeing creative scenes that people make with POV-Ray and if I can do 
some small amount of work to help them, I'm happy.  I look forward to 
seeing your pictures.

David Buck

Post a reply to this message

"David Buck" <dav### [at] simberoncom> schreef in bericht
news:436660f9$1@news.povray.org...
> Well, I'm not part of the POV-Ray development team any more and haven't
> been for over 12 years, so it's not really end-user support - consider
> it peer-to-peer support.  In any event, it's my pleasure.  I always like
> seeing creative scenes that people make with POV-Ray and if I can do
> some small amount of work to help them, I'm happy.  I look forward to
> seeing your pictures.
>

Thanks a lot David! I shall also download the zip and 3.6.0 version and get
to work!

Thomas

Post a reply to this message

David Buck wrote:
High!

> There's a value called EPSILON (see frame.h) which defines a small value 
> and anything smaller would be considered 0 in some of the calculations. 
>  This value is set to 1e-10 which is dangerously close in magnitude to 1 
> / 90,000,000 (the scale of Ringworld).

To which version of PoV-Ray do you refer? When I played around with very 
small objects some years ago (must have been version 3.1), I found out 
that everything smaller than about 1/12,000 unit was not rendered 
anymore... so I wonder that nowdays epsilon should have been lowered 
down to 1e-10, which is a cool six orders of magnitude smaller!

See you in Khyberspace!

Yadgar

Now playing: Tubular Bells Part 2 (Mike Oldfield)

Post a reply to this message