Hi all...

Mark here. I am extremely new to POVray, and Ihave a couple quick questions.
I model an object in Moray, and render it in POVray, but a lot of my
objects(including some .udo files I import) have really bad aliasing on the
edges that I cannot seem to figure out. I have tried all the settings I can
to get these to work. Basic objects like Spheres and cones, etc. don't seem
to have this problem. If I make an object...say a column, that uses a few
merged shapes, the whole thing looks like crap. I had tried raising the
resolution levels, the antialiasing levels, and even the #x#
settings....HELP! I really think I could be good at this stuff if I could
get past these small problems I keep running into. Any help would be
appreciated.

Thanks,

Mark

Post a reply to this message

Hi all,

Mark again. I re-read the original post I made, and after a quick crash
course in ray trace nomenclature, I found that I actually have a bad
"aliasing" problem. The edges of my renderings have a "stepped" or "jagged"
look to them, and I would like them to go away. I have rendered a few
spheres over planes (standard newbie fare I hear) and they turn out nice,
but I can't get a complex model(by my standards anyway) to render nicely. Is
there anyone who can tell me what settings to specify to get the nicest
renderings? If I knew the maximum setting, and the minimum as well as what
the setting is and where to place it, I think I can work out the rest.
Please, can someone help? I have some nice models I have made, and would
like to see them realized in a nice rendering.

Thanks and regards,

Mark

Post a reply to this message

"Mark" <haz### [at] comcastnet> wrote in message
news:3d462ee9$1@news.povray.org...
> Hi all...
>

Hi...

<snip>

Hmm, it sounds as though you just need to turn on anti-aliasing. Assuming that
you're using pov for windows, just select one of the AA formats for rendering in
the output dropdown, or specify +A 0.3 in the command line.

However, you say that when you render a sphere or other primitives, the "jaggys"
are not present, which is a bit confusing, since spheres are very subject to
jaggys.

Post a reply to this message

"Tom Melly" <tom### [at] tomandlucouk> wrote:
> "Mark" <haz### [at] comcastnet> wrote in message
> news:3d462ee9$1@news.povray.org...
> <snip>
>
> Hmm, it sounds as though you just need to turn on anti-aliasing. Assuming
that
> you're using pov for windows, just select one of the AA formats for
rendering in
> the output dropdown, or specify +A 0.3 in the command line.
>
> However, you say that when you render a sphere or other primitives, the
"jaggys"
> are not present, which is a bit confusing, since spheres are very subject
to
> jaggys.

Mark wrote:

> merged shapes, the whole thing looks like crap. I had tried raising the
> resolution levels, the antialiasing levels, and even the #x#
> settings....HELP! I really think I could be good at this stuff if I could

Since he raised the "antialiasing levels", that means he tried that...

I am not a PoV-Ray guru, so I can't really help :-(

Perhaps Mark could post a simple model showing this problem on p.b.s-f, so
experts can take a look.

-- #=--=#=--=#=--=#=--=#=--=#=--=#=--=#=--=# --
Philippe Lhoste (Paris -- France)
Professional programmer and amateur artist
http://jove.prohosting.com/~philho/

Post a reply to this message

"Philippe Lhoste" <Phi### [at] GMXnet> wrote in message
news:3d467bdc$1@news.povray.org...
>
> Since he raised the "antialiasing levels", that means he tried that...
>

Well, it means that it sounds like he tried that.... ;)

> Perhaps Mark could post a simple model showing this problem on p.b.s-f, so
> experts can take a look.
>

probably the best idea.

Post a reply to this message

You should look at the final image, which is saved by POV-Ray to the
directory configured in Moray or POV-Ray itself (probably the same directory
where the scene file itself resides), with some image viewer.
  Moray scales the image to the preview window, which can cause some
artifacts which aren't in the actual rendered image.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

In article <3d4696be@news.povray.org> , Warp <war### [at] tagpovrayorg>  wrote:

>   You should look at the final image, which is saved by POV-Ray to the
> directory configured in Moray or POV-Ray itself (probably the same directory
> where the scene file itself resides), with some image viewer.
>   Moray scales the image to the preview window, which can cause some
> artifacts which aren't in the actual rendered image.

I think none of you guys read his second post (maybe it wasn't there, the
date suggests it was however) and didn't carefully read his first one
either.  he used some incorrect terminology in the first one but still
described his problem accurately....

If you carefully read the first post you find that he reports the problem
only with objects from udo files.  AFAIK (I don't use Moray) those are
simply triangle data files.  So the problem is not with POV-Ray but with the
input data triangles.  I don't know what options Moray has to smooth
triangle meshes, but I guess there is some, so maybe someone could explain
him how to do that?

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

"Thorsten Froehlich" <tho### [at] trfde> wrote in message
news:3d46ac0c@news.povray.org...
>
>
> If you carefully read the first post you find that he reports the problem
> only with objects from udo files.

Well, no, he says that

> but a lot of my
> objects(including some .udo files I import) have really bad > aliasing

which rather implies that the problem is not limited to .udo's.

Post a reply to this message

Hi all,

Mark again. Okay, I see you guys have replied to my original post, and
thanks. I have been wrecking my brain trying to find more info in the
tutorials/manuals out there, and I find that they all assume a little
understanding of raytracing. I have found out how to set the anti aliasing
in povray, and I assume that a threshold of 0.0 is the best one can get. I
rendered a couple scenes at this setting and it improved a lot, but I still
have problems with aliasing. I have also noticed that if I dolly the camera
away from the model it cleans up. I gather that the closer the zoom on the
object, the more it will alias. The scene I would like to render has
several close up views of objects (cone/torus merged, kind of like a dunce
cap with a moulding around the opening for the head) I also noticed that
when I have the threshold set to 0.0 the pixels that were aliased terribly
at no AA were now kind of "off the edge" or seperated from the model. It
looks as if particles of the model are coming off the edge slightly( I
don't think I am explaining this well) Sort of like the model was shedding
dust from itself. My question now is this: Is there a way to make a model
have sharp edges at a very close view, and if so how? If you need more
information from me to help, I will be watching this forum a few times a
day.

Thanks,

Mark

Post a reply to this message

Perhaps if you post some sample images in p.b.i, we can better see the
problem you are having.

-tgq

Post a reply to this message

On Tue, 30 Jul 2002 13:10:35 -0400, "TinCanMan" <Tin### [at] hotmailcom>
wrote:
> Perhaps if you post some sample images in p.b.i, we can better see the
> problem you are having.

Since he is posting in povray.newusers it can be worth to explain p.b.i means
povray.binaries.images group designed for posting images.

ABX

Post a reply to this message

> Since he is posting in povray.newusers it can be worth to explain p.b.i
means
> povray.binaries.images group designed for posting images.
>
> ABX

Mark, what he said.

:)

-tgq

Post a reply to this message

Hi Mark!

As you mention a torus, it's most likely, that you reached the boundaries of
floating-point-precision. There had been problems with tori reported before
and floating point precision was the answer even with quite reasonable
scales of the inner and outer radius. The closed form of the torus somewhat
"exploded" into a bunch of single points IIRC. As far as I know there is no
easy way to solve the problem (which has nothing to do with anti-aliasing).
Changing the scale of the object might help (make it bigger and move the
camera farther away).

As other mentioned, posting some sample-code and a picture of the problem
will help to look into your problem.

Regards,

Marc-Hendrik

Post a reply to this message

Okay, I am posting a small pic in the binaries section, it is .png format so
everyone should be able to view it. The image was rendered at 320x240 with
an AA threshold of 0.0 I hope it is enough for you to see my problem and
help out. By the way, don't judge me on the content...I modeled that in
like 5 minutes just so you guys could see the problem. Thanks,  Mark

Post a reply to this message

In article <web.3d4720047fe0707b634c49e20@news.povray.org>,
 "Mark21" <nomail@nomail> wrote:

> Okay, I am posting a small pic in the binaries section, it is .png format so
> everyone should be able to view it. The image was rendered at 320x240 with
> an AA threshold of 0.0 I hope it is enough for you to see my problem and
> help out. By the way, don't judge me on the content...I modeled that in
> like 5 minutes just so you guys could see the problem. Thanks,  Mark

Looks like that "jagged edge" is just the cone poking through the side 
of the torus. Just reduce the radius of that end of the cone or 
otherwise adjust the cone so it doesn't go through like that.

-- 
Christopher James Huff <chr### [at] maccom>
POV-Ray TAG e-mail: chr### [at] tagpovrayorg
TAG web site: http://tag.povray.org/

Post a reply to this message

Christopher James Huff wrote:
>In article <web.3d4720047fe0707b634c49e20[at]news.povray.org>,
>
>Looks like that "jagged edge" is just the cone poking through the side
>of the torus. Just reduce the radius of that end of the cone or
>otherwise adjust the cone so it doesn't go through like that.
>
The model shown is just an example, but that is the way it will be in the
real scene. I am talking about the jaggies that are on the edges of the
cone itself. Is this perhaps a problem with how an object protrudes through
another? Thanks for the reply, but I need the cone to poke through.

Thanks,

MArk

Post a reply to this message

ABX wrote:
>On Tue, 30 Jul 2002 13:10:35 -0400, "TinCanMan" <Tin### [at] hotmailcom>
>wrote:
>> Perhaps if you post some sample images in p.b.i, we can better see the
>> problem you are having.
>
>Since he is posting in povray.newusers it can be worth to explain p.b.i means
>povray.binaries.images group designed for posting images.
>
>ABX
>
 heheh I figured that one out...Been a news user for quite a while :)

Post a reply to this message

Mark21 wrote:

> Okay, I am posting a small pic in the binaries section, it is .png format
> so everyone should be able to view it. The image was rendered at 320x240
> with an AA threshold of 0.0 I hope it is enough for you to see my problem
> and help out. By the way, don't judge me on the content...I modeled that
> in
> like 5 minutes just so you guys could see the problem. Thanks,  Mark

As others already have pointed out you can use AA method 2 with a high 
recursion level and a threshold of 0.0. This should work but is really 
slow. An alternative is to render the picture 4x bigger with normal AA 
settings and scale it down in a graphic program.

You could also try my modified version of POV-Ray which brings some new 
AA-methods: http://www.povworld.org/povray/aa.html 
It is only POV 3.1 though and there is no windows binary.

- Micha

-- 
http://objects.povworld.org - the POV-Ray Objects Collection

Post a reply to this message

Mark21 <nomail@nomail> wrote:
> and I assume that a threshold of 0.0 is the best one can get.

  Actually no, but that's another story in itself.
  (Well, ok, strictrly speaking it's the best *threshold* one can get, but
it's not the best *antialiasing* one can get.)

> I rendered a couple scenes at this setting and it improved a lot, but I still
> have problems with aliasing.

  Could you please tell us whether you are looking at the actual rendered
image, which is saved on disk when you render it, or at the preview image
which Moray shows scaled inside itself (in the fourth subwindow, or however
you have configured it). It would help a lot to know this small detail.

  If you are rendering at a resultion of, let's say, 320x240, and the
preview window of Moray is, let's say 400x300 pixels in size, the actual
image is scaled larger and pixelation happens. Do you understand this?

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

Micha Riser <mri### [at] gmxnet> wrote:
> An alternative is to render the picture 4x bigger with normal AA 
> settings and scale it down in a graphic program.

  How that's different from using antialiasing method 1 with threshold 0
and four samples per pixel?
  (Method 2 will give you a better result anyways.)

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

Okay,

Well I got it figured out. Thanks for the help...the recursive method works
great for this type of model. Now all I need is talent!

Thanks again,
Mark

Post a reply to this message

Warp wrote:

> Micha Riser <mri### [at] gmxnet> wrote:
>> An alternative is to render the picture 4x bigger with normal AA
>> settings and scale it down in a graphic program.
> 
>   How that's different from using antialiasing method 1 with threshold 0
> and four samples per pixel?
>   (Method 2 will give you a better result anyways.)

With method 1 and threshold 0 you have a fixed number of samples per final 
pixel. But when you scale it 4x bigger and use a threshold >0 you have a 
dynamic number of samples per final pixel with a minimum of 4.

- Micha

-- 
http://objects.povworld.org - the POV-Ray Objects Collection

Post a reply to this message