Hello people...

I've switched mainly to beta34 now, and rendered a few test scenes for a
Xmas card. For the first time, I notice horizontal line(s) which don't
seem related to the scene.

1) They only seem to appear in lath objects,
2) There seems to be some randomness in the intensity of the problem,
3) Independently of the size, the artifact is always 1 pixel high.
4) The effect is not is a spot where a bezier node is defined.
5) Sturm is enabled for both glasses and contents.

Here is an image where the artifact is _very_ visible:

http://jcoppens.com/misc/ok2fault.png (1.1MB)

If interested, I can upload the scene and background pic too.

John

Post a reply to this message

John Coppens <joh### [at] johncoppenscom> wrote:
> Hello people...
>
> I've switched mainly to beta34 now, and rendered a few test scenes for a
> Xmas card. For the first time, I notice horizontal line(s) which don't
> seem related to the scene.
>
> 1) They only seem to appear in lath objects,
> 2) There seems to be some randomness in the intensity of the problem,
> 3) Independently of the size, the artifact is always 1 pixel high.
> 4) The effect is not is a spot where a bezier node is defined.
> 5) Sturm is enabled for both glasses and contents.
>
> Here is an image where the artifact is _very_ visible:
>
> http://jcoppens.com/misc/ok2fault.png (1.1MB)
>
> If interested, I can upload the scene and background pic too.
>
> John

Hi John,
Can't view your image @ http://jcoppens.com/misc/ok2fault.png

Forbidden

You don't have permission to access /misc/ok2fault.png on this server.

Additionally, a 404 Not Found error was encountered while trying to use an
ErrorDocument to handle the request.

Cheers,
Peter

Post a reply to this message

> John Coppens<joh### [at] johncoppenscom>  wrote:
>> Hello people...
>>
>> I've switched mainly to beta34 now, and rendered a few test scenes for a
>> Xmas card. For the first time, I notice horizontal line(s) which don't
>> seem related to the scene.
>>
>> 1) They only seem to appear in lath objects,
>> 2) There seems to be some randomness in the intensity of the problem,
>> 3) Independently of the size, the artifact is always 1 pixel high.
>> 4) The effect is not is a spot where a bezier node is defined.
>> 5) Sturm is enabled for both glasses and contents.
>>
>> Here is an image where the artifact is _very_ visible:
>>
>> http://jcoppens.com/misc/ok2fault.png (1.1MB)
>>
>> If interested, I can upload the scene and background pic too.
>>
>> John
>
> Hi John,
> Can't view your image @ http://jcoppens.com/misc/ok2fault.png
>
> Forbidden
>
> You don't have permission to access /misc/ok2fault.png on this server.
>
> Additionally, a 404 Not Found error was encountered while trying to use an
> ErrorDocument to handle the request.
>
> Cheers,
> Peter
>
>
I can access that image without problem here.
Using Mozilla 3.5.6
I can see that the line appens in a straight part of the objects
There is a second, mich fainter, line slightly under the very obvious one.
In the right glass, there are 3 curved lines that looks like they can be 
other artefacts, or reflection of the original artefact. One bright 
flanked by 2 faint ones. Also visible in the reflection of that glass on 
the second glass.


Alain

Post a reply to this message

John Coppens <joh### [at] johncoppenscom> wrote:
> Hello people...

> I've switched mainly to beta34 now, and rendered a few test scenes for a
> Xmas card. For the first time, I notice horizontal line(s) which don't
> seem related to the scene.

> 1) They only seem to appear in lath objects,

  Lathes have shown artifacts (which look similar to coincident surface
artifacts) at the "seams" probably for as long as they have existed. I don't
know if it's a bug or a limitation.

  Either move the camera up or down a tiny bit, or if that doesn't help,
use a triangle mesh instead of a lathe (it will render faster too). There's
a unitility for that in the object collection.

-- 
                                                          - Warp

Post a reply to this message

On 2 Jan 2010 13:47:35 -0500
Warp <war### [at] tagpovrayorg> wrote:

> > 1) They only seem to appear in lath objects,
> 
>   Lathes have shown artifacts (which look similar to coincident surface
> artifacts) at the "seams" probably for as long as they have existed. I
> don't know if it's a bug or a limitation.

Hi Warp.

'Seams'? Do you mean at one of the nodes defined for the spline? The
artifacts do not coincide with the nodes. The nearest node is at the
widest part of the cup - quite far away.

I've now noticed that in a 800x600 image, the problem is on line 299, and,
in a 1200x900, the problem is on line 449 - exactly the (vertical) center
of the image.

>   Either move the camera up or down a tiny bit, or if that doesn't help,
> use a triangle mesh instead of a lathe (it will render faster too).
> There's a unitility for that in the object collection.

Thanks for the hint. I have my glass definition as Bezier splines, and
the utility seems to accept linear, cuadratic and cubic/natural splines.
Is there some easy way to convert too?

Cheers,
John

Post a reply to this message

John Coppens wrote:
> I've now noticed that in a 800x600 image, the problem is on line 299, and,
> in a 1200x900, the problem is on line 449 - exactly the (vertical) center
> of the image.

Interesting. Try moving or rotating your camera by a very small amount.

Post a reply to this message

On Sun, 03 Jan 2010 21:18:18 -0300
Nicolas Alvarez <nic### [at] gmailcom> wrote:

> > I've now noticed that in a 800x600 image, the problem is on line 299,
> > and, in a 1200x900, the problem is on line 449 - exactly the
> > (vertical) center of the image.
> 
> Interesting. Try moving or rotating your camera by a very small amount.

Here are more tests. I eliminated other elements as much as possible. I
removed the fireworks background, and put in a simple 'Darkgreen' one.

Surprisingly, the artifact spills out of the lathe into the background!

http://jcoppens.com/misc/glass_g1.png and
http://jcoppens.com/misc/glass_rt1.png

In both cases, the camera is looking down ever so slightly:

camera {
   perspective
   location <0, 0.273, -18.5>
   sky <0, 1, 0>
   direction <0, 0, 1>
   right <1.3333, 0, 0>
   up <0, 1, 1e-2>
   look_at <0, 0, -2>
   angle 34   
   translate <15.75, 15.75, 0>
}

Eliminating the 0.273 in location (making the line-of-sight horizontal),
the artifact almost disappears:

http://jcoppens.com/misc/glass_rt0.png

As requested, a slightly rotated camera version (0.1 degrees):

http://jcoppens.com/misc/glass_g2.png

... interesting: the artifact disappears slowly at its first height, and
seems to be appearing a little below.

John

Post a reply to this message