In the image you'll see a glassy blob object with a gray background plane. I
tried sky_sphere and background {}, with similar results, if the AA
threshold is 0.1 or lower. AA values typed onto the two examples. Once the
AA value is high enough the blocky artifact will disappear. Apparently gamma
related somehow.

Curiously, I hadn't noticed such a thing happen before. At first I thought
it was my LCD screen, working at no more than 18-bit color, or maybe the
video driver. Anyone else want to try confirming this as a possible problem?

Bob H.

A simpler SDL for testing is:

global_settings { assumed_gamma 1 } // no correction, no artifact?

camera { location -3*z }

light_source { -100*z, 1 rotate <45,45,0> }

background { 0.5 }

sphere { 0, 1 pigment { rgb 1 } }

/* be sure to use AA of 0.05 or thereabouts to see the affect */

Post a reply to this message

Attachments:
Download 'bad_aa.jpg' (14 KB)

Preview of image 'bad_aa.jpg'

In article <40f578b0$1@news.povray.org> , "Hughes, B." 
<bob### [at] charternet> wrote:

> In the image you'll see a glassy blob object with a gray background plane. I
> tried sky_sphere and background {}, with similar results, if the AA
> threshold is 0.1 or lower. AA values typed onto the two examples. Once the
> AA value is high enough the blocky artifact will disappear. Apparently gamma
> related somehow.

If you post an image to show artifacts, you should really not use a lossy
image format.  Please post this again as PNG, with JPEG it is impossible to
tell what you are talking about and what are just JPEG artifacts.

If possible, please mark the problem in the image as well.

    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

"Hughes, B." <bob### [at] charternet> wrote in message
news:40f578b0$1@news.povray.org...
> In the image you'll see a glassy blob object with a gray background plane.
I
> tried sky_sphere and background {}, with similar results, if the AA
> threshold is 0.1 or lower. AA values typed onto the two examples. Once the
> AA value is high enough the blocky artifact will disappear. Apparently
gamma
> related somehow.
>
> Curiously, I hadn't noticed such a thing happen before. At first I thought
> it was my LCD screen, working at no more than 18-bit color, or maybe the
> video driver. Anyone else want to try confirming this as a possible
problem?
>
> Bob H.

i just noticed the effect monday i think. though i didn't notice that
assumed_gamma was part to of it. also, i've only noticed that it happens
with +AM1 and not +AM2. there seemed to be a definate range of +A0.0x
settings that it occured in, and the size of the artifacts changed based on
the value for +A. i can't remember exactly off the top of my head, but it
was something like +A0.3 to +A0.0000897 or something crazy.

i was also using a sky_sphere. this was on the official 3.6 linux binary
version of pov.

-ross

Post a reply to this message

"Thorsten Froehlich" <tho### [at] trfde> wrote in message
news:40f57c01@news.povray.org...
> In article <40f578b0$1@news.povray.org> , "Hughes, B."
> <bob### [at] charternet> wrote:
>
> > In the image you'll see a glassy blob object with a gray background
plane. I
> > tried sky_sphere and background {}, with similar results, if the AA
> > threshold is 0.1 or lower. AA values typed onto the two examples. Once
the
> > AA value is high enough the blocky artifact will disappear. Apparently
gamma
> > related somehow.
>
> If you post an image to show artifacts, you should really not use a lossy
> image format.  Please post this again as PNG, with JPEG it is impossible
to
> tell what you are talking about and what are just JPEG artifacts.
>
> If possible, please mark the problem in the image as well.
>
>     Thorsten

having experienced this (or very similar) anomoly i can tell you that the
light grey area on the right and lower right is the artifact. my image also
had a lighter area to the right of an object which went away with different
+A value and or by setting +AM2

hope this helps a little.

Post a reply to this message

I've seen this myself, and also determined that it's related to
anti-aliasing. Especially the way that the problem extends to the right and
down indicates that it's related to AA.

However, that same fact makes me wonder: do these artifacts actually appear
on the outputted image, or only on the display screen?

Have you posted screenshots of the display window or the actual output?

 - Slime
 [ http://www.slimeland.com/ ]

Post a reply to this message

"Slime" <fak### [at] emailaddress> wrote in message
news:40f5dfc1@news.povray.org...
>8---8<
> Have you posted screenshots of the display window or the actual output?

Rendered image files, which were pasted together into a single picture.

Here's a png of the simpler example, marked in red showing where the
blockiness appears always toward bottom right directions. AA options were
+a0.05 +am1 +j1.0 used. Contrast enhanced to make it show up better. And
yes, Ross is right, AA method 2 seems to be okay.
The original rendered file attached, also. All these images began as PNG, I
hadn't tried other formats until just now and see that it doesn't change the
artifact.

Bob H.

Post a reply to this message

Attachments:
Download 'aa-gamma_artifact.png' (21 KB) Download 'aa_gamma-original.png' (8 KB)

Preview of image 'aa-gamma_artifact.png'

Preview of image 'aa_gamma-original.png'

Hughes, B. wrote:

> "Slime" <fak### [at] emailaddress> wrote in message
> news:40f5dfc1@news.povray.org...
> 
>>8---8<
>>Have you posted screenshots of the display window or the actual output?
> 
> 
> Rendered image files, which were pasted together into a single picture.
> 
> Here's a png of the simpler example, marked in red showing where the
> blockiness appears always toward bottom right directions. AA options were
> +a0.05 +am1 +j1.0 used. Contrast enhanced to make it show up better. And
> yes, Ross is right, AA method 2 seems to be okay.
> The original rendered file attached, also. All these images began as PNG, I
> hadn't tried other formats until just now and see that it doesn't change the
> artifact.
> 
> Bob H.

Confirmed. It doesn't appear in 3.5, but does in 3.6, In fact.. when I 
changed assumed gamma to 2.2, I could still see vestiges of the artifact.

I have to wonder if it has something to do with the change to do the 
clipping after antialiasing ...

-- 
~Mike

Post a reply to this message

"Hughes, B." <bob### [at] charternet> schreef in bericht
news:40f5fee5@news.povray.org...
| "Slime" <fak### [at] emailaddress> wrote in message
| news:40f5dfc1@news.povray.org...
| >8---8<
| > Have you posted screenshots of the display window or the actual output?
|
| Rendered image files, which were pasted together into a single picture.
|
| Here's a png of the simpler example, marked in red showing where the
| blockiness appears always toward bottom right directions. AA options were
| +a0.05 +am1 +j1.0 used. Contrast enhanced to make it show up better. And
| yes, Ross is right, AA method 2 seems to be okay.
| The original rendered file attached, also. All these images began as PNG, I
| hadn't tried other formats until just now and see that it doesn't change the
| artifact.
|
| Bob H.


I'm using 3.6. I see no artifacts when I render it. -AA, +AA0.01, +AA0.05,
+AA0.3, +AA1, +AA10. All look okay, with no artefacts. I also tried the
settings you suggested (+A0.05 +AM1 +J1.0) no problemo :-\

-- 
Maurice

Post a reply to this message

"SomeOne" <ask### [at] yahoocouknospam> wrote in message
news:40f6f4fd@news.povray.org...
>
> "Hughes, B." <bob### [at] charternet> schreef in bericht
> news:40f5fee5@news.povray.org...
> | "Slime" <fak### [at] emailaddress> wrote in message
> | news:40f5dfc1@news.povray.org...
> | >8---8<
> | > Have you posted screenshots of the display window or the actual
output?
> |
> | Rendered image files, which were pasted together into a single picture.
> |
> | Here's a png of the simpler example, marked in red showing where the
> | blockiness appears always toward bottom right directions. AA options
were
> | +a0.05 +am1 +j1.0 used. Contrast enhanced to make it show up better. And
> | yes, Ross is right, AA method 2 seems to be okay.
> | The original rendered file attached, also. All these images began as
PNG, I
> | hadn't tried other formats until just now and see that it doesn't change
the
> | artifact.
> |
> | Bob H.
>
>
> I'm using 3.6. I see no artifacts when I render it. -AA, +AA0.01, +AA0.05,
> +AA0.3, +AA1, +AA10. All look okay, with no artefacts. I also tried the
> settings you suggested (+A0.05 +AM1 +J1.0) no problemo :-\
>
>

using what scene? were there a sky_sphere{...} or background{...} blocks?

Post a reply to this message

I used the scene provided in the original posting with no modifications.

"Ross" <rli### [at] everestkcnet> schreef in bericht
news:40f7041f$1@news.povray.org...
|
| "SomeOne" <ask### [at] yahoocouknospam> wrote in message
| news:40f6f4fd@news.povray.org...
| >
| > "Hughes, B." <bob### [at] charternet> schreef in bericht
| > news:40f5fee5@news.povray.org...
| > | "Slime" <fak### [at] emailaddress> wrote in message
| > | news:40f5dfc1@news.povray.org...
| > | >8---8<
| > | > Have you posted screenshots of the display window or the actual
| output?
| > |
| > | Rendered image files, which were pasted together into a single picture.
| > |
| > | Here's a png of the simpler example, marked in red showing where the
| > | blockiness appears always toward bottom right directions. AA options
| were
| > | +a0.05 +am1 +j1.0 used. Contrast enhanced to make it show up better. And
| > | yes, Ross is right, AA method 2 seems to be okay.
| > | The original rendered file attached, also. All these images began as
| PNG, I
| > | hadn't tried other formats until just now and see that it doesn't change
| the
| > | artifact.
| > |
| > | Bob H.
| >
| >
| > I'm using 3.6. I see no artifacts when I render it. -AA, +AA0.01, +AA0.05,
| > +AA0.3, +AA1, +AA10. All look okay, with no artefacts. I also tried the
| > settings you suggested (+A0.05 +AM1 +J1.0) no problemo :-\
| >
| >
|
| using what scene? were there a sky_sphere{...} or background{...} blocks?
|
|

Post a reply to this message