Visible disc

[raymoh]

I'm into astronomical art (as an amateur) and want to be as accurate as possible. I wanted to create a visible (star) disc lower than the sun's (0.5 degrees) but it seems thats not possible. If I choose a lower diameter Terragen renders always the 0.5 degrees of the suns disc, despite the slider goes to 0.0000...something. If the disc is lower than 0.005 degrees Terragen jumps suddenly to a disc near 0 degrees. That behavior seems to me a bug. It would be nice to fix that.

[cyphyr]

Works for me.
Are you using the slider or typing it in?

[raymoh]

It doesn't matter.
The diameter will be constant at about 0.4 degrees and jumps to nearly 0 at 0.005 degrees.
Background: I set the disc's diameter in relation to the image size. For example the image size is 1920x1080 and Terragen computes a FOV of about 60 degrees, so the sun's apparent diameter (0.5 degrees) will be about 16 pixel wide. This pixel-diameter is I work with. It seems more accurate to me. And in this context I can't reach my desired 0.2 degrees (or 7 pixels)...:-(

macOs Mojave 10.14.1 on an iMac Pro; Terragen 4.3.18

[WAS]

Are you vieweing through atmosphere? TG has a light focusing issue in atmosphere and clouds.

[raymoh]

I noticed that, but for my "tests" I used an untouched atmosphere and reduced all glow amount and glow power to a minimum. I also reduced the sunlight strength to obtain a clear defined sun disc.

[WAS]

Interesting. It'd be easier to see the issue with an example TGD in this case. You may also have better luck with a light source.

Have you tried something like this?

[Matt]

Try turning off "anti-aliasing bloom" on the render node's Filter tab. Turning this off might cause a jagged edge to the star due to the extremely bright values. Anti-aliasing bloom is designed to overcome this, but has the side effect of making the star appear larger. If you render a larger image, this side effect will be smaller relative to the image size.

[raymoh]

Hi Matt
Your input was great!
Without Anti-Aliasing-Bloom the sun disc's diameter are much more accurate, but it work well only with the Narrow Cubic Pixel Filter.
Mitchell-Netravali and Catmull-Rom are rendering a dark ring around the disc.

[attach=1]
Disc Test +Bloom/Catmull-Rom
[attach=2]
Disc Test -Bloom/Catmull-Rom
[attach=3]
Disc Test -Bloom/Narrow Cubic

[WAS]

The disc will always have a slightly rough appearance without high detail as well as it has no curve and is just a solid disc.

[raymoh]

Do you refer to the Pixel Filter(s)? (Narrow Cubic vs Catmull-Rom)

[WAS]

Do you refer to the Pixel Filter(s)? (Narrow Cubic vs Catmull-Rom)

Honestly I rarely use other AA methods due to artefacting.

[raymoh]

Honestly I rarely use other AA methods due to artefacting.

Sorry when I insist: Does that mean you are working mostly with the Narrow Cubic Filter?

[WAS]

Honestly I rarely use other AA methods due to artefacting.

Sorry when I insist: Does that mean you are working mostly with the Narrow Cubic Filter?

Yes. I rarely change it. Sometimes I'll use Catnum-Roll for close shots of surfaces with very fine low-level detail as it helps create definition between shapes instead of muddling them, but usually not with atmospherics or objects.

[Oshyan]

The black artifacts are inherent to those particular filters, yes (they are "negative lobes", far too much detail can be found here https://en.wikipedia.org/wiki/Ringing_artifacts and here: http://www.yafaray.org/community/forum/viewtopic.php?f=14&t=4744 ).

- Oshyan

[WAS]

The black artifacts are inherent to those particular filters, yes (they are "negative lobes", far too much detail can be found here https://en.wikipedia.org/wiki/Ringing_artifacts and here: http://www.yafaray.org/community/forum/viewtopic.php?f=14&t=4744 ).

- Oshyan

Never seen that second article. Thanks Oshyan.