Sphinx.
 |
I've encountered a rather strange bug - the refraction component seems to modify the precision of the subtree somehow (or the cache of the last component, dunno). See the attached filter - here the refractions output is not connected to anything at all, yet just hooking up its height input will affect other branches of the height source output.
Open the filter and try to disconnect/connect the refraction height input and notice how the brightness/contrast and blend output changes.. 
Edit: Further investigation revealed that this happens no matter what you connect to the refraction input, down to the blur. So it seems to relate to the bitmap based blur source.. btw.. this bug is also present with the height input in the Surface Result. strange bug.ffxml |
|
| Posted: September 28, 2008 10:22 am | ||
|
Sphinx.
|
This image shows what happens - the settings are exactly the same, only difference is the refraction height connection (note that the contrast isn't 100, its a rounding error in the display):
 |
|
| Posted: September 29, 2008 3:51 am | ||
Crapadilla
 |
Weird indeed, especially since the Refraction isn't connected to the Result.
Could this be somehow related to the fact that Refraction samples its input tree with 3 times the amount of samples? --- Crapadilla says: "Damn you, stupid redundant feature requests!" ;) |
|
| Posted: September 29, 2008 5:12 am | ||
|
onyXMaster
Posts: 350 |
Indeed, the Refraction and Result (Surface mode) "Height" inputs forces double-precision, along with changing the default bitmap sampling algorithm on all bitmap-based components on the attached subtree.
I am not sure of the reason behind the observed effect though, since while the extreme contrast adjustments tend to stress the precision-dependent errors, it seems strange that the switch to a higher precision actually increases percieved error. In any case, thank you for pointing out and isolating the issue, we'll look into it. |
|
| Posted: September 29, 2008 5:39 am | ||
|
Vladimir Golovin
Administrator |
Very interesting. We're looking into this.
|
|
| Posted: September 29, 2008 5:48 am | ||
Join Our Community!
Filter Forge has a thriving, vibrant, knowledgeable user community. Feel free to join us and have fun!
33,711 Registered Users
+18 new in 30 days!
153,533 Posts
+38 new in 30 days!
15,348 Topics
+73 new in year!
Online Users Last minute:
14 unregistered users.
Recent Forum Posts:
- Manson Grudge Border by Jaret by Minko
5 hours ago - Kaleidoscopic Variations Vol. 2 by jhantares by jhantares
September 10, 2025 - Polygonal Images Set by Ramlyn by westdale
September 10, 2025 - FF Killing Resolution by SpaceRay
September 9, 2025 - Geometry Spinner by Rachel Duim by Rachel Duim
September 6, 2025 - Channel Glitch by Martin Naumann by SpaceRay
August 31, 2025 - Circuit Glitch by Martin Naumann by SpaceRay
August 31, 2025 - Groovy Rays by Ramlyn by Ramlyn
August 31, 2025 - Is Ultraforge dead? by dpankros
August 30, 2025 - Component Instances by AussieDecalf
August 27, 2025 - Hyperbolic Tiling Challenge by AussieDecalf
August 27, 2025 - Chaos Fields by SpaceRay
August 27, 2025
