I've got some weird gradients going on with bidir, note the part on the top of the image...
bidir:
pathOCL:
blender file:
bidir vs pathocl Shading
Forum rules
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
- FarbigeWelt
- Donor
- Posts: 1046
- Joined: Sun Jul 01, 2018 12:07 pm
- Location: Switzerland
- Contact:
Re: bidir vs pathocl Shading
I have no access to my PC at the moment and cannot look at the object. This is why I ask a few questions.
Is the object smooth or flat shaded?
Do you see a gradient in solid openGL preview in viewport too?
If yes you can avoid this. Extrude top surface in edit mode. Move extruded surface a short way upwards (z-axis) just so that you can distinct vertices in top corners. The openGL shading should be uniform now. The render will be too.
Light and Word designing Creator - www.farbigewelt.ch - aka quantenkristall || #luxcorerender
MacBook Air with M1
MacBook Air with M1
Re: bidir vs pathocl Shading
Thanks, I'm aware how shading works. It would be kinda weird if I didn't by now it would also look the same in OCL rendering as in bidir if that was the case...FarbigeWelt wrote: ↑Tue Aug 06, 2019 7:06 amI have no access to my PC at the moment and cannot look at the object. This is why I ask a few questions.
Is the object smooth or flat shaded?
Do you see a gradient in solid openGL preview in viewport too?
If yes you can avoid this. Extrude top surface in edit mode. Move extruded surface a short way upwards (z-axis) just so that you can distinct vertices in top corners. The openGL shading should be uniform now. The render will be too.
It's a simple flat plane, tried reseting normals, smoothing, flattening normals, reversing normals... and idk what else..
Re: bidir vs pathocl Shading
I'm aware of this problem but I have yet to investigate the source, it must be something in MIS weights.
Re: bidir vs pathocl Shading
Is there some ETA on this? I need caustics(POOL) for my next project that means only bidir+metro+LOT of rendertime...
For me that means either bidir+metro but there is this shading issue, or wait until lighttracing gets vertex merging, not sure which will come sooner and my only other option is fake caustics+cycles and I really don't want to go back to cycles now that 2.8 will have luxcore addon...
For me that means either bidir+metro but there is this shading issue, or wait until lighttracing gets vertex merging, not sure which will come sooner and my only other option is fake caustics+cycles and I really don't want to go back to cycles now that 2.8 will have luxcore addon...
- FarbigeWelt
- Donor
- Posts: 1046
- Joined: Sun Jul 01, 2018 12:07 pm
- Location: Switzerland
- Contact:
Re: bidir vs pathocl Shading
Actually I did not expect anything else from someone with experienced skills. But even close to perfection people forget sometimes the simplest things.lacilaci wrote: ↑Tue Aug 06, 2019 7:12 am
Thanks, I'm aware how shading works. It would be kinda weird if I didn't by now it would also look the same in OCL rendering as in bidir if that was the case...
It's a simple flat plane, tried reseting normals, smoothing, flattening normals, reversing normals... and idk what else..
However, until yesterday I thought to know how Blender deals with surfaces and shading. What was that of an error! I have to post my latest observations this evening maybe someone of the forum has the explanations. There must be some hidden parameters of a triangle‘s surface which are independent on vertices and surface normal and these parameters are responsible for different smooth shading tones.
Light and Word designing Creator - www.farbigewelt.ch - aka quantenkristall || #luxcorerender
MacBook Air with M1
MacBook Air with M1
Re: bidir vs pathocl Shading
I have a feeling that you might have ran into a problem of custom normals in blender, where these are often not exported to luxcore and cause a LOT of trouble especially when you use CAD models or create custom breaks of normals in blender... very annoying but it just might be a blender bug/problemFarbigeWelt wrote: ↑Tue Aug 06, 2019 9:48 amActually I did not expect anything else from someone with experienced skills. But even close to perfection people forget sometimes the simplest things.lacilaci wrote: ↑Tue Aug 06, 2019 7:12 am
Thanks, I'm aware how shading works. It would be kinda weird if I didn't by now it would also look the same in OCL rendering as in bidir if that was the case...
It's a simple flat plane, tried reseting normals, smoothing, flattening normals, reversing normals... and idk what else..
However, until yesterday I thought to know how Blender deals with surfaces and shading. What was that of an error! I have to post my latest observations this evening maybe someone of the forum has the explanations. There must be some hidden parameters of a triangle‘s surface which are independent on vertices and surface normal and these parameters are responsible for different smooth shading tones.