[l1407]

This tends to happen with materials with layers in additive mode, but the specks problem has been improved for the next version coming soon. You can try turning off direct/indirect reflected caustics and see if that helps.

When can you to release the new debug version of the maxwell render

[Bubbaloo]

I am doing some renders for a medical device company and am having to clean the speckles in PS. If it is my sss in the vertebrae discs, anyone have any suggestions how to get rid of the speckles? I used the preset honey sss.

Thanks.

Aaron

Could it be that the meshes with SSS are not closed meshes? If they have holes or unwelded verts, then it may cause problems.

[Asmithey]

It is possible. The models are .step files exported from solid works. I will run the object doctor to see if it comes up with any missing faces.

Thanks,

Aaron

The objects all came out to be clean.

Aaron

[rusteberg]

I am doing some renders for a medical device company and am having to clean the speckles in PS. If it is my sss in the vertebrae discs, anyone have any suggestions how to get rid of the speckles? I used the preset honey sss.

Thanks.

Aaron

Could it be that the meshes with SSS are not closed meshes? If they have holes or unwelded verts, then it may cause problems.

I opened up an old scene with sss the other day created in 2.5 and re-rendered a preview of the material and it was consumed with speckles............ They converge eventually, but have a serious lag time.....

[Asmithey]

I had been just turning off direct and indirect caustic reflections. When uncheck all caustic options the speckles seem to stop appearing.

Aaron

[Hrvoje]

Is there some news regarding the update?
I also experienced this issue in very simple scenes and "regular" materials where after SL 20 or so bright dots start to appear...

Cheers

Hrvoje

[Half Life]

Is there some news regarding the update?
I also experienced this issue in very simple scenes and "regular" materials where after SL 20 or so bright dots start to appear...

Cheers

Hrvoje

This issue was discussed here (in regards to SSS and Dielectrics) -- they are late solving caustics:

http://www.maxwellrender.com/forum/view ... p?p=340525

Increasing the surface roughness will make the missing caustics solve faster, increasing the attenuation may stop them from appearing at all.

The easiest way I've seen to solve the issue is to reduce the contrast of the light source -- The sun (in the physical sky) is incredibly powerful and very small and that drives these caustics to solve late (alot of parallel rays apparently)... however if you increase the scattering properties of the Physical Sky you can reduce the contrast (by spreading the suns power over a larger area) and I've found that solves these caustics faster (or they simply don't appear at all).

The bottom line is it seems high-contrast lighting is the issue and Maxwell will only solve these caustics at very high SL/render times (as this is the nature of unbiased rendering in general).

I really feel as if this is a major issue... hopefully one that can be resolved in future updates to Maxwell, but this is common to all unbiased engines (or so I'm told).

Best,
Jason.

[Hrvoje]

Thanks for the reply.

I have read the topic and reaching such high SL is err...a bit problematic time and computing power wise, so I guess I could test it out without caustics to see if result will be acceptable.
Hopefully something will be done in next update to resolve or minimize this issue.

Cheers

Hrvoje

[mgroeteke]

I am having this as well. Indirect caustics off. No additive materials. Here is the difference between 17 sl and 18 sl.

The 17 SL render was done on one machine. The 18 SL render was a coop with two machines. No sun or sky. This was a studio lighting set with multi-light enabled. 4 emitter planes plus IBL.

Aaron

afaik it's more likely to be an issue appearing in cooperative render mode only... stay tuned for an update.

cheers

markus

[Half Life]

There are several issues being discussed here -- I'm not sure which of them you think is network related, but I can assure you the dielectric and SSS late-solving caustics "noise" issue does not come from any MXI merging issues as I have not used a network for any test renders.

It is a very reliable problem, if you know how to set up for it you can make it happen at will... but it is also a very well-known problem.

Best,
Jason.

[Hrvoje]

No network here, just a single machine rendering simple material (satin preset) with with two emitters in scene. After certain SL (around 20 or so) bright dots appear. Sure, they can be cleaned in post and I unfortunately can't afford to render for another 20-50 hours to see if this will clear up. File has been sent to Mihai for inspection

Cheers

Hrvoje

[Bubbaloo]

No network here, just a single machine rendering simple material (satin preset) with with two emitters in scene. After certain SL (around 20 or so) bright dots appear. Sure, they can be cleaned in post and I unfortunately can't afford to render for another 20-50 hours to see if this will clear up. File has been sent to Mihai for inspection

Cheers

Hrvoje

The Satin preset uses 2 additive layers (1 at 80 and 1 at 100), so I think it's a problem with excessive energy.

[Half Life]

You may find my RAL velvet line on the Maxwell Resources website to be a good solution as it does not use additive blending mode (since additive blending was a noise maker even when I was recording my first Maxwell tutorial video series).

http://resources.maxwellrender.com/scri ... 1_7590.zip

I made alot of those RAL lines as alternatives to the "wizard" materials since they were problematic in various ways.

Best,
Jason.

[limbus]

I encountered this bug as well in many scenes.

The 17 SL render was done on one machine. The 18 SL render was a coop with two machines. No sun or sky. This was a studio lighting set with multi-light enabled. 4 emitter planes plus IBL.

My testing shows that the merging process introduces the white dots. Single render with SL16: no dots. Merged Render with SL16: white dots. Your test seems to point in the same direction.
Hope this will be fixed soon.

Cheers, Florian