Linear workflow - RGB color issue with Lightwave plugin Tweet

[smount1]

This is my setup.

LW2019-OT2019 color test.rar

(9.74 KiB) Downloaded 129 times

BTW, I'm still learning the basics of Octane and I'm not a tech guru guy, so this may be a painfully easy solution that I've simply missed, but any help is greatly appreciated.

I made a plain, mid grey color. I made a default sphere, RGB 255,128,0. I set the render target to a basically default set up, essentially a daylight shot. I experimented with various gamma + response settings: gamma 1 + response sRGB, gamma 2.2 + response Linear, gamma 1 + response Linear (yuk), etc. Nothing seemed to work out right, the color always blows up or changes significantly. Shouldn't you be able to change from a yellow highlight to more of a white highlight color? I know that sounds very amateurish but Octane is famous for having something for nearly everything, under the hood, with practically no post work. I was hoping there was a "white highlight" type button or solution. lol

I know that tone mapping is complex and tricky in Octane, and that I'm looking at things in a Lightwave color picker way instead of an Octane linear space way, but with help, I hope that will change soon.

Ultimately, I just need to be able to match RGB values SOMEHOW, preferably in the Octane plugin, and not Photoshop.

Is there a workaround node or setup that allows a RGB 0-255 color picker values to be converted, read and adjusted as a linear space 0-1 values in the Lightwave plugin. I have a variety of architectural samples, vinyl siding pieces and what not, that I need to be able to match without blowing up their color.

I've been reading various posts on Linear Workflow in Octane and trying to understand how it works, like the following link, but most were rather perplexing to me and didn't help that much. I was hoping there might be a Lamens terms description.

https://render.otoy.com/forum/viewtopic.php?f=21&t=33214[/color][/color]

Another side problem was ARTIFACTS showing up on a simple default 24/12 sphere, with the smoothing options ON in both Lightwave and the material node, why? So much for lower polygon objects? Will I have to sub-patch or subdivide everything now?

Thanyou in advance

[promity]

A little aside, but maybe this information will also be useful to you

https://inlifethrill.com/octane-render- ... Qxt8hmYm0A

[smount1]

That's a great tutorial, thanks for the link. Bookmarked.

It certainly taught me something very useful about handling colors in Octane. The Gaussian Spectrum node gave me much better quality and a more controlled result.

2nd problem -

Regarding my other problem of artifacts on a smoothed surface.

I duplicated the original sphere and subdivided it once. As you can see, there are still visible artifacts, very small artifacts, but they're still there. I tried adding a side fill light, it didn't make a difference, the artifacts remained. Why does this happen when both smoothing options are turned ON and the geometry is fairly dense?

the updated scene

Octane2019_LW2019_color test.rar

(76.59 KiB) Downloaded 136 times

[scooternva]

The documentation on how color spaces work in LightWave+Octane and how to work with a linear workflow are terrible/non-existent (I know this because I'm running into the same problem right now but just haven't had time to delve into it yet). But one thing I can help you with is that issue you're having with the triangular polys because I just ran into it myself:

The steps I've listed below assume that you've set up the standard tabs and shortcuts in LightWave Layout.

• Click the Render Target tab.
• Select Open Editor to open the render target node editor.
• If you don't already have a render kernel attached to the Kernel node in the Octane RenderTarget object, select a render kernel to use and hook its output node to the Kernel node (see screenshot below). My example uses the Direct Lighting kernel.
• Double-click on the kernel to bring up the dialog box where you can edit the settings.
• Look for the Ray Epsilon setting. That's almost certainly what's causing your rendering issue. If your ball is really large, you'll need to increase the size of Ray Epsilon to compensate; otherwise, you'll start getting weird shadowing effects (that's why you're seeing edges of triangles there in your image). Look downthread at the question I posted a few days ago on this subject for more info. In my case, I had a 1 kilometer asteroid doing what your ball is doing. To fix my problem, I upped Ray Epsilon from 1 meter to 10 meters.

OctaneRender is great, but you will quickly find that its documentation leaves a LOT to be desired (that's the G-rated version of what I think of their help manuals). I've actually had some luck with the Cinema 4D documentation created by Polat Yarisci (who goes by username "AOKTAR" here in the OTOY forums); even though it's for the Cinema 4D plugin, a lot of his procedures and advice also apply to the LightWave plugin for OctaneRender.

http://www.aoktar.com/octane/OCTANE%20H ... ANUAL.html

[smount1]

Excellent, thankyou, that was it! two for two

I never would have thought to look there. I've read a few blurbs on Ray Epsilon a while ago and was like "uh... ok", then kept on reading.

It would have been very useful to see a couple of side by side renders of what Ray Epsilon actually does. A picture speaks a thousand words.

I to go through the Cinema 4D manual a fair amount in order to possibly get for information on something. Thats what I also liked about Octane. It doesn't matter which program you use, Octane is controlled the same way.

[smount1]

I was continuing to play with those two spheres and wanted to see what happens when polygons intersect.

ANOTHER artifact appeared. Doh!

You will really need to be a super clean air tight modeler if you use Octane. Not a huge issue but an interesting discovery and one to consider when you're building a scene.

[scooternva]

Yup. You will also discover that sometimes it's forgiving of n-gons and sometimes it will shrug its shoulders, say "I'm OUTTA here", and will refuse to render polygons that have more than four vertices. This will manifest itself in your scenes as matte-black patches on your model where the problem child polys are located. After I coverted a model created in LightWave 9.6 to OctaneRender I spent a LOT of time going through the model fixing stuff like this. Perfectly flat polygons with 0.0% flatness but that had lots of points (example: the flat circle capping a cylinder) were loads of fun. Such polys wouldn't be a problem in LightWave, but in OctaneRender they would sometimes render and sometimes not.

Another fun little "quirk": At one point I accidentally set a few light-emitting diffuse surfaces to not cast shadows. This wasn't immediately apparent as an issue, except when all of the sudden areas of my model that should have been in near total shadow seemed to "glow". It was those damn light emitters... they were casting a directionless glow in every direction on my model! That was SUPER fun to figure out. Aggravating doesn't being to describe what that was like to hunt down and correct.

But man, once you've figured out the quirks and learn to model cleanly (and create your materials "the Octane way"), the results and speed speak for themselves!

[smount1]

It would certainly be nice if Octane eventually supported n-gons. Perhaps in time it will. Fingers crossed.

Fortunately I've been quite fastidious with making models out of quads, rarely even a triangle. A habit I developed from my learning subpatch modeling days.

Whenever I start getting aggravated with something not working out in Octane, I just take a peak at the gallery and remind myself "one day that will be my work in there"

[juanjgon]

Hi,

The Octane for LightWave plugin already supports n-gons. You have even some options related to this feature in the Octane Object Properties custom object plugin, but in any case, the n-gons support is always enabled by default. It is true that we have found some rare polygons that can't be tessellated by the plugin, but usually, this feature works fine.

Thanks,
-Juanjo

[smount1]

Excellent!

Thanks for the info Juanjgon