Octane Materials take forever to preview in viewport

[Hurricane046]

Hello fellow Octane enthusiasts!

We've been dealing with strange problem at work - whenever we open a scene in C4D it takes up to a minute or two to cache all the materials in the viewport (depending on scene size).
We can of course save it without material previews (N~Q shortcut) but we'd like to keep the previews on at all times if possible. The problem is the minute or two lags the program.
Whenever this "caching" process is going after scene opening it makes entire C4D stutter and you really have to wait until it's done before you can start doing your thing.



Any ideas how to make this instant, as it is with the native C4D mats? Or at least speed it up somehow?

[aoktar]

It's a background process to calculate viewport textures for the materials and will not block you completely. But yes will cause some small lags, there's not a way to avoid it unless disable the OPENGL previews. I suppose you are not using high resolution for textures, I'd say to keep them at default otherwise.

[Hurricane046]

Thanks for the quick answer! I was hoping there is some way to keep them cached (they calculate once and save to cache so they don't need to be calculated on each startup), but if it works as you said it's not that huge of a problem. Thanks again and render on!

[ShivaMist]

It would be nice to find a way for the process to make less of an impact in big scenes, either having a cache system or an adjustement to let C4D be more responsive while rendering the materials.
When we have to switch heavy scenes to get objet/materials into another scene that where it will take its time especially with PBR textures.

[aoktar]

It would be nice to find a way for the process to make less of an impact in big scenes, either having a cache system or an adjustement to let C4D be more responsive while rendering the materials.
When we have to switch heavy scenes to get objet/materials into another scene that where it will take its time especially with PBR textures.

Yes would be. I've spend a lot of effort for materials area for this and more in past years.

[KeeWe]

Experienced the same over the last copuple of years. It's probably because the textures got bigger, but for some projects it's a pretty big time consumer. Ahmet, whats the bottleneck here? CPU is almost at idle (2-3% load), the program is stored on an nvme, the preview size is set to default. Is it the fairly old 1080/2080 ti GPU?

Getting a perfoamce boost in this area would be a huge deal.

Btw, maybe I'm wrong but I think opening a scene in standalone loads a lot quicker.

[aoktar]

Experienced the same over the last copuple of years. It's probably because the textures got bigger, but for some projects it's a pretty big time consumer. Ahmet, whats the bottleneck here? CPU is almost at idle (2-3% load), the program is stored on an nvme, the preview size is set to default. Is it the fairly old 1080/2080 ti GPU?

Getting a perfoamce boost in this area would be a huge deal.

Btw, maybe I'm wrong but I think opening a scene in standalone loads a lot quicker.

Hi KeeWe,
Standalone doesn't make any calculation for previews, etc... Don't compare apple and oranges.
As I wrote that I've spent much time for providing a stable combination for materials area for previews/etc... But there are not yet a quick/stable way to provide a performance leap. It's a dangerous part, don't want to touch anything unless having a large time and a stable build. Otherwise we can destroy much things while trying to do this.

[KeeWe]

You are right, I guess I meant the loading times when hitting the render button. I usually don't use standalone, I just tested it with a project a couple of months back and noticed the speed improvement when actually sending the scene to render. My bad.

And you are absolutely right... before breaking something, it's better having it a bit slower.

[aoktar]

You are right, I guess I meant the loading times when hitting the render button. I usually don't use standalone, I just tested it with a project a couple of months back and noticed the speed improvement when actually sending the scene to render. My bad.

And you are absolutely right... before breaking something, it's better having it a bit slower.

Yup sending processes are not same also. We're making a lot of pre-processing to prepare an Octane scene before sending it to GPU side. This makes the differences what you see.

[aoktar]

I have an old method to eliminate the lags. But it was disabled because of some problems. Reach me by PM if you wish to test it and tell me your license/C4D versions. I can give a test build.