WIP sloooooooow render

I recently upgraded to the WIP in anticipation of a final release. After upgrade I opened some files that had been previously edited in Toucan era Rhino for Mac. A few modifications were made and the files prepped for render. I turned on skylight, otherwise made no changes to render settings (no lights, only simple materials). At first it said it would take an hour, so I redirected and worked on other things. After over four hours the render was apparently frozen at 92%. I wasn’t able to get a screenshot of Activity Monitor but memory was in the vicinity of 6.5 GB and CPU >400%. Rhino was unresponsive, but ultimately I was able to cancel the rendering.
I frequently render in Flamingo on the windows partition, however would like to be able to do basic rendering in native Mac. Can anyone point me to the problem?

2011 MBP

Model Name: MacBook Pro
Model Identifier: MacBookPro8,2
Processor Name: Intel Core i7
Processor Speed: 2.2 GHz
Number of Processors: 1
Total Number of Cores: 4
L2 Cache (per Core): 256 KB
L3 Cache: 6 MB
Memory: 8 GB
Boot ROM Version: MBP81.0047.B27
SMC Version (system): 1.69f4
Serial Number (system): C02G83BHDRJM
Hardware UUID: 7967B36F-CBBF-5A6C-A2ED-50587125EE20
Sudden Motion Sensor:
State: Enabled

escallators2.3dm (10.4 MB)

took about 1.5hrs on a 3.5ghz quad…

(click on it for actual size)

it’s 3000px wide… with a 5year old 2.2ghz laptop quad, i think you should expect longer render times… if you need to do a lot of renders like this, probably a good time to consider 6+ core macpro…


that aside, i’m not sure why it failed to complete for you… 400%cpu is good as it’s saying rhino is using all the cpu power available… i don’t know what the 6.5GB means though… possibly ran out of memory? is there any sort of message/alert that notifies of memory problems (question is in general-- not directly at you brooklynsculpt)


offtopic-- what’s brooklynSculpt?

Thanks, Jeff.
That still seems like a pretty long time for a printable render to me, 4 hours sounds more like a G4 not an i7. It seems to me the skylight is the main culprit as most renders in toucan with multiple materials and lights took under an hour, and most renders in Flamingo are quite good at 20-40 minutes. Anyone at McNeel care to chime in on this?
Out of curiosity I popped this file into the latest Wenatchee release and with no added lights it completed in under a minute, although as you might expect it looked terrible. I’ll give it a try in Flamingo later and report back.

In both Wenatchee and now WIP I’ve had serious problems with Rhino gobbling up memory and stalling. I upgraded to WIP after reading this post:

The file that I was working with at the time was almost completely curves (about 16k) and I was seeing very similar responses. Could this be related?

And, Jeff - I am a sculptor, and lived in BK for a time.

The skylight can add time for sure. A couple rectangular lights will render faster and also have soft shadows but these shadows will be based on the distance and size of the rectangular light from the model. You may also need to adjust the ambient light color in the render properties and tweak the rectangular lights shadow intensity to get a similar effect to the skylight. The ambient occlusion in Toucan was faster than the skylight true but only at the lower gathering setting in my experience. In any case, Toucan is not will be developed going forward (reasons explained in other posts if you’re interested). The skylight in Rhino Render supports HDR lighting though so this is pretty powerful and wasn’t possible in Toucan so that’s a plus. All and all, Rhino render is a very basic renderer and once plugins are supported you’ll have more options on the Mac for sure. In your model, and for the look you illustrate of soft uniform shadows, I’d use Vray or Brazil with irradiance mapping or auto-occlusion render cache respectively. Or use a GPU based engine like Octane and process the raytracing on an Nvidia card with a lot of Cuda cores. You can either use biased techniques for speed or hardware upgrades in my experience but both can make a world of difference.

Here’s an example of what I’m talking about using Brazil’s auto-occlusion render cache. 3000x1500 @72dpi in 11 minutes on an i7 quad core.

And this is a 4.1 minute render on an old Core2Quad Q6600 processor with Cinema4D R12
With a skylight and Global illumination and Ambient Occlusion turned on.

The Q6600 is less than half the speed of an i7 K2600,and 1/3 of an i7 K4770

Speed is the reason I really dislike Rhino render.