OctaneRender 1.0 review - page4

Comparing OctaneRender 1.0 to V-Ray 2.0, KeyShot 3.x and Arion 2.0.x

I would like to mention something before we go and compare the final images. OctaneRender is render engine and it’s impossible to compare it and his function for example to V-Ray where V-Ray is not only a render engine but a complete render solution. Moreover, all of these renderers are in a constant development even in the time I was writing this review. For example Arion is now presenting his new version of Arion 2.0.4 with highly optimized new sampling method and KeyShot on the other hand presented a while ago the new interactive KeyShotVR and is going to release a new version of KeyShot, 4.0. Still, we can compare the final result of render quality and render time because these factors should be optimized in all four renderers at maximum.
 

OctaneRender 1.0 vs V-Ray 2.0 vs Arion 1.6.x vs KeyShot 3.3 vs MentalRay

Highly quality renders and 10 to 50 times faster? This is really bold statement by OTOY. So let’s take a look if it’s really true and the GPU rendering technology is really faster than CPU. To be able to test the quality and render speed of OctaneRender 1.0  I decided to compare it to V-Ray 2.0, KeyShot 3.3, Arion 2.0.3 and MentalRay. The renders were pointed on these three main factors; render speed, render quality and the real time (RT) renderer. As the object for comparison I choose a Home Theater System which will be later also as a tutorial consisting from four different objects and materials;  metal material with texture based look, a plastic material with mate and reflective shader and some other textures symbolizing the logos and other stuff on the objects.

The testing PC had the following specifications:

  • CPU – i7 2600K 3,5 GHz
  • RAM - 16 GB
  • GPU – GTX560 TI 2GB DDR5
  • Windows 7 x64
  • 3ds Max 2012 x64

The overall scene was built from the “Home Teather” and a plane (a ramp) on which was in all four applications (scenes) were applied a neutral material consisting just from a diffuse colour which wasn’t taking or generating reflections. In all applications were used an hdri map for the environment and reflections. The render set-ups and comparing conditions are listed in the tables bellow.

MentalRay

  • Lightning: One Photometric light
  • Materials: Arch&Design materials
  • Environment: hdri map
  • Resolution: 1024x568
  • Mental Ray Settings:
    • Samples: 4-16
    • FG Precision: custom
    • Bounces: 2

KeyShot 3

  • Lightning: hdri map
  • Materials: Presets materials (metal, plastic, emitting)
  • Environment: hdri map
  • Resolution: 1024x568
  • KeyShot Render Settings:
    • Samples:                   64
    • Bounces:                     8
    • Antialiasing:                2
    • Shadow quality:          2

Arion 2.0.3

  • Lightning: skylight
  • Materials: RC Live Presets (metal, plastic, emitting)
  • Environment: hdri map
  • Resolution: 1024x568
  • Arion Render settings:
    • Bounces:                     6
    • Unbiased method

V-Ray 2.0

  • Lightning: V-Ray dome light, 3x V-Ray lights
  • Materials: V-Ray shaders (metal, plastic, emitting
  • Environment: hdri map
  • Resolution: 1024x568
  • V-Ray Render Settings:
    • DMC Sampler, custom
    • Primary & Secondary bounces
    • Indirect illumination, custom

OctaneRender 1.0 (3ds Max)

  • Lightning: mesh emitters, Octanelight
  • Materials: mix and glossy materials
  • Environment: hdri map
  • Resolution: 1024x568
  • OctaneRender settings:
    • Direct light method
    • 6400 samples

OctaneRender 1.01 (Stand-alone)

  • Lightning: mesh emitters, Octanelight
  • Materials: mix and glossy materials
  • Environment: hdri map
  • Resolution: 1024x568
  • OctaneRender settings:
    • Direct light method
    • 6400 samples

 

So, let’s take a look on the results. Next to the final render I did also two more to see how the quality and time will change with changing the settings lower or somewhere higher. The final renders which I used for comparison together with the rest of the outputs are in the gallery bellow. The “final” tag indicates that this picture was used for the comparison and the “low” or “high” tag was used to point out if the setting was higher than the final render or lower. The images are sorted from lowest render time to the highest where the render time is embedded also in to the image.

Conclusion

OctaneRender is for me a very pleasant surprise. The render power, quality and GPU speed is very impressive and opens new ways of rendering. Comparing the final results either from the plug-in or the stand-alone version we can say that the rendering method based on GPU was good step and can compete with other renderers based on the CPU. Moreover, the constant development on both versions ensures that the renderer will evolve and new features and performance improvements will be added soon. Looking above on all renders I can’t choose any winner, all are good, all renderers have their own advantages and different hardware needs. Choosing just one without optimizing every single scene and also the hardware is impossible. Still, I don’t see a fair way how to say the final conclusion about who is the winner of this comparison so I will let you to decide this.

Popular