Learn about balancing render quality and render time.
- [Instructor] Understanding the Arnold Renderer settings is essential to an efficient workflow. You can dial in exactly the level of quality you need for a particular situation. Render samples and ray trace depth are adjustable for each contributing render component, such as diffuse or specular. Proper application of these settings can optimize render times, by directing computing resources toward render components that contribute most significantly to the final image. In other words we don't want to devote computing resources to render components that aren't showing up much in the final render.
So, let's open up the render settings dialogue and take a look. Render setup. Go to the Arnold Renderer tab. And at the top is the rollout labeled Sampling and Ray Depth. We've got a bunch of parameters down here, mainly we have two columns, Samples and Ray Depth. Samples is the visual fidelity. Higher values mean better quality and less grain to the rendering. Ray depth is the number of times a ray can bounce in the scene.
And for the individual components, such as Diffuse, Specular, Transmission, et cetera, there is an appropriate ray depth parameter. If we increase the number of diffuse rays, for example, then we'll get more bounce light in the scene. Let's take a look at this without any adjustment, we'll just do a render as-is. So just click on the Render button. The rendering with default sampling and ray depth values looks a bit dark, and that's because the number of diffuse bounces is only one by default.
That's the first thing I'm going to increase. I'll bring the number of diffuse bounces up to five, and when I render this, it'll be much brighter because we're going to get a lot more bounce light. Likewise with the specular rays, we need more of those because I've got some metallic and glass surfaces in the scene. Increase the number of specular rays to three. The transmission rays is set to eight by default, and as we'll see later in the chapter on materials, we will need at least seven rays for the vase and the water here.
And we've got a default of eight, so we're fine, and we shouldn't get any black areas in the rendering. We can bring the transmission rays down to seven because that's all we need, but it won't really make any difference in this case. At the bottom however, we have the Depth Limits and the Ray Limit Total will need to be increased, because we currently have seven plus three plus five, or 15 rays. Let's bring that up to 20, just for safety. So now we'll be able to actually render all of those rays.
On the left we have the number of samples, and again that's the visual fidelity. And we could dial in individual values for these particular components. We can also set the overall quality here, with the Camera Anti-Aliasing Samples field. The default is three, which looks very grainy. We're going to bring this up to five and that should be a medium-quality rendering. Additionally, because we've got so much grain here in the bounce light, we can increase the number of Diffuse samples.
Bring that up to a value of four. And now we've got pretty good settings for our production render. There will be a little bit of grain, because these are not super-high values, those are pretty conservative values for a production render. Be careful increasing the number of samples, because the calculations will go up very quickly as you make small adjustments to the number of samples. On the right here, you can see columns that list the minimum and maximum number of samples, and there's a specific calculation that yields these results.
You can see that the Camera Anti-Aliasing here is five, and over here it says 25. That tells me that this value is being squared. Five times five is 25. Below that there is a diffuse sample value, and we increase that in order to remove grain in the shadows and the bounce light. And its minimum number of samples is reading out as 400, and that's because this value is being squared as well. So, the calculation for the number of diffuse samples is the camera sample squared, multiplied by the component samples squared, for the diffuse component.
And that's going to be 25 times 16, which is 400. Here's that formula. The minimum number of samples per component, is the camera sample squared, multiplied by the number of samples squared for that component. And the minimum number of samples shows the number of samples for just one ray. The maximum number, over here, shows the number of samples you might have at the maximum ray depth.
And we can see the ray depth here. And the formula for calculating the maximum number of samples, is to take the minimum number of samples and add to that. And you're adding the camera squared, multiplied by the ray depth value minus one. And that's how the min and max number of samples are calculated. Let's do a rendering of this at full quality. It's going to take quite a lot longer. We'll go ahead and execute a production render.
The production render is vastly improved in quality, compared to the default parameters. Especially since we've increased the ray depth on the diffuse component. The entire shot is now brighter. That's how to optimize the Arnold Renderer settings, specifically the sampling and ray depth for a production rendering, and it wraps up our chapter on Arnold concepts.
- Arnold rendering concepts
- Arnold lights such as quad, spot, and distant
- Modifying Arnold object properties
- Filtering light with the gobo filter modifier
- Image-based lighting with Skydome
- Daylight simulation with Physical Sky
- Arnold Standard Surface material parameters
- Diffuse, opacity, and bump mapping
- Rendering refractions with Transmission
- Building an Arnold shading network
- Test rendering with utility map
- Mesh subdivision and displacement at render time
- Atmospheric perspective with scene environment fog
- Rendering a spherical environment with VR Camera