The Visual Difference Between Linear and Non Linear Workflow
By Neil Blevins
Created On: Sept 21st 2013

If you don't know a lot about Linear Workflow (LWF) yet, you may wish to read Wouter Wynen's Linear Workflow tutorial and my Linear Workflow In Vray For 3dsmax Cheetsheet before reading below. While these are max/vray specific tutorials, the theory is the same for all renderers and 3d apps.

This lesson comes out of a discussion we had on the vray forum, there's plenty of tutorials discussing how to use Linear Workflow, and why you should use Linear Workflow, but there isn't as much information out there on exactly what the visual difference is between a Linear and a Non Linear workflow image. Lets take a deeper look at exactly what is happening.

The following two images are made with a neutralized lighting rig. That means that if you provide a material that's 50% grey, the end result in your render is 50% grey. The Non Linear image has a input color of 128,128,128 and the Linear image has the same color but with a gamma of 2.2, in order to achieve the same final color in the two renders (this is discussed in more detail in my Linear Workflow In Vray For 3dsmax Cheetsheet). This way, we can compare apples to apples.

Each has a single area light from the upper right, no sky or fill light and no GI.


Feel free to save the images and flip between them to see the difference more clearly.

You'll see the main difference is that the Non Linear one has a much softer and gradual falloff to black, and the Linear Workflow image has a larger area of flat bright color and a sharper falloff to black.

Alfa Smyrna on the vray forum I think said it best: "I see LWF as adjusting the falloff curve for the transition from black to white. I perceive it as a falloff for greys so gamma=1 is a more steep transition but especially for interiors with small openings, LWF makes the greys look more similar to real life distribution of light."

My renders above support that idea, the two workflows have a very different falloff curve for the transition from black to white.

Here's another example, this time something more complicated. The linear workflow image has less contrast than the non linear workflow image...

In this example, I apply a levels in photoshop to the Linear Workflow image and get back an image that's a lot closer to the Non Linear image.

Here's a similar example, but this time with texture maps...

Similar results, although the Linear image run through the Levels Color Correction is not as close to the Non Linear render because the levels color correction is adding contrast to both the lighting of the scene as well as the textures.

Here's a more complex lighting situation, a skylight using an hdr image map with a number of GI bounces (the ground surface is 50% grey at the material level and in the final render)...


This is a great example of the downsides of not using Linear Workflow, the image on the left has plenty of contrast, but notice that the environment map and the texture map I'm using look like this in photoshop...

So the texture maps in the left render are way too dark. With Linear Workflow, the texture maps look a lot closer to the real colors, but the lighting is less contrasty.

BTW, here's me applying a Curves color correction in photoshop to the image...

It's not as contrasty as the non linear image, but has deeper blacks than the linear image.

Now, whether you like the more contrasty image or the less contrasty image is a matter of taste, it's an artistic choice. Mathematically, Linear Work Flow is better for a number of reasons, and it's easier to integrate cg elements into real photography using Linear Workflow. But artistically, you may be after something different. Afterall, realism may not be your goal. Even in live action films the Directors Of Photography will shoot on a film stock that specifically increases contrast, or will crush the blacks at the color grading step. If you like the final look of the linear workflow image, then great, you're all done. But if you want to get your contrast back, here's your options...
  1. Don't use linear workflow. You'll keep your contrast, but your texture maps will tend to look wrong and you'll have to do a lot of extra steps since almost all renderers and 3d apps now default to linear workflow. Also, you may get incorrect results with displacement maps, for example, if you export a displacement map out of mudbox to use inside 3dsmax, you want to set its gamma to 1, or else you'll get incorrect results, since you want the absolute numeric value of the map, not a visual value.
  2. Use Linear Workflow, get contrast by color correction. So take your LWF image and bring it into photoshop or a compositor like Nuke or Digital Fusion, and use color correction to increase the contrast of your image to your liking. Some common tools would be Levels, Curves, gamma, exposure, etc. Also, you may want to save your image out of the renderer as a floating point exr with a gamma of 1 since that'll provide you more data to do your color correction on, and then the last step is applying the 2.2 gamma in your composite.
  3. Use Linear Workflow, adjust your lights. In more complex lighting situations, you can get some of your contrast back by reducing the fill light in your scene and/or reducing the multiplier on your secondary light bounces (if using GI)
What your choose is up to you, but hopefully this lesson will give you a better understanding of the problem, which hopefully will lead to a solution that works better for you.

This site is ©2013 by Neil Blevins, All rights are reserved.
Back to