AI MASKING TO THE EXTREME: LIGHT DEPTH IN LUMINAR NEO 1.25

Gone are the days when photographers have to use manually brushed masks or dodge and burn techniques to sculpt lighting into their photos! Advancements in machine learning and AI have allowed multiple programs to introduce new tools to simplify your workflow and make editing faster. As an example, let’s take a brief look at Luminar Neo’s latest AI tool: Light Depth.

UNDERSTANDING DEPTH

Before we dive into how to use the new Light Depth tool, we first want to touch on depth within an image, otherwise known as Z-space.

Take a simple line chart like the one illustrated here.

When you move left and right on that chart, that’s the X-axis, and when you move up and down on that chart, that’s the Y-axis.

Z-space, or movement along the Z-axis, is actually moving into and out of the 2D graph on your chart.

Why is this important? To put it simply: this is how depth is calculated in an otherwise 2D image. Instead of seeing a flat photo of a lake in front of beautiful mountains, Luminar Neo’s new Light Depth tool (and others like it) are using AI to calculate the actual objects at multiple slices of depth within the Z-axis of your image. In turn, it produces a mask that you can move along that Z-axis to isolate only your lake, or only your mountains. And it does it all automatically!

BUT HOW?

Machine learning and AI have come a long way, like we mentioned earlier. Luminar Neo is using advanced programming to not only recognize objects in your image based on their silhouette, but also the depth and space in between multiple objects in your photo. It can also lean into EXIF data like focal length and aperture to understand the spatial relationship between objects in the foreground and objects in the background.

LET’S GET PRACTICAL

Below is an example with a car in a pit lane garage inside of Luminar Neo. To an older application, this is just a 2D image, and we’d have to manually mask the car to isolate it from the background. But with Luminar Neo’s Light Depth tool, it’s as easy as selecting the Light Depth tool and entering an initial Amount. In the examples below, we can see Amounts from 17 to 100. It’s important to not be too heavy handed when selecting an Amount so as not to make the mask jump out at the viewer.

From there, you can move the circular controller in the middle of the Light Depth preview to show you, in real time, how the scene is being observed by AI and how the light can wrap around the foreground car and affect only the background, if you’d prefer.

In this way, you can interactively shape the light around the car and adjust the lighting Amount to fit your tastes. You can even widen or narrow the scope of the lighting change, and also change the warmness or coolness of that light.

If you tick down Advanced Settings in the Light Depth tool, you can find other options to change the color temperature of the near or far planes in the Z-axis, and even brighten or darken them individually.

All of that is done in seconds with no manual masking at all!

SO WHAT’S NEXT?

AI is doing amazing things already to be able to recognize and isolate shapes like cars, flowers, animals and people in images. In the future, it’d be amazing to see these Light Depth tools being able to recognize that a car is in the image, and not only affect the lighting of the object but also the specular highlighting, or perhaps even changing the color of the paint but maintaining or even boosting the pearlescent factory coat within that paint.

With the rate of AI development and machine learning progress, updates and new features to our favorite photo and video tools are mere months away, not years. The sky is the limit!