Introduction
This document describes what is the Vignetting effect, and how to correct it in Aximmetry.
By correcting the Vignetting effect, you can achieve better quality keying in the case of Tracked Camera and Virtual Camera setups.
Vignette correction can also help you achieve better quality color correction when creating LUTs in the case of LED wall setups.
What is a Vignette?
The effect called Vignette can cause a “darkening” effect towards the edges of the image of the camera. In the picture below, you can see what a white surface can look like in your camera.
A vignette is caused by the characteristics of the camera and lens. Because of the vignette, during keying, the individual green pixels can have different shades in the input from your camera, even if your green background is evenly lit and it is one homogeneous surface.
The Vignette Correction could solve this problem by removing the darkening towards the borders of your camera's picture.
How to Apply Vignette Correction in Aximmetry?
For Vignette Correction to work, you need to capture footage with Aximmetry of the corresponding Input.
The process is different in the case of Tracked cameras and Virtual cameras:
-
Tracked Cameras
To begin in tracked camera compounds, locate the VIGNETTE panel next to your INPUT panel, and turn on the Capture parameter:
After that, text will appear on the video output, guiding you through the whole vignette correction process:
The capture process requires an ArUco marker with ID 70 from the 6x6 standard dictionary at 180 mm Marker size.
You can generate it at: https://chev.me/arucogen/
Print the ArUco marker at a size of 18 cm x 18 cm.
-
Virtual Cameras
To begin in virtual camera compounds, locate the INPUT panel, and use the Capture Vignette parameter:
This will capture a vignette image based on the current zoom and lens characteristics of your camera.
Capture Guidelines
You will also need a homogeneous (uniform color) surface that can fill the entire camera frame during the capture:
If you don't have a big enough homogeneous white light source, then you can try using uniform bright colored cardboard (the cardboard should have an even surface). The cardboard should be evenly lit with diffused white lighting to avoid the appearance of any color gradient.
To get the best results, your camera's picture should be close to overexposure. But any part of the picture being overexposed will significantly reduce the quality of the Vignette Compensation. To determine if the camera is overexposed or not, we suggest turning on the zebra ( https://en.wikipedia.org/wiki/Zebra_patterning ) in your camera. When the zebra is turned on, you should increase the exposure until the zebra appears, and then turn the exposure just a little bit down until all of the zebra disappears. Also, while the zebra is visible, it can give a good idea if your surface is homogeneously colored and evenly lit, as the zebra should only appear in the middle of the picture, where there is no vignette. If the zebra pattern is more numerous in one direction, then your surface is probably not evenly lit.
The Difference that Vignette Correction Makes
In this session, you can see a direct comparison of the keying with and without Vignette correction. The pictures were made with the same camera, lens, and lighting setup; the only difference is the Vignette correction.
Preview during keying without using Vignette Correction:
Preview during keying using Vignette Correction:
Keyed result without using Vignette Correction:
Keyed result using Vignette Correction:
As you can see, after applying the Vignette correction, the Keying is much clearer around the corners of the green screen; therefore, there is no need to increase the ‘Low Cut’ value any further (as it would be, when not using Vignette correction).
