Sharing our experience as optical engineers specializing in #augmentedreality, me and my partners in JOYA Team want to create a common language, a database that can be shared and used by anyone who wishes to learn and understand the specifics of augmented &virtual reality systems – our optical terms library. If there is a term you want to learn about - leave a comment and we promise to give our interpretation of this term.
The next term is Uniformity
In Augmented / Virtual / Mixed Reality Systems, an image source is projected using an optical system to infinity (or a finite distance) to be viewed by the user’s eye, producing a virtual image. This image quality within the defined System Pupil (more about this term will be detailed in a separate post on System Pupil) can be expressed in different parameters, Uniformity is one of the parameters that we classify as a “Non-Imaging quality parameter”. When talking about Uniformity, it usually refers to uniformity through the Fielf of View, so when the user is looking at the projected image, we refer to the uniformity in different areas of the image. We can also talk about uniformity through the system pupil, we will do so in a separate post.
The Uniformity is a very generic term, we can talk about different types of Uniformity:
Large / Small area Uniformity
Brightness Non-Uniformity Illustration (Large area)
Color Non-Uniformity Illustration
All of these are important aspects of image quality and user experience. It is very important to define these requirements based on user’s needs and perform the non-imaging design and simulations in order to evaluate these parameters in the ealy design stages. One of the things we specialize in is the non-imaging design and simulations using the tools that are required for this.
Traditional optical design software tools (such as Zemax, CodeV, OSLO etc.) are made to design the optical systems, emphasizing explicitly on the image quality parameters that we classify as “Imaging quality parameters”, such as MTF / CTF; Distortion; AID; Field Curvature etc. These tools have very limited simulation abilities of the non-imaging aspects and parameters of optical systems, the main such parameter is Uniformity. This means that, traditionally, the first time the Uniformity is evaluated is when the prototypes are assembled and tested. If then the results are not satisfactory and the requirements are not met, little can be done to improve the design. Many products just go to the market with the non-imaging performances on the “best effort” basis, just updating the requirements to fit the masured results, compromising the user experience.
The recently made available optical design software tools for non-imaging and illumination design, such as LightTools (from Synopsys) provide the ability to model, analyze and optimize the image Uniformity parameters in any system at the early design stage. This tool is also used for evaluating other non-imaging parameters and artifcts, such as Ghost & Stray Light (more on this in the detailed post about Ghost & Stray Light).
Uniformity is specified in [%], Color Uniformity is unitless. It’s common to see the following requirements (if any are specified at all):
Uniformity: < ±20%
Color Uniformity: ΔR (u’, v’) ≤0.02
There are many different formulae for Uniformity or Non-Uniformity, so it’s useful to specify the parameter calculation formula together with the requirements values.
We find very suitable the following formulae:
Max – Maximal Brightness across the FOV;
Min – Minimal Brightness across the FOV;
Avg – Average Brightness across the FOV
Specifying the testing points for the Max/Min/Avg calculation has a large impact on the results
u’1, v’1 – color coordinates at any specific FOV point (point1)
u’2, v’2 – color coordinates at any other FOV point (point2)
The Uniformity requirements are defined based on the experience assuring the requested image quality. Another good source for these requirements can be found in the VESA (Video Electronics Standards Association) IDMS - Information Display Measurements Standard.
Our addition to the Uniformity specification:
Uniformity shall be defined for a given measurement pupil size, which correlates to the user’s eye pupil size and depends on outside lighting conditions (more on this will be detailed in a separate post dedicated to System Pupil).
Uniformity testing points locations shall be defined in oder to cover the whole FOV.
It is very important to define the Large area and Small area Uniformity requirements separately.
The Brightness and Color Uniformity shall be defined for each system prime color, as well as for the “white” color.
The Uniformity is the main optical system non-imaging quality measure and the Uniformity Requirements depend on the system’s use and the image content. Here are several different cases that impact the system Uniformity requirements in order to create an optimal design and user experience:
In all AR/ VR/ MR system types, it’s very important to consider the human vision characteristics in order to produce the requirements that will ultimately create good user experience. Since the human eye is more sensitive to contrast, and hence to the small area non-uniformity, small area non-uniformity requirements shall be tighter than the large area non-uniformity.
In case of an augmented reality system, when the projected image is combined with bright direct scene view, the scene illumination causes the projected image colors to appear washed out and the color space is significantly reduced. Each prime color brightness of the projected image must be high enough to maintain discernable color, this also drives the brightness uniformity requirements. For the same reason, the prime colors coordinates have to be maintained uniform through the FOV, so the colors don’t loose theis purity across the image. Some of the systems use local or global outside scene dimming in order to enhance both the image contrast and color.
In case of a mixed reality system, when the projected synthetic image is combined with a direct viewing camera / night vision image, the two images’ Brightness and Color Uniformity working point can be artificially adjusted by playin gwith the 2 images settings and their.
The image quality perception of symbolic or text image is different from that of a full video image. Thus, in case of a virtual reality system, when a full video image is projected, the large area image uniformity is important, so the requirements have to be tighter. The color space, on the other hand, is undisturbed by the outside scene, so the color uniformity requirements can be loosened.
Our definition of Uniformity (example):
Large Area Uniformity
Brightness Uniformity (for each prime color): ≤ ±20%
Background Uniformity (“black”): ≤ ±50%
Color Unifomity (for each prime color): ΔR (u’, v’) ≤ 0.02
Small Area Uniformity
Brightness Uniformity (for each prime color): ≤ ±10%
Background Uniformity (“black”): ≤ ±20%
Color Unifomity (for each prime color): ΔR (u’, v’) ≤ 0.01