|
| Purpose of Latency Meter |
The effectiveness of virtual environment systems depends critically on the end-to-end delay between the user's motion and the update of the display. When the user moves, the graphics system must update the images on the display to reflect the proper projection of the virtual world on their field of vision. Significant delay in this update is
perceived as swimming of the virtual world; objects in the virtual world appear to follow the user's motions.Delays arise from a variety of sources:
Several of these times are subject to variation from subtle changes in the system. For example, users might change parameters of the tracking system in an effort to reduce jitter and not realize that such changes increase delay. Or programmers might make seemingly innocuous changes in their programs that effect rendering time (e.g. a change to the OpenGL state).
- tracker measurement time;
- transmission of the measurement to the graphics system;
- rendering time;
- synchronization delays at several points in the system; and
- time required to scan out the image to the display.
Publication:
Miller D., Bishop G., Latency meter: a device for easily monitoring VE delay, in Proceedings of SPIE Vol. #4660 Stereoscopic Displays and Virtual Reality Systems IX, San Jose, CA, January 2002.More Details:The conference was Photonics West 2002 > Electronic Imaging > The engineering Reality of Virtual Reality (4660B)
How it works Results Background People
| How it works [top] |
Our latency meter works as shown in figure 1 by observing the user's motion and the display's response using high-speed optical sensors. When the user rocks back and forth, the display exhibits a similar but delayed rocking of objects in the user's field of vision. We process the signals from the optical sensors to extract the times of very slow image change corresponding to the times when the user is nearly stopped (just before reversing direction). By correlating a sequence of these turn-around points in the two signals we can accurately estimate the end-to-end system delay.
![]()
Figure 1: System overview with interactive system and latency meter. In this case the user (actor) is in a virtual tavern environment.We have constructed simple sensors based on 1D CCD arrays (such as found in optical scanners) and have acquired data in real-time from a working VE system. Figure 2 shows the sensors and fixture. We have processed the data offline and have validated the delay estimates with independent measurements made using Mine's method [Mine 1993]. We have developed an online version of the algorithm that will continuously acquire data and display estimated latency. Our offline method has several tuning parameters whose effects must be better understood. We are also investigating a confidence measure that will allow simultaneous display of the latency and an estimate of its accuracy.
![]()
Figure 2: 1D CCD arrays. Two sensors are used with cylindrical lenses to record vertical and horizontal data.
| Results [top] |
We are using the latency meter to measure the end-to-end delay of several UNC interactive projects. The interactive projects involve tracking the user and displaying the corresponding view. The different systems listed below use a variety of trackers and displays. Typical results are 50 to 150ms of delay depending on the settings of the tracking system. When possible the latency meter results are verified with Mine's end-to-end latency measure, developed previously at UNC. Knowing system latency will help developers evaluate system improvements.UNC interactive projects measured:
- EVE Walk Through
- Lamp Shader a.k.a virtual spray paint
- Office of the Future
- NanoManipulator
- Ultrasound Augmented Reality
| Background [top] |
Several groups have measured end-to-end latency of their virtual environment systems [Azuma 1994, Jacoby 1996, Liang 1991, Mine 1993]. In our experience these measurements have involved complicated and expensive equipment that is not dedicated to the task. In our laboratory it generally requires over a day to collect the necessary equipment, modify the software, and make the measurements. Needless to say, the measurements are not made often.
| People [top] |
The latency meter is being developed in the Tracker group by:
- Gary Bishop - lead investigator
- Greg Welch - investigator
- Dorian Miller - implementor
- MSL engineers
Note on logo:
The logo is of the comic cowboy Lucky Luke, whose slogan is "he can draw his pistols faster than his shadow". Of course it is not possible to draw faster than your shadow; however, this would be the case if a user were tracked by an interactive system that generates the shadow. |