The responsiveness of a display is a critical factor in user experience, particularly in interactive applications like gaming and virtual reality. Display latency, the delay between an input and the corresponding visual output on the screen, can significantly impact the sense of immersion and control. Interestingly, the time it takes to send a packet to a remote host is often half the time reported by ping, which measures a round trip time, highlighting the nuances of latency measurement in different contexts.
One way to measure display latency involves using a high-speed camera and a simple program that toggles the display color in response to user input. This method allows for a frame-by-frame analysis of the delay. In a typical setup, this might involve a program polling a game controller in a spin loop, clearing the screen to a different color, and swapping buffers whenever a button is pressed. By video recording both the game controller and the screen at 240 frames per second (fps), one can precisely count the frames between the button press and the visual change on the display.
This testing methodology, while straightforward, reveals significant differences in latency across various display technologies. For instance, when testing with an old CRT display at a 170 Hz vertical refresh rate, and under optimal conditions (vsync disabled, minimizing driver overhead), a color change is often observed within two 240 Hz frames after the input. This suggests a minimal latency, possibly around 8ms, associated with the USB HID processing of the game controller input. Further investigation is needed to precisely quantify input path latency.
Desktop LCD monitors often exhibit considerably higher latency, sometimes taking 10 or more 240 Hz frames to display a change. In contrast, the Sony HMZ-T1 head-mounted display, when tested, averaged around 18 frames, translating to over 70 milliseconds of total latency. This particular test was conducted in a multi-monitor setup, which can introduce additional latency due to driver overhead, potentially accounting for a couple of frames.
Latency is not solely a matter of software or processing delays; some latency is inherent to the display technology itself. LCD panels, for example, require time for liquid crystals to change orientation, with response times ranging from 4 to 20 milliseconds depending on the specific LCD technology. Single-chip LCoS (Liquid Crystal on Silicon) displays inherently buffer at least one video frame due to their conversion process from packed pixels to sequential color planes. Similarly, laser raster displays necessitate buffering to convert raster return patterns into back-and-forth scanning patterns. Frame-sequential or top-bottom split stereo 3D displays also face limitations in mid-frame updates.
OLED (Organic Light Emitting Diode) displays stand out as potentially offering the lowest latency among non-CRT technologies. The eMagin Z800 OLED display, for example, demonstrates latency comparable to a 60 Hz CRT, outperforming other non-CRT displays tested. This inherent speed of OLED technology makes it attractive for applications demanding minimal latency.
The relatively poor latency performance observed in some displays, like the Sony HMZ, often stems from software implementation choices rather than fundamental hardware limitations. Features commonly found in TVs, such as motion interpolation, necessitate frame buffering – often at least one frame and sometimes more. Other features like on-screen menus, format conversions, and content protection mechanisms can also introduce buffering delays. While some of these could be implemented in a streaming manner, the simpler approach of buffering between each subsystem can lead to a cumulative latency of half a dozen frames or more in certain systems.
This situation is not immutable. Display latency issues are largely addressable through improved software engineering and a focus on minimizing unnecessary buffering. It is crucial for display manufacturers to prioritize and optimize for low latency, particularly as display technology becomes increasingly integral to interactive and real-time experiences. Continued pressure and advocacy for low-latency displays can drive improvements across the industry, benefiting users in various applications, from gaming to professional visualization.
