What is the latency of a Hall Effect keyboard? How is it composed?

People often have the misconception that an 8000Hz polling rate equates to a 0.125ms low latency. However, this is not the case. Polling rate refers to how often the computer queries the keyboard for key state changes.

For example, 8000Hz means the computer queries the keyboard every 0.125ms, and 1000Hz means the computer queries the keyboard every 1ms.

Here is an interactive example:

The slider represents the moment when the keyboard detects a key press. The dots on the progress bar represent the times when the computer queries the keyboard. The filled portion of the progress bar indicates the latency duration.

From this, we can see that polling rate and keyboard latency are not directly related, as polling rate is just one component of latency.

In conclusion, polling rate and latency are two different concepts and should not be simply equated. A higher polling rate does not necessarily mean lower latency.

Latency composition of Hall Effect keyboards

The following steps do not follow a strictly sequential order.

1. The time required for the sensor to respond to magnetic force changes, i.e., how long it takes to change the output voltage. Except for the cheapest sensors, this time far exceeds the requirement to achieve 0.125ms latency.
2. The time consumed by software and hardware filtering, which typically takes the longest time.
3. The time required for the controller to detect the latest voltage. This can vary greatly depending on the technical approach.
4. The time for multiple voltage sampling rounds required to achieve 0.01mm precision (even precision below 0.05mm requires this, though the number of rounds needed differs).
5. The time needed for the controller to calculate, such as determining the relationship between voltage and distance, whether a key should be pressed or released, or retrieving the specific key code.
6. The time until the next polling, which is described in the previous paragraph. The reduced latency of having 8000Hz is only apparent here.
7. The time required for the computer to process the data.

Due to the non-linear relationship between magnetic force and distance, the change in voltage at the top of the travel is minimal. Why?

Some Hall Effect keyboards take more extensive samples at the top of the travel to achieve 0.01mm precision, leading to latency differences between the top and bottom of the key travel. Additionally, noise from the keyboard's circuit design and components becomes significant under such small voltage changes. This necessitates more complex filtering algorithms, increasing computation time and latency.

Our Efforts

Our keyboards feature targeted circuit designs and high-quality, low-noise components to reduce sensor noise. At the same time, we use higher performance controllers combined with optimized algorithms to maximize processing speed. By achieving high precision across the full key travel, we have greatly reduced or even eliminated latency differences between the top and bottom of the key travel.

To some extent, EDGE 80 and 60 V2 provide users with an experience that approaches the performance limits.