Controllers provide a haptic feedback channel
Aliases: controller vibration · trigger haptics · handheld force cue
What it is
The virtual panel has already been hit; the highlight is still fading in; the motor in the palm has already ticked. A handheld controller supplies a controller haptics channel: the system can write “contact / threshold / empty” onto the skin so the eyes do not have to police every state change.
That tick is not the mechanical click at the end of key travel. The click follows the finger and happens whenever the key exists. The motor follows virtual events and stays silent if nothing happened. The same piece of plastic is both an input device and a small tactile display.
Why it happens
Virtual objects have no surface normal force. Collisions, snaps, and empty magazines would, in the real world, arrive through skin and muscle spindles. In XR those forces are absent, so vision is drafted as the evidence that “you touched it.” A short pulse takes the somatosensory path, on the order of ten milliseconds, and closes the loop before an animation finishes. The eyes can stay on the scene; confirmation does not need peripheral catching of a button flash.
Bandwidth is narrow. Linear resonators are good at brief ticks, poor at sustained shape or texture. What they can mark are events — contact onset, a crossed threshold, an error — not the material of a virtual surface. Mark every event and the channel becomes noise; people mute haptics and the path dies.
Studying it
Keep the task and compare haptic confirmation against visual-only confirmation within subjects. Align event dispatch to motor onset at millisecond resolution.
Independent variables: haptics on/off, pulse duration and amplitude, sync with the visual highlight, event class (hit / miss / continuous scrape). Dependent variables: missed confirmations, repeated mis-actions, the instant people judge “I hit it,” body-related items on a presence questionnaire, the rate at which people turn vibration off.
Watch ceiling effects. Large, high-contrast buttons already suffice visually; haptics will not show. Shrink the target, or delay the visual highlight by a few tens of milliseconds, and the channel’s contribution appears. Detection-threshold studies (two intensities discriminable?) speak to device capability, not to whether a cue should exist.
Where it stops holding
All-in-one headsets often ship weak motors in large shells; a tick that is crisp on a desktop controller becomes a thud. Gloves, sweat, and a loose grip cut transmission to the skin. Sustained low-frequency buzz numbs the hand and masks later short pulses. The channel assumes the device is being held; set it down or switch to bare hands and the source is gone. Lab actuators are often calibrated; consumer firmware has its own minimum playback length and clipping, so a paper waveform is not what arrives in the palm.
Applying it
- Fire short pulses on state flips only: select, snap, empty, error. Do not leave a resident buzz on hover, drag trails, or a background heartbeat.
- Give hit and miss different duration or amplitude so a person with eyes closed can tell success from failure.
- Align motor onset to the event, not to the first frame of the animation. The animation may lag; the tick on the skin may not.
- Ship a mute, then check that visual confirmation still stands when it is used — muting is common, not an edge case.
- How to check: have someone wear the headset with the outward view visible to you. Cover their in-headset highlights and judge from the hand’s pause whether they think they scored a hit. With haptics, that pause should land on the pulse. If they still wait for the highlight, the channel is not in the confirmation loop.