The brightness step at don and doff needs dark-adaptation time

What it is

The visor has one luminance; the room has another. At don, a brighter visor is a flash of glare; a darker visor makes the rods that were working in the room start over. Doff reverses the step. Dark adaptation from cones to rods can take ten to thirty minutes to the floor; practically, “I can see the step in the room” still wants tens of seconds to a few minutes. The luminance step is an instantaneous jump, not fatigue that crept in during the session, and not tears boiled off by fewer blinks.

Stairs, doorways, driving, all happen in the minute or two after doff, while adaptation is unfinished.

Why it happens

Light bleaches photopigment. A sudden dark: remaining cone signal is not enough, rods wait on pigment regeneration, threshold falls slowly. A sudden bright: dark-adapted rods saturate, glare wipes detail, until cones take over and the pupil stops down. Headset spectra often peak in a band (OLED spikes, LCD backlight) that is not the room’s colour temperature, so a chromatic adaptation sits on the luminance step and object colour and contrast look wrong for a while.

Video see-through parks camera exposure in between: at don the “room” is already the camera’s dynamic range, and only at doff does the real room appear, so the step can be larger than in optical see-through. Optical see-through always leaks some real light, a smaller step, but killing the backlight or switching mode still jumps.

The curve is steepest in the first tens of seconds. A product that treats doff as instant return of environmental vision sends people into real obstacles in their blindest tens of seconds.

Studying it

Don and doff between a controlled visor luminance and a controlled room luminance. Measure detection threshold (a just-visible step, a sign) as a function of time, plus glare discomfort. A pupillometer can log the stop-down.

Independent variables: direction and size of the luminance step, spectral difference, how long the visor was worn (pre-adaptation), age. Dependent variables: threshold-versus-time curve, time to “I can see the room”, hesitation or missed footing on the first steps after doff.

Age is a strong variable: regeneration is slower past fifty. A lab of twenty-year-olds will overestimate “you can walk immediately after doff”.

Where it stops holding

When visor and room luminance and spectra are close, the step can be small enough not to wait for. A dark VR cinema doffed into a bright office is the large step; the reverse — bright office into a dark visor — is the unseen menu at don. Night-vision and safety patrols that already live in the dark may see better after doff than inside; the complaint then is glare in the bright room. Recovery times from one visor luminance in the lab do not carry onto a product with ten user brightness steps.

Applying it

• Do not boot into a full-white frame. Ramp from dark or mid-grey to content luminance so pupil and pigment have a few seconds.
• Do not schedule stairs or a street crossing at the instant of doff. Leave a visor-off perch that can still be felt (a seat, a rail), or make “doff” dim the visor first and only then allow take-off.
• Video see-through returning to real vision needs the same ramp; do not slam camera gain to full or to zero.
• How to check: in the intended environment, doff from content luminance and time until a pre-placed single step is named stably. If typical users need tens of seconds, the product flow must keep that time; a line of copy saying “mind your surroundings” does not.

Related

• Same group: N1.15.1 Blink rate drops during focused viewing and the ocular surface dries · N1.15.3 Staring at one depth for long accelerates accommodative fatigue · N1.15.4 Visual fatigue shows as a performance drop before subjective discomfort
• Nearby: N1.08 Headset Mass and Prolonged Wear · N5.08 Video See-through and Optical See-through
• Search terms: dark adaptation · light adaptation · luminance step