Anchor distance must scale with body size, not a fixed length

What it is

Forty centimetres in front of the belly is a palm’s reach for someone 190 cm tall, and near end of arm for a child of 120 cm. Anchor distance must scale with body size, not a fixed length: how far a panel sits from waist, chest, or wrist should be a fraction of this arm or this torso, not centimetres hard-coded in world space.

Hard-coded centimetres calibrate for one stature. Everyone else receives not “the same near-body,” but a reach difficulty that has been scaled up or down.

Why it happens

“Thirty centimetres forward” in a body frame looks relative and is still an external length. What is actually relative is “half a forearm,” “0.3 of hip width.” Arm length, torso height, and shoulder breadth vary; the same absolute offset lands in different functional zones for different people — an adult’s comfort, a child’s maximum reach, or the reverse, in a child’s face and on an adult’s belly. Arrival is compiled in chain length, not in metres. An absolute offset forces a reach out of proportion with the person’s own body.

The headset’s default user is often the mid-stature adult in the calibration room. Offsets are tuned to that chest and shipped in everyone’s config. Tall users then find the panel on their body; short users have to lean out. If the tracking origin is the head and a fixed “90 cm down to the waist” is pushed from it, a child’s waist is pushed toward the knees and a tall user’s toward the chest — a fake waist worse than none.

Studying it

Recruit people with a wide span of arm length or height, place a panel with an absolute offset and with an offset as a fraction of arm length, and have them reach and look.

Independent variables: offset in centimetres versus as an arm-length fraction, stature (arm length / height), waist versus chest as the site. Dependent variables: whether shoulder–elbow angles land in comfort, one-shot arrival success, ratings of “too near / too far,” the performance gap between stature groups.

The gap between stature groups is the main result. Under absolute offsets it should grow with the stature span; under proportional offsets it should shrink. Measured only on university students, the two curves nearly coincide and the claim looks empty. Bring in children, or people sitting in adult chairs, and absolute-offset failure appears.

Where it stops holding

Shared objects in the room cannot be placed by one person’s body proportions — tall and short would then see two misaligned worlds. Proportion applies only to entries pinned on this person, serving this person. A user-adjustable absolute offset (pull the panel in or out) can partly replace automatic proportion, if the adjuster itself can be found and does not reset to the hard-coded centimetres on next launch. On extreme statures (very small children, very long arms) a simple linear fraction also warps, because joint proportions do not scale uniformly with height. Without leg or hip tracking, “by height” can only be guessed from head height; sitting is then estimated as a very short stander, and the proportion goes wrong.

Applying it

• Write forward and vertical offsets of riding entries as a fraction of arm length or seated height, not as “0.4 m forward” in a config.
• On first don, measure arm length or let people park the panel where the hand naturally rests, and store that distance as this person’s fraction — not as centimetres to be issued to the next user.
• An offset pushed down from the head to the waist must use this person’s torso height, not a universal 90 cm.
• How to check: take one clearly tall and one clearly short person on the same absolute offset. If one has it on the belly and the other is at a locked elbow, the absolute has already failed. After switching both to the same fraction of their own arm, both elbows should sit in a similar comfort band — otherwise proportion never actually entered the layout.

Related

• Same group: N2.07.1 The body supplies a location reference that does not need vision · N2.07.2 Body-anchored interfaces travel with the person · N2.07.3 Choice of reference frame decides what happens after a turn · N2.07.4 The head is the worst body site for an anchor · N2.07.5 A waist anchor is more stable than a shoulder anchor · N2.07.6 A wrist anchor brings the interface into view at the cost of raising the arm
• Nearby: N3.01 Locking Modes · N3.08 World-locked, Body-locked, and Head-locked
• Search terms: body-scaled anchor distance · anthropometric offset · arm-length proportion