Rename "yaw" and "pitch" in virtual fitting

[peterhollender]

"yaw" and "pitch" are more appropriate for rotations from an object's point of view. In our case, we're talking about orbiting a transducer around an approximately-spherical head.

I think what we want is either "latitude - longitude", or "azimuth - elevation". I kind of like "latitude-longitude" because it mirrors coordinates on a globe, and we can simply define the "nose" position as the "equator"/"prime meridian"

[ebrahimebrahim]

Much clearer! And the ears are the poles. The left ear is the north pole, I think.

[peterhollender]

Much clearer! And the ears are the poles. The left ear is the north pole, I think.

Eh... guess I should have said "longitude-latitude". For usability I'd pick the directions so that the top of the head is the north pole, the subject's right ear is (checks Google maps...) the Galapagos Islands, and the subject's left ear is the Indian ocean.

[ebrahimebrahim]

One reason to not put the poles at the top of the head is that a search grid that is regularly spaced in terms of spherical coords will have closely packed points at the pole. Maybe that's the reason you had done it that way in the matlab software?

[peterhollender]

Hmm... so maybe it's more of a longitude/latitude specification of the center, and then relative "local longitude/latitude" around that point? Gets undefined at the north pole, but as long as you have a "up" direction, it's readily defined

[peterhollender]

Actually it wouldn't even be undefined at the north pole, since you could use the longitude setting to disambiguate the "north" direction