Re: Convergence as a Depth Cue

[P3D Gabriel Jacob <jacob@xxxxxxxxxxxxxx>]

LeRoy Barco writes

>     IMO, convergence is exceptionally weak as a depth cue.

Yes this is true, as opposed to what is commonly thought. But I would
like to make some qualifications on the convergence cue. See below.

(some stuff deleted)
>     So I must queque(:=)) with those who think the self-contained image
>disparaties are the overwhelming contributors to the mental stereovision
>construct.

I am not sure if I'm interpretating correctly but disparities are a very
important contributor to reconstructing stereovision.

Okay now for some rambling thoughts. Warning, stuff below is due to
33-1/3% conjecture, 33-1/3% experimental analysis and 33-1/3% experience
(rightly or wrongly) ;-).

I think convergence plays a factor in apparent depth perception with
certain important qualifications. It's commonly thought that one
determines depth of objects by convergence but more precisely is due
to disparities. I feel convergence cues be used for rough approximations
in determining depth by the brain. This is not to say it's not an
important cue though! In the first 30 feet or so, I think what happens
is that the rough distance is determined by convergence (accomodation
is another factor) and then the +/- few (maybe 1) feet is due to disparity
determination.

As we know in slide mounting (and such) too much disparity results in
the brain refusing to fuse the images. This is where convergence comes
to the rescue, and the eyes reconverge to another point closer to the
place where the huge disparity existed. So to recap, I think convergence
and disparity depth determination are both important, especially in 
closer distances. Convergence being used to zoom into the depth in question
and disparity information to resolve the "fine" depth. Question arises
then, how can we judge rough distances with convergence when it is only
a rough estimate and disparities are only good for resolving a few feet
of depth (at close distances, at farther distances then it plays a bigger
factor, disregarding motion parallax) at each convergence point. This
is done by the other visual cues such as occulsion, size of object, etc.

This leads to another question of apparent size of images due to convergence.
Will leave that for another post.

Gabriel


------------------------------