Re: Retinal Resolution

[T3D Jeremy Hinton <Jeremy.Hinton@xxxxxxxxxxxxx>]

T3D john bercovitz  <bercov@xxxxxxxxxxx> wrote:

>OK.....   8-)  If we do so well with the neural and vascular
>networks in the optical path, how well would we do it they
>weren't there?  I know, everyone hates a question like that.
>8-)

Aliasing. Fine stripes and sharp edges giving artefactual bands.
I think it looks rather like Moire fringes. It is possible to 
create stripe patterns on the retina without attenuation by the 
optics, using interference patterns from a laser, so the idea has 
been tested. (but not by me!)

We need a Stereo application to stay on topic..... these stripey 
patterns are periodic gratings, and may not be great stereo stimuli. 
If the same spatial frequency and orientation was presented to each 
eye, it would be possible to see them at almost any depth (wallpaper 
effect). If the spatial frequencies were a little different, the 
perception should be of an oblique plane.

There may be benefits to improved optical resolution with the same 
retinal sampling, but I am not aware of them. Of course if the optical 
quality was better, it may be worthwhile evolving a finer receptor mosaic.



Jeremy



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