| From: | wafflemiester <evamarie@**********.NET> |
|---|---|
| Subject: | Visual perception functionality (was Re: Thermographics and glass) |
| Date: | Fri, 10 Jul 1998 17:19:26 -0500 |
Warning- meandering technical info below.
>
> Nexx Many-Scars wrote:
> >
> > In a message dated 98-07-09 13:55:27 EDT, you write:
> >
> > > So, my point is that orks and dwarves should have real troubles using
> > > any kind of sight augmentation (except for cybereyes, IMHO, because they
> > > can switch the thermo vision option off).
> >
> > How do dwarves and trolls handle cyber-eye thermo? After all, its not going
> > to be the same as their natural thermo, but their brains would already be set
> > up to see things that way, unlike the brains of elves, orks, and breeders...
> >
> > Nexx
>
> The way i see it is that when you got your eyes replaced, visual input
> still has to go through the same neural pathway to get to the brain and
> be interpreted.
>
Not nesecarilly. If the eyes feed a simsense signal (as is implied by
R2's "simsense veritigo" flaw), the sense feeds need not match any
"normal" visual perception. Then it could look like ANYTHING, or even
bypass visual perception asa means of perception (you could "know" your
surroundings). This seems unlikely (low essence cost=minor simsense
interaction, I'd think), but gives some idea ofthe flexability involved.
> So, let's say that a special area of the brain treat the IR signal, and
> another the light one. In this case, thermo vision is considered as
> another sense, sharing the same sensors as normal sight.
>
That makes some sense. However, brain regions are also quite flexible,
and can "learn" new functions.
> We have two options:
> 1. Signal is discriminated at the eye level. There is two kinds of
> receptive cellules in the eye, each kind taking care of one kind of
> stimulus. So IR and light signals go by their own particular pathway to
> the brain to be interpreted in different places.
> Your cyber replacement will have to be wired to the right neural
> channels.
> 2. Signal is discriminated at the brain level. Our two kinds of cellules
> emit signals that share the same pathway, like phone calls on the cable,
> but are somehow sorted by the brain to go to their respective processing
> places.
> Your cyber replacement will have to give its output in the right format.
Niether of these even quite captures how a NATURAL eye functions,
because there is some "procesesing"in they eye, in that the retinal
light sensing nuerons (rod AND cone) do NOT connect directly to the
brain. Instead, they feed signals into intemediary nueral levels
located deaper in the retina that do comparative functions like edge and
motion detection. Those then link to the optic nerve, which goes to the
optical lobes. There is no "format", but there are definate rules of
processeing you could work with (or completely alter, I suppose).
This opens some intersting posibilites for cyber-eye design and retinal
modification. If I personally were augmenting low-light or thermo
vision, I would have the new sensors feed into the edge detection and
especially motion detecting intermediary nureons. This mimics natures
designs- almost all creatures see motion in the dark much better than
still objects, and can notice it without knowing WHAT they see. For
example, my pet snake reacts with a hunting posture sometimes when her
hot air blower goes on- she "sees" heat in motion (using a simlar system
as pit vipers, but more primative in the constrictors), and thinks
"food", even though she doesn't "see" anything else. Of course,
snakes
are stupid, but humans getfooled by optical illusions also- its an
artifact of nureal function, one that shows how information is conveyed
to the mind.
As an example, try some image processing experiments (I used paint shop
pro, but DON'T use a compressed image). Take a picture, and pushdown the
color (gamma), brightness, and contrast levels. It will look like
"nightime". Now, run an edge detection filter. It can improve things a
lot, but may look confusing, without the original image. Even better,
take two similarly treated frames from motion, and enhance the difrence
between the two. What you are seeing is the "extra" information that
lowlight or thermo systems could give you- and, in my games, that
ultrasound represents with colored lines on goggles or even better in a
displaylink.
Why did I say NOT to use a compressed image? Well, jpeg compression is
designed to eleminate some information that your retina won't process-
it reduces color variations, simplifies grey scales, and compresses
information about edges and shading. At high compression, this produces
noticable artifacts, but at low levels it just removes stuff most folks
won't notice. Mpeg compression does a similar thing with motion.
Thatmight alteryour results a lot. (COOL, huh? Jpeg is "simsense"
already, in that its computer processing methods based on / modeling
human perception.)
If designing cyber-vision augmentations, I'd do it this way so as to
give maximum benefit with minimal interfrence to "normal" vision. Maybe
I'm talking about alpha or betawear here, since this would require some
mapping of the patients normal visual nervous funtion and targeted
stimulous. Then again, so does a datajack.
> IMHO, option 2. would be harder to realize, as you are adding processing
> equipment to the cyber eyes (and it has the same price for everybody),
> as opposed to option 1. where all you have to do is to put the cybereyes
> outputs where they belong.
As I see it, option one would be just feading new info to the existing
retina- in effect, replacing (or modifying) all the mechical "camra"
parts ofthe eye, and putting a "screen" over the retian (probaly of
direct electrical stimulus). Option 2 is closer to my idea, and would
be more complex knowledgewise, but IMO simpler implant wise- and the
computer power needed to map nueral response would be easily availible.
>
> An alternative way to get around this is to say that visual information
> is treated indifferently whatever the race, that the same brain areas
> are involved. The visual input is differenciated directly at the
> reception level by different kind of cellules producing an unified
> signal that takes both ir and thermo stimuli into account and blend them
> together. This information is then interpreted by the normal visual
> brain areas. For all intents and purposes, it would be like seeing
> colors no other race can see.
Thats my impression, but in fact the brain area used matters little-
the brain can re-learn visual functions to an extent. As for the colors
seen, who knows what any color looks like, subjectively? I know my own
two eyes see slightly diffrent colors.
There is a LOT more info produced in the visual cortex than just what
"colors" and 2-d shapes are present. The idea of seeing an "edge" or
"motion" with no other associated image stimulous is normally bizarre,
but not impossible- optical illusions create the effect quite often. In
the case of cyber-eye sensoriums (ornatural thermo / lowlight), the
"invisible line" effect in the clasic illusionary triangle would
actually be useful information. Invisable mages spotted with thermo, or
shapes in the dark, might be noticed as as sharply defined "edges" and
"motions" yet still have no associated "colors", for example.
-Mongoose- who wouldn't mind having cybereyes if he could tweek the
image processing effects. Free hallucinations, cool! Just make a
backup before you mess around... HMM, maybe I'll do some animated
optical illusions on my new system. Heh.
