[Prev][Next][Index][Thread]

Re: (meteorobs) Re: Dark Adaption

On Thu, 15 Feb 2001, Robert Lunsford wrote:

> I found your reply most interesting but it brought up the question; can
> the human eye really detect changes in angular velocity in every meteor?
> 
> I don't believe that it's possible in meteors of short length or those
> of short duration. To me it would only seem possible to see these
> changes in meteors lasting several seconds.

Changes in angular velocity, are probably unconsciously utilised in the
human visual system, but effects like this can probably be overriden.
I didn't write that changes could be detected in every meteor and it appears
that only a small proportion of people (and not necessarily on every
occasion) actually utilise this information.  As I said, in the cases where
people have reported an 3D orientation of a fireball to me, it has mostly
been correct.

When the observer is specifically concentrating on the 2D great-circle
path, the 3D orientation would probably not register.  I suspect that the
only time we become aware of velocity changes is when they are marked,
as in long duration meteors as Bob suggests.  The computational process
(referring to computers and I guess the neural system) operates on the
changes (more correctly the first differential of the angular velicity) and
when these changes are zero, the object is moving at right angles to the
observer.


Steve Harrison wrote
> >Even within a static viewpoint
> >there are many *cues* to depth in the visual field, from object overlap,
> >surface texture to haze obscuration.

>Ummm... let's take a really good example and go look at some NEAR pix of
>Eros' surface. If you have no idea whatsoever how large any of those
>features really are, then you cannot possibly begin to guess how far NEAR
>was from the surface when the pic was snapped. THAT'S about as close to a
>"static" 2D example as we're likely to come, I suspect.

There is a missunderstanding here about what I was talking about.  I was
referring to *cues* to *depth*, not absolute distance determination.  In
other words, there are various items of information that can be interpreted
(mostly correctly, butt not always so, hence "cue") as differences in
distance from the observer.  This adds information on the third dimension
that makes the image more than just 2D.  I'm not saying it makes it
truely 3D.

However with regard to a NEAR image of Eros, there is indeed information
that is suggestive of depth differences, from surface texture, obscuration
and shadows, to mention a few off the top of my head.  Add the dimension of
time and from a static observational perspective the rotation of Eros
would be correctly interpreted to give the 3D shape.

> >  But more importantly, vision is a
> >dynamic process; we view from a moving platform.  Thus dynamical parallax
> >information gives a powerful direct cue to the third dimension.  Animals
> >that do not have binocular vision use head movements to determine distance.
> >While it is true that this is neither as certain or as immediate as
> >binocular vision, it is incorrect to imply that single eye vision is
> >strictly 2D. Even within the retina, computations are performed, so what we
> >see, even with one eye from a static location, IS more a model than a
> >photograph.

>Actually... I rather suspect that what really happens is that the brain
>shifts the eye's focal point back and forth in an effort to gain an idea of
>distance.

Yes, accommodation is a more direct means of depth perception but is only
relevant for close objects.

> If the brain receives conflicting information in the form of a
>totally unfamiliar object whose dimensions are completely unknown from
>memory, it has no idea, except by changing the focal point, just how far
>away, or how large, the object is. A good example of this is when we look
>through a telescope at a known close-by object: by manually shifting the
>focus back and forth, we eventually gain an understanding of the distance.
>Then by comparing the conceived distance to the size of the object, we gain
>an understanding of the object's size. Our brain does the same with our
>eyes.

This cannot apply to distant objects where other cues to depth are necessary.

>Anybody remember waking up really, really drunk, or perhaps stunned
>by a fall or blow to the head? Remember how long it took you to "focus" on
>a familiar object? Until that "known" object, with known size and distance,
>actually produced the right "numbers" in your head, you felt totally out of
>it, right?

This is not relevant to the issue.  A malfunctioning system is unlikely to
correctly interpret any information.

> >To make this more on topic, when we look at the "starry dome" we are seeing
> >a 2D surface.  There are NO cues to depth to the human eye.  It is often
> >assumed that a meteor is also a 2D phenomenon because we cannot utilise
> >binocular vision over such huge distances.  However I contend that there
> >*are* cues to depth, and that these tend to be used correctly in the most
> >part.

>Yes, I agree.. but I propose that the PRIMARY cue, for such 2D objects for
>which no known distance or size information is previously available, is the
>brain's attempt to shift focus until the object makes sense.

This cannot work for distant objects as I stated above.

> >Again,
> >if the assumption is that an object doesn't change size, then the apparent
> >size on the retina is a cue to relative distance.

>Umm.. I propose the major cue is the brain's foreknowledge of where it has
>to shift the eye's focal point.

same again

> >  However the information
> >on angular velocity alone can be used to calculate orientation (the radiant) 
> >(and Pete Gural and I have written programs to do this).

>Yes, but with only one source of information (that is, a single "eye"), how
>do you differentiate between one object moving x degrees across the FOV and
>another object but at a very different distance, both of which present the
>same size to the retina and the same angular movement?

You don't.  The information is processed to give *orientation* and time of
closest approach, not absolute distance.  The radiant of a meteor is the
orientation of the meteor.  If one wished, one could place real world
constraints on the velocity of the meteor such that it had a prior
elliptical orbit before hitting the atmosphere.  The range of velocities
between the parabolic limit and 11 km/sec, give a range of absolute
distances.  In reality, atmospheric deceleration complicates this, but
it gives a first approximation.

Cheers, Rob

To UNSUBSCRIBE from the 'meteorobs' email list, use the Web form at:
http://www.tiacdot net/users/lewkaren/meteorobs/subscribe.html
References: