> First, the human eye is only able with many precautions and much effort
> to distinguish intensity differences of more than 5 %.

You're too optimistic here! Those 5% may be valid under dayight
conditions, but under nighttime conditions, when estimating the
brightness of stars, we need a 10% intensity difference to
distinguish it.

> Although our displays have 256 levels,

Again, the displays don't have these levels -- they are nowadays almost
always analog devices. It's the video display adapter (or "VGA card")
in your computer which has 256, or some other number of, levels.

> therefore, we are able only to distinguish about 20 levels of intensity
> more or less simultaneously.

Again, that depends on the intensity of each level. You here seem to
assume that the levels increase linearly in intensity. But they
could just as well increase logarithmically, such that each level is
10% brighter than the previous level - in that case we'd be able to
distinguish all levels from one another.

> Divided evenly among 10 magnitudes, this would be about 0.5 mag steps.
>
> Then, do you really want to be 6th mag stars in GUIDE output to be as
> difficult to see as in the real world?

Definitely not. But i think it would be quite hard to make a point on
the video screen really shine with an apparent magnitude of 6. It's the
same with planetariums - the artificial stars on the planetarium dome
shine brighter than the corresponding stars in the real sky.

> To be visible at all, the lowest brightness of that 20 level scale would
> have to be rather large, at least several mm across, and that is exactly
> the problem we meet when representing Sirius in the diameter scale. Not
> very attractive, I think!

I though we were discussing representing stellar brightnesses with points
of varying brighness but the same size, not points of varying sizes.

Now, if the faintest dots on the screen would have to be several mm
large to be seen at all, shouldn't the same be valid for the real stars
in the sky? I.e. wouldn't a star of magnitude 6 have to be, say, at
least 1/10 degree or so in apparent size to be visible to our eyes?

> There is one thing we might try, however, that is to use the intensity
> handle as well as the radius handle to represent star brightness. When we
> would represent the weaker stars in say, 4 brightness levels, the radius
> requirement could be a little relaxed.
>
> I am asking myself if I want to try and find the weakest stars in that
> representation, but on my display anyhow, a dim 1.5 mm dot might be a better
> approximation to what I see than a lopsided four-bright-pixel approximation.

I was considering 1-pixel approximations.... after all, real stars are,
to our eyes, point sources, not small squares or circles.

----------------------------------------------------------------
Paul Schlyter, Swedish Amateur Astronomer's Society (SAAF)
Grev Turegatan 40, S-114 38 Stockholm, SWEDEN
e-mail: pausch at saaf dot se or paul.schlyter at ausys dot se
WWW: http://hotel04.ausys.se/pausch http://welcome.to/pausch