On Mar 24, 2009, at 11:15 AM, Fabien Chéreau wrote:

2009/3/24 Timothy Reaves <treaves@silverfieldstech.com>:

On Mar 24, 2009, at 9:24 AM, Fabien Chéreau wrote:

I don't really get the meaning of this bug.
The star rendering does increase the stars halo size, but it does it
so that the relative luminosity is consistent between stars. It is not
an easy task to change this behaviour, but what you could do is to
adjust the luminosity scaling parameters when your plugin zooms in.

On Tue, Mar 24, 2009 at 14:51, Timothy Reaves
<treaves@silverfieldstech.com> wrote:

       I don't think anyone is bothering to actually look at the bug

tracker, so I'll follow up here.  This one is annoying, and should be

fixed soon (I hope).


The problem is that stars are point sources of light.  They are not 1 arc
minute across.  This bug is in the zooming code, not in the plugin.  If it's
the luminosity code, then THAT code needs to take zoom level into account.

As you says, stars are point sources, their radius of their disk
should have a size of 0.000000000000000000000000000001 pixels...
The thing is that what you see with your eyes is the Point Spread
Function of your eye + optic instrument as an optical system.
The current code tries hard to simulate what a user would expect for
zoom ranges scaling from naked eye for large FOV to hubble images for
small FOV.

The absolute size of the halo is arbitrary, and is set by the option
in the View window.


This doesn't address the defect though.  The rendering code should adjust for zoom level, not what initiates the zoom.