[sm3600-devel] Calibration scan

[Marian E. <eic...@co...>]

Hi Glenn,

> > Then we should get an option in scantool for getting a really raw origin
> > scan (without homing and that stuff).
> 
> I have been thinking about a scheme where we capture a single
> 600dpi image of the first 50mm or so of the bed and then use that
> image to find the edges of the glass, and get the black and white
> values for the brightness compensation. This would require a
> little bit of image processing but we could probably get the code
> from the Internet.
> 
> What do you think?

I am not *really* happy with this concept, because it will slow down the
initialisation phase in a form, that I will not really like (hey, I am
not the most patient use out there, ok? :-) ).

Ok, the psychological matter may be more important, because the caching
of such an area will be really expensive, and I come from those old 1 KB
Sinclair machines :-)

There might be a conceptual problem with this idea, too, and I'd like to
discuss it.

If there is a white or - worse - gray original laying on the bed, there
may be no really good chance to detect the edges of the glass safely
(even with spacial transformation magic).

This is the very reason, why I am using the home position marks as
landmarks. Of course, this may not be the real solution as well, so we
should try every promising better concept.

               Mit freundlichen Gruessen / Yours sincerely
    Dipl.-Ing, Dipl-Theol. Marian Eichholz, Koeln, Germany
                                     pos...@fr...
                              mailto:eic...@co...
                         http://privat.schlund.de/eichholz