Re: filmscanners: Vuescan long pass mode

["bjs" <bjs1@home.com>]

----- Original Message -----
From: <EdHamrick@aol.com>
To: <filmscanners@halftone.co.uk>
Sent: Friday, February 09, 2001 3:05 AM
Subject: Re: filmscanners: Vuescan long pass mode


>
> I've been saying on this mailing list for more than a year that
> overexposing the CCD would have charge bleeding problems.
> This limits the usefulness of the long exposure feature on
> some scanners.  It's a hardware limitation, and there's not a
> lot I can do to work around it.
>


I agree, although Vuescan long pass currently seems to make some very simple
assumptions about the CCD charge bleeding characteristics.   At the very
least it would be nice not to have Vuescan include charge bled pixels in its
output as it currently does.  The examples I posted earlier provide enough
information for a more sophisticated approach.  Whether it's worth it or not
I dunno.

Clean filters, grain reduction filters or longpass algorithms are all great
ideas but none of them work perfectly so I guess we have to accept their
flaws given today's SOTA.  I'd cast my one tiny insignificant vote for
improved longpass algorithms rather than working on better clean filters
though.

Byron