On 1101245076 seconds since the Beginning of the UNIX epoch
Michael Kukat wrote:
>

>1) The routine to look up matching clockchip programming for a resolution is
>quite a dirty hack. There should not only be a scan for H/V resolution but als
>o
>for vertical frequency. So all table entries could be present at the same time
>,
>and one has a chance to get the 60Hz for TFTs without recompiling the xsrc tre
>e
>:) Okay, i used hexedit to change the table in the binary :) What is the best
>place to address hints/patches for this? Here or directly to XFree?

For the ELSA or for the TGA?  The TGA code is in the kernel and
not the XServer last I checked.  The routines to look up the modes
are a reproduction of the basic logic that DEC used to configure
these cards based on the DIP switches on the back...

>2) The image quality of the TGA2 is not really the finest. For me it looks lik
>e
>the video signals are a bit too high, as black is not really black and white i
>s
>much brighter than from the ELSA. And there are some strange stripes and
>something which looks like echos (but left of the real signal, not right as
>usual). Does someone know if the TGA cards have different video signal
>impedances? Did someone has experience with soldering some resistors to the
>output or other tweaks to get this working a bit more useful?

You might have sync-on-green turned on for the card.  Can't remember
if I left it as the default configuration.  If so, the easiest
thing to do would be to compile a kernel after hacking those lines
of the driver and see if that resolves the issue.

--
    Roland Dowdeswell                      http://www.Imrryr.ORG/~elric/