Abekas, SGI and rgb

Mark Rosenstein mbr at flash.bellcore.com
Tue Oct 2 06:58:57 AEST 1990


We have an Abekas A60, that I am trying to dump images to
via rcp from a 4D/25. Abekas's normal mode is images in yuv, but
they are supposed to be able to take images in rgb format, of which 
I foolishly rendered a bunch. I have not sucessfully transfered them,
and before I convert them all to yuv myself, I'd like to check if 
anyone has successfully done this.
 
When I do the transfer to the abekas in rgb mode, I get a dot pattern imposed
on my image. There are 32 dots across the screen, and 7 scan lines
between the dots, and the rows of dots are angled down about 10 degrees.
The dots are at the same place on different images (though at different
places on different fields).

To get a minimal case,
I conduct the following experiment. Frame 420 on the abekas is fine.
>From my PI:
rcp a60:420.yuv /tmp/420.yuv        #transfer from abekas hostname a60 to my PI
rcp /tmp/420.yuv a60:421.yuv        #transfer back
and frame 421 is fine. The next experiment
rcp a60:420.rgb /tmp/420.rgb        #now do transfer in rgb mode
rcp /tmp/420.rgb a60:421.rgb
and I get the bunch of dots.

Confounding data is that my PI is on a different net than the Abekas.
Also repair has been through and changed out a bunch of boards on
the A60 to no avail. [No one to my knowledge has tried this type of transfer
here before, though].
Finally, from a Sun only feet away from the Abekas, I can do the above
experiment and only get say one dot per entire frame. Interestingly,
I can rcp the rgb file from my sgi to the sun and then over to the
Abekas, with again only say one dot per frame. 

Has anyone had any relevant experience? 

Thanks.

Mark.



More information about the Comp.sys.sgi mailing list