LaserSoft Imaging

In the past few days PsCS2 has started crashing on an accelerating basis as it tires to import a scan. As of today this is now happening every second scan. It is getting a little aggravating.
This happens after the scan is finishing or finished and then PsCS2 crashes before the image opens up on the monitor.

Set up: Photoshop CS2, using OS X 10.3.9 running on a 733mhz G4, 1.2Gb RAM installed with 72% allocated for Photoshop. All other programs are turned off.

i am using the newest release of SilverFast Ai6 Studio as a plug-in version for the Nikon Coolscan LS 5000 ED. This particular set of scans is being made at 4000ppi, 1.3 x .9 inches. The file size is approxiamtely 55mb for 8 bit per channel RGB scans.

I have dumped and reset preferences for SilverFast numerous times now but that hasn't fixed the problem.

In SilverFast Ai6 Studio I am using Digital ICE (normal) and doing a 2 pass scan.

I have also posted this question in the Adobe photoshop user forums.

Hi Ellis,

I wonder if you're allocating too much RAM to Photoshop. Sometimes this can cause O/S instability (especially when performing RAM instensive tasks like scanning). Have you tried lowering the RAM setting in Photoshop? I would knock it down to 50%... it's generally unecessary to set Photoshop RAM above 60-65% and I'd only recommended that if you have more than 4GB of RAM installed.

Good luck.

In addition to the things Helpful Corn mentions, you should also consider the location of CS2's scratch disk. If you do not have enough free disk space it will cause problems as well. Adobe recommends using a drive other than the one your OS uses for swap space, and the freer the disk space the better.

Thanks for the suggestions. I reduced the % of RAM allocated to Photoshop CS2 from 72% to 59% (650MB), quit PsCS2, rebooted the machine, and relaunched PsCS2. Either or both moves solved the problem

I have two scratch disks separate from the main (applications) HDD and there is plenty of scratch disk space available -- about 20Gb total.

It sounds like Helpful Corn's suggestion solved the problem. Thanks for letting us know!