| Author |
 Topic  |
|
|
diaz
BIAC Alum
USA
212 Posts |
 Posted - Jun 28 2004 : 6:50:50 PM
|
| I had two different display problems with cigal on both computers at the 4T which I could not fix. On the XP computer cigal would load and run the experimental files, but wasn't displaying all of the stimuli. On the win98 computer, for every file I tried to load, it would initially start loading the file, but then give an error 42 'stack (internal memory space) overload error, and if you tried to do anything else, it crashed. We tried rebooting the computers, but had the same problem. Also, all of the files run fine on the NT computer in my office and they are not substantially different from the files I used in Mask.04 (which ran on the win 98 computer at the 4T). Though they are between 20 and 60 seconds longer than the Mask.04 files. Any thoughts, I have another subject tomorrow and would like to resolve this before then. |
Michele T. Diaz, Ph.D.
Associate Director
Brain Imaging and Analysis Center |
|
|
diaz
BIAC Alum
USA
212 Posts |
Posted - Jun 29 2004 : 10:14:55 AM
|
I determined that the Cigal display problem on the XP computer at the 4T is related to the new font designation convention. If I use the new way to specify fonts (e.g. font "courier new bold" 30 in the header and font 0 0 0 0 just before the first stimulus) this results in not every stimulus being displayed. If I revert back to the old way of designating fonts (e.g., fontnum = 41 in the header). The display looks fine (displays all of the stimuli). Again both methods work on my office NT computer, so I'm not sure exactly what is going wrong.
|
Michele T. Diaz, Ph.D.
Associate Director
Brain Imaging and Analysis Center |
 |
|
|
jim.voyvodic
BIAC Faculty
138 Posts |
Posted - Jun 29 2004 : 12:07:37 PM
|
Curious. The "old" way of specifying fonts is a bit limited and machine dependent because CIGAL simply asks the computer what fonts are pre-installed, makes a table of those, and then you can access them by number (eg. font 41 is the 41st font in
the list the computer reported). The advantage is that these fonts are definitely there and supported, the disadvantage is that they may vary from machine to machine (or operating system). The "new" way (added a year or two ago) lets you specify a particular font by name and size and then load that (as CIGAL's font number 0). The advantage is that you have a better chance of getting the specific font you want. I haven't played with it enough to see whether all symbols are supported, or if there are subtle interactions, or just bugs in CIGAL. Within CIGAL's documentation we refer to the old and new way of specifying fonts as using "hard" and "soft" fonts, although these terms may be technically incorrect. I have never experienced a problem with the soft fonts working erratically, and wonder if there is a pattern to the text strings you are using; perhaps some symbols are not defined in all fonts. There may be a memory or resource leak in the program which only shows up after many repeated calls to the font. If you send me your stimulus file I will try to figure out what's going wrong.
Jim
|
|
|
| |
Topic |
|
|
|
Cigal Display Problems
