CompuServe Thread

#PAR backup?

4 messages in this thread
#139793From: theo van bruggenDec 5, 1994 6:19 AM
All, I spend 4 days and nights on the backup/restore problems of the PAR, even installing a PAR slavedrive to make my tests possible. I found out the following things : Often (but not always) the ANI-files are corrupted at frame 896. When you restore these files to the PAR you loose everything from 896>> till the end of the file. What did i test ? I had an animation of 7500 frames. After backup/compare of this file i got the message "file is different". Then i split them in 10 files of 750 frames each. When i did a backup/compare of those 10 files, 2 files were different. I renamed the 2-files on the PAR and restored the 2 files (which were different) to the PAR. When i played them, there were corrupted after a certain frame. Then i exported a few frames as TGA's to my DOS-drive. The TGA's all looked like fractals. AND NOW THE STRANGE THING I couldn't find (visual) which were the frames that caused the problem. So i deceided to split the 2 files of 750 frames into 4 files of 375 frames. When i did a backup/compare of the 4 files there was no difference anymore. (I expected ofcourse that 2 files still would give me problems) Again i renamed the 4 files on the PAR and restored them. They all played 100% ok. All this looks very strange (but it is true !!!). I noticed the following things. When an ANI-file is corrupted it always is on a frame numbered X*128. When the compare program gives a message "files are different" you can count on it (so you have a problem; when you restore these files all frames are bad after a certain frame). (btw i used a Colorado Powertape 4GB SCSI-2, and to make sure it wasn't my tapedrive or tapesoftware also an Everex Excel 250 tapeunit). When you split this bad ANI-files again and again, it sometimes solves the problem. Sometimes going through your frames one by one you visualy can see the bad frame, removing this frame solves your problem. (So i'm sure it has nothing to do with the backup/compare/restore-procedure itself) MY SUGGESTIONS : (i learned it the hard way, loosing a lot of stuff) Backup/compare……….until you don't get the message "files are different" anymore, try to split and split and split when you have problem-files until you have a file which is 100% ok, or you have dropped your problem to a few frames. Never delete projects, even not empty ones, untill you are at the point to reformat your PAR (although it looks that the problem is solved in the 1.33B release). ANOTHER THING I NOTICED Sometimes when i start the PARTEST.EXE the program tells me i have a NTSC-board and the next time a PAL-version( which is true). Strange !. Hope this will be of some assistance to you all. CIAO FOR NOW Theo
#139911From: Christophe Van OyenDec 5, 1994 3:37 PM
Just make sure DPS learns about your findings theo. So far I never had an animation larger then a 1000 frames I was able to copy to a DOS drive and copy back that did no go corrupt at frame 896??? The Partest problem you discribe, I'll check that out after disconnecting from Cserve. Christophe
#139916From: Christophe Van OyenDec 5, 1994 3:52 PM
I just checked out when I run partest I always get PAL, without any doubt, I know I always got PAL when I ran it before. So to check what you said. I just ran about 10 times PAR test with AutoMATic Search. I even restarted my PC 3 Times turning power off and then rerunning Partest and I do not have any mention of NTSC, always straight away PAL board. Christophe
#139990From: ROBERT RITGERDec 6, 1994 12:13 AM
Theo: I appreciate all of the work you put into the PAR mysteries. /__________From your message___________\\ When an ANI-file is corrupted it always is on a frame numbered X*128. __________________________________________ I can verify that x*128 is where my animations frazzle. Bob