Please Phil Taylor help us with the "wierd aircraft" pr...

[jordanal 70]

Good luck Bert, let me know if there's anything I can do to help. Please note, I have detailed what I did over the past week to get these results in thread in the Hardware forum here:http://forums.avsim.net/dcboard.php?az=sho..._id=27027&page=(so far, 5 basic steps)Also be aware, Nvidia yanked the 162.15 beta drivers yesterday from their website. Rumor over at Guru3d forum is, they were full of corrupted profiles. Hmmm, go figure, Ryan just got done explaining all that to them, so I hope they're coming out with a new Beta with good profiles. Not a big deal as we blow away the default FS profiles in the driver and re-create them. Anyway, if you can't find the 162.15 driver, which I recommend you use, you may still be able to find them over at www.guru3d.com.

[Guest bertvankampen]

Hi Al,Thanks for your advise; I certainly will follow up when i start all over again with a clean HD, winXP...etc.RegardsBert

[tf51d 284]

FWIW, I don't have the problem with aircraft corruption, but I do have scenery corruption, mostly when using default aircraft at low levels after flying about 5-10 minutes. I had this problem also with the RTM release, which I traced to the BufferPools statements in fsx.cfg. When I didn't include these statements the problem went away. Since SP1, the BufferPools default has been upped to 4000000, so even leaving these statements, I still see the problem. I tried reducing it to 1000000 (Which was the RTM default) with same results. Last week I reduced it to 400000 and so far, with no noticeable graphic degradation, I haven't had the problem since, so I'm knocking on wood, the problem is gone! Phil has originally stated the problem was most likely caused by having too high settings for my system, which I admit I run pretty high. I have 2GB mem, 180GB hd space free and ATI 512MB X1950XTX, so I ran memstatus to see if I did run out of system resources. It showed a max of 1.6GB physical memory used, and 380mb video memory used. High yes, but that still leaves 400mb available (In addition to 4GB virtual memory) and 132mb video memory available, so I don't think lack of system resources are the problem. However there may be program limits in FS itself that maybe. Only Phil can tell us that.

[PhilTaylor 45]

I am pretty sure I mentioned we can also fail due to memory fragmentation, where a contiguous block of the size we need isnt available even though the aggregated amount free is decent sized. You could be running into that. By reducing the bufferpool size, you just reduced the contiguous size we seek so that might be the effect you are seeing.

[tf51d 284]

>I am pretty sure I mentioned we can also fail due to memory>fragmentation, where a contiguous block of the size we need>isnt available even though the aggregated amount free is>decent sized. You could be running into that. By reducing the>bufferpool size, you just reduced the contiguous size we seek>so that might be the effect you are seeing.Thanks Phil! As I said, reducing the Pool size to 400000 seems to have solved the problem for me, with absolutely no negative effects, either graphic or performance wise, so I'll keep it there!