+- Sitni Sati Forums (https://forum.afterworks.com)
+-- Forum: Products (https://forum.afterworks.com/forumdisplay.php?fid=4)
+--- Forum: FumeFX [max] (https://forum.afterworks.com/forumdisplay.php?fid=8)
+--- Thread: Weirds Bugs Make FumeFX Nearly Unusable - Everything Solved (/showthread.php?tid=1579)
Weirds Bugs Make FumeFX Nearly Unusable - Everything Solved - DasFruitLoop - 11-21-2014
Hello there, for my first post I bear less than stellar news as my first personal project with FumeFx has insofar been a complete flop due to a repeatable bug.
Following this tutorial to a T http://cgi.tutsplus.com/tutorials/creat ... -cms-21151 (including all of the settings) I begin to run a simulation but after only 3 or 4 frames the simulation freezes and my CPU usage drops to 50% and stays there. I soon discovered however that if I continue to use my computer as the simulation runs that it will continue running. Be it surfing the web or painting in Photoshop but the moment I stop and do not touch my computer for a few minutes then the simulation will freeze. Playing videogames does nothing unfortunately.
The reason I know the simulation has stopped and merely not stopped responding is that a single frame takes 2 and a half to 3 and a half minutes to simulate and deposit the data into the default folder, an hour without anything being deposited is pretty damming. By using my computer continuously I have managed to simulate 39 frames averaging the aforementioned times but I had to answer the door and the simulations stopped and I had to start over from thirty, speaking of which.
How to I set FumeFX to resume over increments of one frame and not 10, or to be disingenuous for the sake of emphasis, is there anyway I can I brute force this behaviour, even at the cost of stability because at this point it is not really saying much?
Lastly, I cannot have the preview window open for more than a few minutes or Max will crash and take my simulation with it.
My specs are: Windows 7 Professional 64BIT - Latest updates installed
I7 3930K @ 4.2GHZ -Non Hyperthreaded - (the cpu never goes above 65-70 degrees C with no other system instabilities)
64GB of RAM @ 1600MHZ
256 gig SSD OS Drive with 150 Gigs left
1TB Work/Cache HDdrive with 650 Gigs left
2 Titan Blacks - Non SLI - Stock Clocks - Driver Version 340.52
This bug is repeatable regardless of the scene I am working on and very unwieldy to work around, I am sincerely sorry for sounding so biting :$ but I have spent the last week trying to solve this issue on my own without the foggiest hint of success and thus turn to the DEVS and community for their assistance as I genuinely want to use this software to its fullest potential.
EDIT: I cannot believe it but I just noticed that in the 5 or so scenes I have attempted to run the simulation with, a tiny bit of geometry was sticking out of each container, huh. I will try running another simulation and let you guys know the results but in the meantime my bed calls me with sweet, sweet lullabies.
EDIT: One thing I forgot to mention is that I have tried playing around with the snapshot options to no avail as well in case that was supposed to fix the 10 frame increments of resuming simulations.
EDIT: Even after ensuring that no geometry intersected the borders on the container my simulation did not make it past frame 6.
EDIT: Found the source of the problem and fixed it and this is how: The problem is caused by mixing RAM kits. Even though I bought identical 32GB packs there were subtle differences between them that caused my system to not post a few years ago. On the GSkill forums I was told to up the voltage from 1.5V to 1.65V which solved all of my issues. Due to a hard-drive failure I had to reset my BIOS and never bothered to change the RAM voltage until now. Doing so fixed the issues, I assume it was because the program when exceeding the capacity of an 8GB DIMM Stick bled over into another from a different kit, encountering subtle differences in each stick's architecture, timing, and voltage and thus could not maintain stability. Neat!
I figured this all out after encountering an ancient thread from another website where someone experiencing the same issue was told it was a problem associated with populating all of his Motherboards DIMM slots and the OP's inference that it had to do with using 2 different brands.
P.S. Still cannot use the preview window without crashing Max.
EDIT: Preview window started working out of the blue, no explanation and absolutely no issues ATM.
Re: Weirds Bugs Make FumeFX Nearly Unusable - Everything Sol - sitnisati - 11-25-2014
Hello,
Thank you for your post... and the solution as well
I am glad that everything is working properly now.
As for the preview window, we're beefing it up with new debug mode for the FumeFX 4.0.
Sometimes it's just that driver has some small glitch and it does not support what it should and on our side, we maybe don't check all and every possible scenario.
It'll be much easier to handle and debug such situation in the next release.
Thank you.
Regards,
Kresimir Tkalcec