I routinely test BIAB using only 512m ram during the beta tests, I pull out all the sticks but two 256s at some point just to see if it will still do what it is supposed to do and to date it always does.

Things that are likely to use more RAM:

*Forte Coyote DXi synth. Loads its soundfonts into RAM. Not advisable to use with "small" amounts of ram, although with XP and 1g ram, should run okay with the "stock" soundbank it ships with. Tested to do so here.

*RealTime DX Effects -- These load directly into RAM. If you have an Audio track going on and also have a DX Effect invoked, this may prove problematic when using less than 1gig of ram. But nothing is "etched in stone" here. You should try it out and see what you get on your own system, for there are way too many variables involved.

*Single Density Ram vs DDR Double Density Ram -- If the motherboard is designed to use DDR ram but you have loaded up the two available slots with mismatched ram sizes, you may be taking a performance hit here as the motherboard will automatically revert to Single Density upon bootup if the BIOS detects two differently sized ram sticks in the slots. The ram count may look to be higher, but you are only going to get one memory cycle per clock cycle instead of two.

I tend to doubt that this particular reported problem is due to lack of ram at this point. More than likely a softsynth glitch using the Render process, or perhaps a hard to duplicate program bug, perhaps just a single instance of a corrupted songfile or something like that.

TIP: If the song plays back fine, try using another recording program such as the free Audacity or Powertracks, Realband, etc. to RECORD BIAB as it plays back from beginning to end of song to get your stereo wav file of the song playing right. This bypasses the auto-Render stuff entirely, which seems to be where the glitch is.


--Mac

PGmusic FAQs, Tutorials and Updates! click here

You must be Audiominds.
www.audiominds.com