Bits is bits again

I was going to get started on our 'fundamental' conversation started yesterday. But reading through the PS Forums I ran across an explanation by our own Ted Smith about some of the issues affecting the sound of digital audio. I wanted to post it here, so informative that comment. When asked about why a RamDisk and its various setting might make an audible difference, Ted responded with this:

As a handwaving example the RamDisk can allow fewer disk reads/seeks during playing. Seeking fast costs energy, the faster the more energy. Each seek adds a big current spike/dip to the power supply. Moving these spike around will change when their effects happen. Reading a file all at once can cause fewer seeks over all and hence less interference. That interference can show up in the system in various ways – thru the AC used to power each unit of your system, thru RF (which often shows up because of ground loops, as well as high frequency hash over the interconnects, etc.), thru jitter via the DAC… Another possible mechanism: Using a RAM disk can cause the file to be more contiguous in physical memory in the PC, perhaps causing fewer memory cache collisions with the rest of the running system (or more appropriately, perhaps not being as interfered with by the rest of the system.) Noise from the PC affects all of the components of your system thru varying paths to varying degrees. It's a very complicated system and figuring out from first principles all of the interactions is like using physics to predict chemistry. Possible in principle, impractical in the extreme. A DAC can only do so much to help the whole system and jitter attenuation is among the things that a DAC can do to help, but jitter itself is a secondary effect, both at the source(s) and the destination.

Well said, and thanks Ted.