Okay, so I am adding another "hard drive" to my computer this week, since all the newer sample libraries are getting so freakin' huge (plus the fact that I'm such a "sample junkie")....and wanted to ask a few questions, please.
I went ahead and purchased a really big 400GB Seagate hard drive to add to my computer (which already has its drives full of Giga files) ....and wanted to know if it should be "partitioned" when I install it.
The only thing this new drive will be used for is samples for GigaStudio. There won't be any programs running off of this hard drive... it's all for storage.
The way it is defined... "Partioning" is meant to divide memory or mass storage into isolated sections.
Do I need this drive "divided" since it's gonna be used for just one purpose? Are there direct advantages/ disadvantages to partitioning a drive when using computer based samplers? What do ya'll think?
I'm pretty sure partitioning will gain you nothing and may actually cost you space and performance. I can't give you any details since I long ago gave up on partitioning drives when one went bad on me and cost a friend of mine weeks of data.
I disagree. I think you should partition and put your more valuable items on different partitions. If one partition breaks, or you loose a directory sector, the other partitions are recoverable. Thats why I put my personal files on a different partition than my OS. Also if I want to upgrade OS, I can reformat that partition and reinstall OS software w/o losing my personal files.
As for sample streaming. Its probably best not to, as the arm of the drive needs to move greater distances when two or more samples are across multiple partitions. I do partition my drives (probably not a good idea) for sampling, and then run disk defrag on them. I am old fashined and cant break my habit
You should, however, format a sample drive for 32k clusters, not the standard smaller cluster size. This is generally the optimal cluster size, 64k is sometimes not as fast. Larger clusters get more data off more quickly.
The only advantage to small cluster sizes is reducing waste when dealing with many small files. That's not a problem with sampler data.