multisampling

I hope this hasnt already been covered too often, but if you have like 128 mb memory installed, wouldnt you just sample a sound for each semitone, at least for shorter sounds, when creating a multisample? Doesnt all this pitching/detuning stuff come from ages when memory was scarce and expensive?

I ask this because I read in Stereoids very helpful faq about multisampling und also downloaded some nice free fender rhodes piano and other pgms from the net, but especially with that rhodes one it is a pity that there are notes created from detuned samples, because it just sounds like sample, not like real rhodes, if you listen carefully, and i am not even speaking about sampling different velocities.. If you detune a sample by even just a halftone, you will get a different result compared to the original sound at that pitch.

I wonder the sampling cds sold by the founders of this site and other ones are different?

It really depends on the sound, I think: some sounds can handle detuning without sounding strange: harmonically simple sounds, even electric pianos Sometimes you want it to sound strange, actually, I play around with Mellotron strings, which sure don't sound natural.

The worst are voices, which you can't detune much without sounding like Darth Vader or a Smurf*, and that has to do with Formants. (When you sing higher or lower, there is a Formant component to your voice that hardly changes - that gives your voice its character, and it's why you don't sound like you if you detune it.)

* If you don't know what a Smurf is, use Google if you really want to find out. I wouldn't bother, personally.

Yes, but that only means it sounds ok because it sounds similar to what you would expect. It doesnt mean its really the same sound. An acoustic piano is the best example, and therefore takes so much memory to sample.
But that does not only cover acoustic instruments, also synthesizers often produce lots of wave-spectrums around your selected note/frequency, and if you simply pitch that sample up, it will not give the same result as sampling the higher note from the synth. And then you have the 44khz resolution, which means that even a sinus wave is rastered into little "corners" when sampled. So if you sample it very low you should be able to pitch it up and it should sound ok. but the other way round -sampling it at a high tone and pitching it down will produce some noise.

stereoroid wrote:
* If you don't know what a Smurf is, use Google if you really want to find out. I wouldn't bother, personally.

I know the smurfs. When I was a kid, they had same hit singles. Recently they showed this one guy on tv who made one smurf record. It was his only hit, but made him a millionaire, he lives in a castle now and makes horrible music.