HZPiano Wouldn't sampling all 88 notes be an equal or less amount of work while still resulting in a more authentic sample set?
More authentic, yes. Less work, no. I record roughly 1/3 of the keys, so consider the time recording and editing: I have to concentrate to perform the notes at the seven velocities, pedal up, and pedal down. This is actually only about 1.5 hours of actual recording - assuming no mess ups, no gardeners in the background, no pots and pans clanging in the kitchen, etc. And it's fairly intense concentration to detect background sounds in case I need to record the note again, which increases the time. If I did all 88 keys, I would need to concentrate for, say 4.5 hours, which is rough for me, and also more inconvenient for the piano owner as I'd be there for more time.
Now, when I get home, I cut up the samples into roughly 2600 separate files - I have some automation to do this, but 88 keys would be 3x the time. But the really biggest time saver is when I edit each note/sample. There are 2600 files per microphone pair, and until now, I could edit them all on a per note basis (group all mics together). At this point, I need to noise reduce each note on each microphone separately, because any click that I made when I touched the cable, or plane flying by, or car driving by is another 30seconds to 2 minutes of meticulous editing on 12 microphones. Most libraries are done in a recording studio, where this isn't an issue. They can spend hours and hours in the studio focusing on the recording. I do the opposite. I spend less time recording, but much more time editing.
QuasiUnaFantasia If I understand it correctly, once the technology has been developed, it can be used on all future piano's to be sampled. So, in the long run, there could a lot gained.
That's right. My aim for the past two years is to make a template in which I only need minor tweaks to make a nice piano. I'm at the point now where I have all of the basic features of a pro library, and now I can just sit back and do the recording /editing, and not so much the programming. I'm thinking that after recording, I could probably put out another piano in about one month, where-as each programming trick took multiple months to achieve. The problem will be finding good subjects to spend the time on.