Two stages to this one:

• Very simple implementation allowing the ptb backend specifically to do auto resampling by having _setSndFromArray create an AudioClip, do the resampling, then setup the AudioSlave using that clip's samples (in a new _setSndFromClip method)
• Refactor to give all backends the same behaviour for processing files, arrays, notes, etc. into an AudioClip object (as nothing up to this point needs to be backend-specific), then have each backend implement a _setSndFromClip method (generally the same stuff as in _setSndFromArray but with thisArray = clip.samples at the beginning)