Sorry Brett - I know I promised not to spend any more time on this topic!
I think it's generally an excellent and informative article, although perhaps not as complete as it might ideally be. It mostly covers things that have already been discussed at length (although not in such detail in one place) in many posts over time on this forum.
It does largely focus on the compromises of wet sampling, which is perhaps fair enough since Dr. Pykett has covered some of the key compromises of dry sampling in other articles, although it would be nice to see some of the advantages of wet sampling included too, most significantly:
- That wet sampling makes it possible to play at home (ideally on headphones) a sample set of a specific (e.g. famous or historic) real organ in its original acoustic, with the per-pipe differences in room acoustic (spatial information) reproduced.
- That wet sampling can yield per-pipe differences in acoustic, which with dry sampling could only be achieved with one speaker per pipe, or one impulse response/reverb unit/instance per pipe.
- That the end result of playing back a wet sample set sounds no worse (and no better) than listening to a CD recording made of the real organ from the same microphone positions.
- When listened to via headphones and recorded binaurally, then the result is essentially identical to what you would would have heard with your head in the position of the microphones (apart from slight differences in microphone response, non-linearities of the organ, etc.).
Many of Dr. Pykett's shortcomings of wet sampling are results of playing the output back through speakers in a room, i.e. that the listening room's acoustic will unnaturally change the result, and he seems to have dismissed headphone listening as relevant for digital organs. However, for Hauptwerk's wet sample set home users, headphone listening is often practicable and preferred, and it's the recommended approach by many wet sample set producers, precisely because it allows the original acoustic to be reproduced properly. (An anechoic, near-field listening environment being second-best.)
With (good) binaural recording and headphone playback, I'd say that it's only really the inability to hear the sound field change in response to moving one's head that's a significant compromise in terms (specifically) of acoustics.
Edit: P.S. Conversely, there are some other significant compromises of wet sampling that the article doesn't mention, such as the fact that non-static effects like swell box movements, wind fluctuations and tremulants can't be reproduced as accurately if the arey applied artificially to samples that are already wet (since they should theoretically occur before the effects of the acoustic/reverb).
Best regards, Martin.
Hauptwerk software designer/developer, Milan Digital Audio.