Am 07.03.2012 00:37, schrieb Per Boysen:
Please let me note that I did not mention such analog summing in my answer to Mark's question. I totally ment digital summing when mentioning summing as one example of digital signal processing that may give a better sonic resultMy point in my question was that it seemed to me that you mixed up sample rate (i.e. kHz) and sample depth in your line of argument, which, to paraphrase, went "it makes sense to use 96kHz for the editing process because when mixing down multiple tracks, that's better because of summing". My point: it makes sense to use e.g. 32IEEE instead of 16 bit, even if your target format is 16bit, but this has nothing to do whatsoever with the sample rate, because summing is done for each sample individually.
Other than that: I'm all for 24/48 in the recording process - 24 because of the obivous reasons mentioned in the article and 48 because I still remember ABX experiments (albeit with now outdated 80s hardware) which found an improvement of 48 vs 44.1, at least for trained ears (and the DAW calculates with 32IEEE anyway, and yes, I still believe that 64bit is rubbish).