The DTS encoding of this one seems better then that of "Fear of a Blank Planet" (especially than the title track of that one). It is not certain why, the sound of this album is different, a different DTS encoder was used, so who would know. 24/96 has nothing to do with case at hand, as the 5.1 tracks on the DVD-A are 24/48. Though DTS can be quite good, still the lossless MLP beats it here.
BTW the stereo MLP is in 24/96, I don't know why, there appears to be no content above 24kHz.
I can think of a very good reasoin why the DTS encoding is superior.
On FOABP, I used the then latest DTS encoder - the DTS-PSE. On this one, I used the very latest build of the DTS-HD MAS suite in legacy mode.
DTS keep refining their codec and it keeps getting audibly better.
How good the DTS streanm is compared top the lossless will depend on the mix, and the complexity of it. DTS remember uses a fixed bitrate which must inevitably mean variable quality.
MLP is VBR, meaning the bitrate adjusts to the complexities (given the 3 rules of MLP) of the mix & varies the bitrate. I have seen 5.1 at 24/48 MLP as low as 2.5Mb/sec, and a lot higher.
Regarding the 96k sample rate on the stereo version.
There are many reasons for using a 96k sample rate, and not one of them has anything to do with erxtended frequency response that you cannot hear & your speakers almost certainly cannot reproduce even if your amplifier can. The main one is the filtering in the cvonverter design, as a gentle slope LPF introduced anywhere in the reproductionh chain will eliminate the ringing or pre echo effects of aliasing. The filter design is critical here, and sadly it is cheaper to make a converter sound good at 96k rather than 48k.
Another excellent reason for upsampling is when an oversampling process is used. The numbers alone make sense:
For 44.1kHz operation, a 128x oversampling converter operates internally at a whopping 5.6448MHz! Then it is downsamploed to the target rate, at which time the ultrasonic crap gets filtered out. This process can yeild an S/N ratio of as much as 120dB - or close to the operational (as opposed to the theoretical) limit of most gear.
There is an awful lot of information on this subject at
www.digido.com - Bob Katz' excellent site - or in his book "Mastering Audio - The Art & The Science".
http://www.amazon.com/Mastering-Aud...=sr_1_1?ie=UTF8&s=books&qid=1270659736&sr=8-1
So as this is not the right place, I will shut up now...sorry.