Possibly depends on the type of music you listen to, the card and headphones. I tried it with on with For Victory or Death by Amon Amarth and it made it sound horrible, really distant/tinny. My card is an Auzen Bravura, headphones are at a700's. Trying to find some songs I have with clear channel mixing..
edit: Start of Angel of Death by Slayer is a good example, with CMSS-3d on it (subtly) mixes the two channels - not equally so as to completely ruin the mixing, but it is noticeable. Plus the general drop is audio quality for me is terrible, it seriously sound ****. Crystalliser on the other hand is alright, but doesn't make that huge of a difference really.
edit2: Actually googling around has some interesting stuff: (Hidden for those that don't care)
http://www.tomshardware.com/forum/176741-28-turn-cmss(tl;dr: Do whatever you prefer.)
astrallite wrote:Your question is an interesting one, and I will try to tackle that based on an "objective" standpoint from physics theory, since there is no commonly agreed with characteristics of "better."
If music is recorded in 2-channel format--what are talking about if the source you are playing is 2-channel, and specifically 2-channel--then it would be optimal to play it back in a 2-channel format. It has discrete channel information for left and right channels. The complex sum of this output will help you get an idea where the individual instruments, and soundstage is.
Playing it back from a matrixing system (where left and right channels are collapsed to the center and transients are thrown to the rear) would mean:
a) you've lost some stereo seperation. The center channel is playing back the a sum of the left and right channels. If there was an original center "pointsource" from the sum of left and right channels, an additional "pointsource" has been thrown into the mix and the combined effect is ambiguous. There might be an equilibrium condition where if you were perfectly lined up, a single pointsource might present itself. But much more likely, you are sitting somewhere closer between two channels (center and left, or center and right), and further from another pair. Under any of these conditions, instrumental location and 3D soundstage will present two, perhaps even three identical pointsources when the original stereo sum was only 1.
b) Very arbitrary use of transients from the front soundstage that are just played back in the rears. Some muddying of instrumental location will invariably happen. Things that happen in the rear soundfield would usually be delayed from your ears as they reach the back of the room and reflect back to your ears. Now you have a doubly effect of direct and first reflections, which would have an ambiguous effect.
c) Any music that was not recorded specific channel dominant (some recordings don't have a center sound stage, but have vocals dominant in one channel, such as Coldplay's X&Y which is explicitly left-channel dominant) are now center channel dominant. This might sound *strange* to your ears. Also, if they *were* specific channel dominant to begin with, CMSS's front soundstage collapse into the center would be *very* strange, because this would doubly focus the vocals on one side of the front soundstage.
I'm sure there are a lot more specific issues but I can't think of any right now. But music is subjective, regardless of the physical output. If you enjoy it, don't listen to what others say.
More discussions:http://www.head-fi.org/t/302413/cmss-3d ... -and-music