Originally Posted by wilbur_the_goose
Looks like Dolby is eliminating Center Spread for unknown reasons:
Here's some background on this issue from someone who was there.
When PLII was being developed by Jim Fosgate and Dolby, there was a conscious effort to address not only Lt/Rt movie soundtracks but, for the first time, stereo music. The "Center Width" control was added to do just as Gene mentioned -- prevent the steering of the main vocals to the center speaker from collapsing the stereo effect.
Why does this happen? In a 2-speaker system, the center vocal itself can produce a razor sharp phantom image, so that ought not be a spatial problem if reproduced from a single center speaker (and it's not -- something now proven with the advent of the 5.1 format). The problem is with wideband logic decoders (where wideband means the steering matrix operates without any frequency selectivity -- no filters in the audio path). When vocals are being steered to the center output, to some degree so are other, lower level stereo/spatial sounds. More on this later.
The Center Width control would, as Sanjay explained, reduce how much of the derived center signal was removed from the L/R channels, at the extreme setting (max width) the removal being totally cancelled, leaving the phantom image completely intact. Making this adjustable (8 steps to choose) let users dial in the degree needed to achieve improved center imaging vs stereo effect, and to maintain the tonal quality of the L/R speakers in cases when the center speakers were not on par. (This is before modern room EQ was able to help that significantly.) This is all highly subjective so we gave the user the final say. At any setting of the control the total energy of the vocals is maintained, only their distribution across L/R and C varies.
Aside from the above, there are other differences between the PLII Movie and Music modes in how the surround channels are handled, again to optimize the different needs of the content, but we can save that for another time if anyone is interested.
Back to the wideband steering issue. The idea of running multiple decoders, each covering a portion of the frequency spectrum, was considered decades ago as that could theoretically prevent dominant sounds in one frequency range from affecting other signals at other frequencies. It was just too taxing on the DSPs of the day to ask them to run 4 to 10 surround decoders in parallel, plus a filter bank. Nor would that ensure optimal results. Sounds good on paper, but when breaking up the spectrum and redistributing it, it does not necessarily reconstruct exactly as one would want for musical integrity. The early attempts made that abundantly clear.
Today’s DSPs are vastly more powerful, so multiband upmixers easily “fit the profile” of the hardware. A case can be made that with very good frequency selectivity it is possible to extract center vocals while not compromising the stereo effect, and having tested DSU and Center Spread on/off in the AVM 60, I’d acknowledge that this has been achieved. With well matched L/C/R ideally positioned and listener at the MLP, the difference is virtually indistinguishable. Couple that with a move to eliminate any upmixing “flavors” from the roster (Movie, Music, Game), and with the drive to eliminate user controls (Center Width, Panorama), all in the name of simplifying use for the end user, there can be what some might consider the ideal solution from an upmixer – one size fits all. Unfortunately, we do not all have perfectly matched or arranged L/C/R. And that's not the only difference -- what often comes out the surrounds from DSU is not at all the same in spirit or effect as what we hear from PLII or AuroMatic. Totally different philosophy.
I suspect that I am not the only one who rejects this “one for all” concept. If that were true, we would find no joy in having the vastly different portrayals afforded by DSU, Neural:X and AuroMatic upmixing options. And as different as they are, in my humble opinion none of them do as good a job for stereo music as PLII (or PLIIx, the 7.1-ch version). Not sure how many of you have heard PLII in a well-matched, musically clean playback system (i.e. with Dirac), but there is no other upmixer that derives as much spatial immersion with as much respect for the purity of the source material as PLII. Yes, I am biased as I had a hand in the development (well, maybe a pinky). But that’s what gives me an understanding of the innards.
My dream for an immersive surround processor is one that is just as interested in music as it is in movies. And as good as AuroMatic 2D is (and I said exactly that years ago when I owned the Marantz AV7702), it’s use of artificial reverb is something that rubs me the wrong way when I am in serious music mode. Just to name one niggle.
If the HTP-1 were to include PLIIx, and support multichannel Roon playback, I’d see it as a serious surround music contender. Until then, the Classe SSP-800 remains untouched by Monoprice, Storm, Datasat, or Trinnov.