The "7th channel myth" and thoughts on extra channels

Discussion in 'Archived Threads 2001-2004' started by RichardMA, Jul 7, 2002.

  1. RichardMA

    RichardMA Second Unit

    Joined:
    Apr 16, 2002
    Messages:
    446
    Likes Received:
    0
    This "7th channel" misunderstanding is helped along by manufacturers who do nothing to dispel the idea of "7.1" channels.
    Currently, there is no such thing. THX decorrelation applied to the two monophonic rear channels
    results in a "kind" of 6 and 7th channels, but they are
    not true channels because they are not capable of proper
    steering, and each has the information the other uses,
    but at different times. Right now, the most
    channels are supplied by DTS ES Discrete, which has 6.1
    discrete channels as opposed to DTS Matrix, Dolby EX
    and THX EX which provide only 5.1 discrete channels and
    a 6th matrixed channel derived from the left and right
    surround channels. THX decorrelation was first implemented
    in Pro-Logic decoders using THX circuitry some time ago.
    The difference between a discrete channel and a matrix
    channel mainly concerns separation. Discrete channels,
    as in Dolby Digital 5.1 or DTS 5.1 or DTS ES Discrete 6.1
    have separations of around 70-80db. Matrix derived channels
    (such as all the channels in old Pro-Logic, or Dolby EX's
    centre surround channel) are only separated by about 35db
    (Circle Surround, another matrix method is about 40db)
    so you hear "bleed" in these channels from the ones they were derived from and vice-versa.
    Interestingly, all matrix circuits are NOT created equal.
    Depending on the processor, some matrix channels in some
    systems are not as well separated as in other systems.
    There are various schemes out there for decoding matrixed channels. Dolby and DTS provide "flags" in the software
    to help with the correct "steering" of the matrix channels.
    But there has been some problems with these flags and
    THX EX (THX's version of Dolby EX). Dolby stopped using flags and I don't know if they've started putting them back in yet. What it means is that these channels can be derived from any multichannel source, whether there are flags embedded in the software or not. Some of the schemes like Circle Surround or Lexicon's Logic 7 can even derive multiple channels from 2 channel PCM sources, but that basically resembles the old Pro-Logic method and all those
    channels have very weak separation.
    If people want to have "lots" of channels, Circle Surround
    (and probably any matrix decoder, even old Pro-Logic receivers) claims to be able to derive Sony SDDS information
    (8.1 channels in theater but not available to consumers)
    from Sony DVD movies. There was an article in WSR a while
    back from the president of Smart Industries (Circle Surround) who claimed that the full 8.1 SDDS channel information is mixed into some standard 5.1 Sony movies,
    only needing to be extracted by a decoder like his.
    Of course, if it is true, you don't get 8.1 discrete channels like full theatrical SDDS, but you would get
    8.1 channels consisting of a discrete left-centre-right
    front channels, left right discrete surround channels,
    one matrix rear centre channel and two matrix front left
    and right channels, higher up than the standard discrete
    left and right front channels, kind of like Yamaha's pre-existing system.
    To this end, you can interpose a Circle Surround decoder
    between any two discrete or maybe even matrixed channels
    to see if you can derive differential information. Yamaha
    also uses this method with it's flagship receivers to
    create a 7th and 8th channels positioned wider and above
    the front left and right speakers. Parasound's (discontinued?) CSE 6.1 add-on decoder which allows older
    5.1 gear to have 6th and 7th monophonic channels also has
    outputs for two channels positioned on the ceiling about
    mid-point in the home theater. This, they say are the two
    channels Dolby elected NOT to provide for real theaters
    (owing to cost of extra speakers) when they implemented their Dolby EX scheme for movie theaters. I think "The
    Phantom Menace" was the first theatrical film to contain
    Dolby EX flags.
    So it's an interesting time for home theater and you can
    experiment with extra channels if you buy things like
    Circle Surrounds Jr. units and use them between your
    discrete channels.
    Theoretically, though I don't know exactly how this
    would work out, you could have a situation where
    you have 5 or 6 discrete channels and matrix channels in
    between or above or below those channels. This might
    result in a home theater environment where there are
    10-14 or more channels. One method, "Ambisonics" apparently
    has produced a 10 channel home theater system, but they are
    all matrixed channels so it's kind of like a "super" Pro-Logic, not IMO preferable to a discrete system with fewer channels.
    The thing that is the crux of the whole issue is this;
    Which matrix channels are truly channels (albeit with reduced separation compared to discrete channels) but are
    CAPABLE of providing accurate steering? In other words,
    I know that the 6th rear channel in a DTS ES or Dolby EX
    or THX EX system provides true, direction behaviour
    (I've tested it) but will additional matrixed channels derived by interposing a decoder between discrete channels
    provide real steering of sounds? THX EX's decorrelation
    for the 6th and 7th rear surrounds does not provide this.
    It only provides a "difference" designed to fool the ear
    into thinking there are 2 channels. This IMO, is no more than filler or dipole speaker-like diffusion. It does serve a purpose (designed for larger listening environments
    where one speaker would not serve the each seat well) but
    in fact, it is not much different than simply taking the 6th
    channel and splitting it to multiple speakers, as is done
    in theaters with surround channels now.
    In fact, Dolby only recommends splitting the 6th channel
    into two speakers in large listening environments, probably
    to prevent a dilution of the directionality the 6th
    channel provides.
    But further experimentation with different movie soundtracks
    and separate matrix decoders is needed to see exactly what
    is possible as far as extra matrix channels goes.
    Provided separation can be kept to at least 25db (about the
    same as stereo on old vinyl records) and PROVIDED the
    information is not simply diffused sound, that it's truly directional or not clealy audible (important!) in the standard channelsthen these extra channels should be
    sought out to provide a more accurate more enveloping home theater environment.
     
  2. PatrickTy

    PatrickTy Extra

    Joined:
    Jun 26, 2002
    Messages:
    19
    Likes Received:
    0
    >>.The difference between a discrete channel and a matrix
    channel mainly concerns separation. Discrete channels,
    as in Dolby Digital 5.1 or DTS 5.1 or DTS ES Discrete 6.1
    have separations of around 70-80db. Matrix derived channels
    (such as all the channels in old Pro-Logic, or Dolby EX's
    centre surround channel) are only separated by about 35db
    (Circle Surround, another matrix method is about 40db)>There are various schemes out there for decoding matrixed channels. Dolby and DTS provide "flags" in the software
    to help with the correct "steering" of the matrix channels.>Some of the schemes like Circle Surround or Lexicon's Logic 7 can even derive multiple channels from 2 channel PCM sources, but that basically resembles the old Pro-Logic method and all those
    channels have very weak separation.>I think "The
    Phantom Menace" was the first theatrical film to contain
    Dolby EX flags.I know that the 6th rear channel in a DTS ES or Dolby EX
    or THX EX system provides true, direction behaviour
    (I've tested it) but will additional matrixed channels derived by interposing a decoder between discrete channels
    provide real steering of sounds?
     
  3. ling_w

    ling_w Second Unit

    Joined:
    Sep 3, 2001
    Messages:
    426
    Likes Received:
    0
     
  4. PatrickTy

    PatrickTy Extra

    Joined:
    Jun 26, 2002
    Messages:
    19
    Likes Received:
    0
    >>As far as Ambisonic for large space and varied audience seating, it could be implemented with the deployment of enough channels in the XY axis so that adjacent speakers are at the limit of perceivable difference in horizontal imaging. >And even with std Ambisonic, it is still 1000x better in reproducing realistic 3-D sound than any of the bare bone unprocessed or matrixed format.>It (Ambisonics) is not a steering, logic matrix or any of whatever you have heard of.>Ambisonic is doing 10ch playback system because its technology easily...
     
  5. ling_w

    ling_w Second Unit

    Joined:
    Sep 3, 2001
    Messages:
    426
    Likes Received:
    0
     
  6. PatrickTy

    PatrickTy Extra

    Joined:
    Jun 26, 2002
    Messages:
    19
    Likes Received:
    0
    >>Are you listening to UHJ straight? B-format with the WXY information could be extracted from the UHJ format. Not the most optimal thing, but what else is there that could do 10 channels from a 2ch carrier? I heard many UHJ encoded discs (mostly from Nimbus) and don't notice the phasey sound. Phasey sound occurs in Superstereo though.
     
  7. ling_w

    ling_w Second Unit

    Joined:
    Sep 3, 2001
    Messages:
    426
    Likes Received:
    0
    Pan-pot mixing is never good. The concept of side imaging is non-existent. Decoded ambisonic signal includes the equivalent of pan-pot mixed signal in addition to the wavefront simulation. So even off axis, the imaging by loudness intensity between speakers will still be there.
    Second order encoding records the higher order spherical harmonics of the sound. 0th order is W, 1st order is XYZ, and 2nd order consists of the RSTUV components. So a total of 9 channels are needed. A graphical representation of the higher order spherical components could be seen here:
    http://members.tripod.com/martin_lee.../harmonic.html
    Ambisonic in off-axis and its relation to second order harmonics encoding is quoted from
    http://www.york.ac.uk/inst/mustech/3...omogeneous.htm
    However the degradation as the listener moves off centre is relatively graceful and sufficiently similar to what would happen in a natural soundfield as not to be unduly disturbing to a listener [26]. Daniel has shown [24] by a statistical analysis of the pattern of localisation errors which occur as a listener moves off centre that the wide area performance is very significantly improved by simply going from first to second order, all other things staying the same.
    A description of 2nd order from
    http://www.helsinki.fi/~ssyreeni/dsound/dsound-c-12

    In an ideal first order system the sound will always come from the correct hemisphere, with a cosine directional weighting. With higher order systems, we can approximate point sources, which is what we need in order to present a stable spatial image to a larger audience. This implies a broadened sweet spot even in a second order ambisonic setup, although the level of realism even in a first order system is amazing.

    A paper that compares first to second order ambi encoding:
    A simple paper:
    http://audiolab.uwaterloo.ca/~jeffb/info/aes/paper.html
    A more complicated mathematical paper:
    http://audiolab.uwaterloo.ca/~jeffb/thesis/thesis.html
    So second order improves imaging dramatically for a wider audience, but also requires much more processing capability. A third order encoding was also written up by Michael Gerzon, as could be seen here:
    http://www.york.ac.uk/inst/mustech/3.../secondor.html
    But the need of 16 channels and massive decoding power probably puts it out of reach for everyone.
     
  8. Daryl L

    Daryl L Supporting Actor

    Joined:
    Sep 26, 1999
    Messages:
    762
    Likes Received:
    0
    HUH!!!!! LOL! Dang this is interesting. But whats SQ? When I got my Integra DTR-7.2 and added a back surround channel dad asked why do I need so many speakers. I explained 5.1 and 6.1 discrete encoding/decoding on dvd's to him. The played THX's Tex and Jungle EX demo's for him. He replied' "Hmmm That's nice" [​IMG] Then after hearing the plain fly from front center to back center speakers in the Jungle Demo he said all I need now in a speaker in the ceiling between the front and back centers to fill the gap of the flyover. I sold my Integra and got the Harmon Kardon AVR8000 for logic7, added a second back surround speaker. Now all I need is my ceiling mounted center speaker! [​IMG] Any Suggestions? LOL
     
  9. ling_w

    ling_w Second Unit

    Joined:
    Sep 3, 2001
    Messages:
    426
    Likes Received:
    0
    Daryl,

    That is what all the Ambisonic talk is about. The number of playback channels is not determined by how many channels it is recorded in. B-format recordings (requiring only 3 channels to carry the information) done in the early 70's could be played back in systems that consists of 20 channels (and this is discrete channels), there could be many multiples of height channel, not just for flyover, but ceiling reflections in concert hall/church etc. There are ambisonic systems with 64 channels out there, and all utilizing the same recordings as every other ambisonic user.

    As I said before, to understand Ambisonic, you have to unlearn the junk formats they have been throwing at you.
     

Share This Page