Martin Leese wrote:
Stefan Schreiber wrote:
...
Or distribute 3-4 channel UHJ, which is the stereo-compatible form of
FOA. Of course this proposal didn't "catch" on, even if this should
work. (I see this doesn't work for a CD distribution, but this is the
only case by now. But if you chose physical distribution, you still
could include the UHJ and/or B format version on some extra DVD. In this
case you would not need 2-channel UHJ at all...)
3- or 4-channel UHJ is only stereo compatible
if it plays on stereo systems.
Exactly, this was the proposal. You "hide" the 3rd and 4th channel into
a (max. 320 kbps) AAC stereo stream. (You can look back to the threads
if you wish...)
Again, in short form:
http://en.wikipedia.org/wiki/Advanced_Audio_Coding
AAC supports inclusion of 48 full-bandwidth
<http://en.wikipedia.org/wiki/Bandwidth_%28computing%29> (up to
96 kHz) audio channels <http://en.wikipedia.org/wiki/Audio_channel> in
one stream plus 16 low frequency effects (LFE
<http://en.wikipedia.org/wiki/Low-frequency_effects>, limited to
120 Hz) channels, up to 16 "coupling" or dialog channels, and up to 16
data streams
I think it would be possible to use the data streams as extension
channels. (We speak of the T and Q channel in the UHJ format. These are
not direct "speaker" channels, whereas L and R are. In this sense, I
believe it is not just a kind of "hack" but the appropiate way to
combine two audio and two "data" extension channels. A UHJ decoder
brings LRS(T) back into WXY(Z) form.)
Compare to AVC-MVC (for example): The 3D data stream (second video
stream) is embedded into some "plain" AVC video stream, which can be
read and decoded by a conventional AVC decoder...
The 2nd (and more general) suggestion or "option" was to use certain
properties of the (MP4) container format, to add 2 or more further audio
extension channels to the L/R channels, L/R and the audio extensions
being even more separated. (In different container areas. In case of
doubt, don't forget that parallel stereo and 5.1 audio streams within
the same video file are presented about in the same way. In our case,
some 3/4 channel UHJ decoder would look for the T/(Q) channels, and
convert LRT(Q) to B format.)
http://en.wikipedia.org/wiki/MP4
MPEG-4 Part 14 or MP4 is a digital multimedia
<http://en.wikipedia.org/wiki/Multimedia> format most commonly used to
store video <http://en.wikipedia.org/wiki/Digital_video> and audio
<http://en.wikipedia.org/wiki/Digital_audio>, but can also be used to
store other data such as subtitles
<http://en.wikipedia.org/wiki/Subtitles> and still images.[1]
<http://en.wikipedia.org/wiki/MP4#cite_note-1> Like most modern
container formats
<http://en.wikipedia.org/wiki/Container_format_%28digital%29>, it
allows streaming <http://en.wikipedia.org/wiki/Streaming_media> over
the Internet <http://en.wikipedia.org/wiki/Internet>. The only
official filename extension
<http://en.wikipedia.org/wiki/Filename_extension> for MPEG-4 Part 14
files is .mp4, but many have other extensions, most commonly .m4a and
.m4p. M4A (audio only) is often compressed using AAC encoding (lossy),
but can also be in Apple Lossless
<http://en.wikipedia.org/wiki/Apple_Lossless> format.
(You also could use other container formats to bundle AAC stereo files
and "extensions", like Ogg etc.)
Note that I have proposed some potential extension of UHJ to higher
(Ambisonics) orders, when I proposed a (simplistic...) Ambisonics
1st/3rd order format I have called sound field format. (Proposing also
some form of HOA which would be backward-compatible to stereo. In this
case I spoke of 3h and 3h1p .AMB files, brought into some LRTQ +
extension form. My question if you could not just code this into LRTQ -
UV PQ - the latter 4 channels from the B format hierarchy - has not been
answered on this list. ;-) - This idea came basically from Jörn
Nettingsmeier, who suggested that the stereo downmix from the .AMB
components could be derived just from the 1st order components, with few
or no obvious disadvantages.)
4-channel UHJ,
for example, is most likely to be interpreted as
quad, and how this is mixed down to stereo
depends on the system, I think. You might
hear only the two front channels (which is what
you want), or the "rear" channels might be
mixed into the front (producing a mishmash).
I did not propose to introduce some 4-channel audio format, but an
extended two-channel format. (= extended stereo)
To "implement" this idea, I, you, BBC or anybody else would need the
respective and full standard definitions, which I currently don't have.
(Secondly, I didn't see any specific "demand" here. Too few reactions,
and then I am certainly not too bored. )
Perhaps with DVD-V and/or DVD-A there is a
way of using metadata to specify the downmix
to stereo (I don't know whether this is possible,
and would be interested to hear).
Speaking of 5.1 (possible) or of quad (I think not possible.)? I don't
think that DVD-A downmixing is related to any metadata, though.
Best,
Stefan Schreiber
_______________________________________________
Sursound mailing list
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound
