Fons Adriaensen wrote:
On Mon, Sep 15, 2014 at 08:32:24PM +0200, Jörn Nettingsmeier wrote:
There is no established standardized main microphone for Auro 3D. In
fact, practices vary wildly, so that different Auro 3D productions
have totally different degrees of localisation, crosstalk, and
channel decorrelation. This has dramatic effects on up- or
downmixing algorithms, if format conversions are required later in
the production chain.
Which in mnay cases will be near impossible, except maybe by some
manually controlled magic and trickery, ad-hoc for each case.
Mpeg-H 3DA tries to achieve this, automatically. Results will be seen...
One vocal proponent of Auro 3D as a classical recording format
insists that a huge square of omnis is the only acceptable miking
technique. No comment.
The only thing you can hope to do with such a recording is to
deliver the signals to the speakers.
If you can use spaced omnis for 5.1 recordings, the same should be
possible for (to 3D) extended 5.1. If not, why not?
The format just defines speaker positions, not signal relationships.
As does e.g. 22.2. At least 5.1 had *some* paradigm behind it, or
could pretend to have one. That can't be said of any of the 'umpteen
speakers' formats.
I honestly would suspect that this is wrong. 22.2 has 10 speakers on the
horiz. plane (5.1 + side speakers + L/R wide + B), + upper and lower
layers. This might be a lot, but you can't say that 5.1 is more logic
than 7.1, Auro-3D or 22.2. Otherwise, please give some reasons!
It is easily possible to create a higher-order recording even
without a native HOA microphone, as long as it's understood that any
ambient microphones are used such that they do not contribute to any
co-incident holophonic image, but just add highly decorrelated
diffuse field information which is to some extent "artificial". In
that sense, they are not "ambisonic", they just use the Ambisonic
format to include some decorrelated ambience with sufficiently low
crosstalk.
And that's not really anything new - this has been done for ages
with plain old stereo.
When going that way - panned discrete sources plus a separately
generated diffuse ambience - the AMB format remains the one of
choice. At least the one that leaves open lots of ways in wich
the result can be modified in useful ways later.
I must agree with Stefan's point about 4th order AMB using the
Eigenmic or similar. It doesn't make sense at all. We would be
much better off with e.g. a second order mic followed by some
logic decoding (a Harpex on steroids).
Thank you, as you just confirmed that I had at least < some > point... A
12 capsule 2nd order mike is easier to assemble than a 32 capsule mike
(Eigenmike), easier to test, easier to calibrate and probably smaller.
(Some form of Harpex for 2nd or 3rd order would be highly interesting, a
further agreement.)
I can't help but having the impression that the whole 'spatial
HD' audio field is in a state of complete uncontrolled chaos.
Just today I was reading the MPEG specs for HOA - if ever any
standard bears all the marks of design by committee this one
does. It's depressing.
The HOA "codec" is transport-only and actually non-HOA (is this a form
of DirAC?), whereas the HOA rendering stage is a kind of re-construction
of a high-order soundfield. Or did I get this wrongly?
Which version of Mpeg-H 3DA did you see, BTW? CD or DIS? (I know only
CD, "committee draft".)
Best regards,
Stefan
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20140916/b2bc0418/attachment.html>
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit
account or options, view archives and so on.
