I agree that height is an individualized pinna function. So you need a
recording and reproduction system that produces a soundfield in the
listening room where height sound actually comes from up above so everybody
can enjoy it at once. But there is a difference between direct sound
sources and hall ambience sound fields. I know of no music recordings of
heavenly choirs or similar. Some movies have direct sound from above and
virtual reality may be another area where there are discrete height objects.
But for the vast majority of CDs, LPs, SACDs, MP3s, Internet Downloads, etc.
there are no elevated instruments or voices. Where height is slightly
important, is in concert hall ambience. For ambience in the horizontal
plane you can just record a rear pair, shielded from the ceiling and front
and use a binaural loudspeaker to deliver it with the key localization cues
intact. For ambience height you want surround speakers driven by 3D hall
impulses that are sorted by where the hall reverb is coming from. So you
can have as much height ambience as you like if you can afford all the
speakers. For 2.0 media that do not have an ambient rear pair you can use
impulse responses to drive multiple surround speakers in the horizontal
plane. But note that the human brain prefers hall ambience to come from the
side, the rear, the ceiling, and the front in that order of preference.
Ceiling ambience is mostly mono and thus of lesser interest to the brain so
most listeners do not miss the height speakers when they are turned off if
the horizontal plane ones are numerous.
Angelo Farina has provided us with such sets of 3D IRs. I, being a fanatic,
listen to my collection of frontal 2.0 classical music media sounding with
the rear quarter sphere of the Sydney Opera House via 24 surround speakers
some six being elevated. A group of the La Scala IRs can be downloaded from
the Ambiophonic website if you are really into this. Obviously, it would be
easier if all front stage music was delivered in a 4.0 format where the rear
pair represented just the rear horizontal circle of hall reverb so only two
speakers would be needed to, as above, reproduce it with the hall reverb
ILD, ITD and most of the pinna cues intact.
For movies that do have elevated direct sound effects encoded, you can use
the mono elevated speakers they specify, but still use binaural loudspeaker
methods for all the sound in the horizontal plane. Finally, if you only
listen to vocal soloists with a guitar, you don't need any of this,
including stereophonics.
-----Original Message-----
From: Sursound [mailto:sursound-***@music.vt.edu] On Behalf Of Augustine
Leudar
Sent: Saturday, November 04, 2017 6:26 AM
To: Surround Sound discussion group
Subject: Re: [Sursound] 3D Browser disco
Hi,
Yes for horizontal localisation you may get away with ILDs and ITDs - but
for height information the individual shape of the pinna is key and the
system of migrating frequency peaks and notches caused by the direction
dependant filterning caused by pinna shape . By filtering I mean the
frequency peaks and notches which change in accordance with height, a
filtering system that each individual brain has learnt, definitely work
better with individuslised HRTFs.... Interpoliation between HRTF data sets
is one option but there are others......
Post by Ralph GlasgalIt doesn't take a millionaire any longer. If you follow the rules of
binaural psychoacoustics you don't need to measure any HRTFs to have a
recording and loudspeaker reproduction system that delivers normal
binaural hearing cues to any reasonable standard and does not make use
of the stereophonic effect, or Ambisonic equations, but is a sort of
home-friendly Wavefield Synthesis. Like WFS the listening area is
much less restricted than that of a normal stereo system and since no
HRTFs or head tracking is involved multiple listeners can be accommodated.
The first rule is that in any such chain, there must be one head
shadow but only one and it need not be the end listener's. By head
shadow in this rule we are just considering ITD and ILD in the
frequency range where these affect localization. Since we nod,
rotate, and lean and sound goes over, under, around the front, around
the rear, etc. the ITD and ILD are quite smeared and so the brain is
used to this and can adjust to an ITD, ILD head shadow not its own.
So it does not matter how the ITD and ILD are recorded or synthesized
as long as they are normal values for a frontal or rear stage object
and then are delivered unaltered to the ear canals. So no
personalized HRTFs so far or head tracking.
The second rule operates in the higher frequency range where it is
only the pinnae that determine localization. The rule is that in any
recording and reproduction chain there must be one and only one set of
pinnae and that set must be that of the end listener. Since few mic
arrays have pinnae, including the Soundfield, there is seldom a
problem of having two sets of pinnae in a system. (can use most dummy
head mics without the outer ears.) So the only problem in having the
remaining part of a soundfield binaural, is in making sure that all
the sound sources come to the ear from a direction that is pinnae
proper. So for example central solo frontal sound should come from
the central front. All reflections or direct sound coming from the
rear should have directional components not only for the pinnae, but
also be delivered with binaural ILD and ITD values that are not then
allowed to be altered by the listener's head dimensions.
It turns out that in the horizontal plain it is not too difficult to
provide what the pinnae need in directionality without making any HF
HRTF measurements. So I and my converts easily obtain solid full
circle localization of direct sound in the horizontal plane from
ordinary 4.0 recordings like SACDs, DTS 5.1, SQ, UHJ, Dolby Atmos,
Auro 3d, etc. using just four speakers and only a modest amount of
numerical processing.
Direct
elevated sound sources can be accommodated using similar rules but
that needs more research and real music recordings to test with. Hall
ambience height is easily accommodated as is all concert hall reverb,
also in a binaural friendly way.
The rules are different for headphones or earphones, however. Lots of
AES and other papers on all this, including several on the Ambiophone
3D mic array that anyone can copy and use to make better 4.0
recordings of live sound.
-----Original Message-----
Sent: Friday, November 03, 2017 3:17 PM
To: Surround Sound discussion group
Subject: Re: [Sursound] 3D Browser disco
great sound design but no cigar on the binaural front - thats really
not going to get solved until a quick and convenient way of measuring
HRTFs is designed - I have several in case any millionaires out there
are interested ....
Post by Ralph GlasgalHave you tried https://cloud.aria3d.com Using a Chrome browser,
Christos Tsakostas, a Greek audio researcher produces an up to 180
degrees wide stage in the horizontal plane starting with ordinary
5.1 or 2.0 recordings and you can play your own UHJ's this way.
There is also no reason why this technique could not soon produce a
full circle of direct sound using just four speakers from 4.0/5.1
media including Ambisonic media decoded to 4.0 which I have heard this
way via DTS media.
Post by Ralph Glasgal_______________________________________________
Sursound mailing list
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe
here, edit account or options, view archives and so on.
--
Dr. Augustine Leudar
Artistic Director Magik Door LTD
Company Number : NI635217
Registered 63 Ballycoan rd,
Belfast BT88LL
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20171104/465
a5a41/attachment.html>