I'm a mathematician and programmer working on experimental surround sound techniques. Some of the ideas I'm working on require dozens of channels. These channels are often highly correlated and are very well suited to compression. I'm new to Vorbis. Does it attempt to address such issues? Thanks, --Richard --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
"Richard W.E. Furse" wrote:> I'm a mathematician and programmer working on experimental surround sound > techniques. Some of the ideas I'm working on require dozens of channels. > These channels are often highly correlated and are very well suited to > compression. > > I'm new to Vorbis. Does it attempt to address such issues? > > Thanks, > > --RichardI believe that the current goal is to implement ambisonics. From August> Ralph Giles wrote: > > > On Mon, 14 Aug 2000, Jean-Marc Valin wrote: > > > > > I saw from the interview on Slashdot that there is no support for channel > > > coupling in Vorbis. I would like to help work on that. Has anything been started > > > about that? I don't know yet whether if could work "in real life", but I have > > > started to work on a way to do adaptative coupling between two channels (by > > > finding the best transformation to predict one channel from the other). > > > > As I understand it, the plan is to implement joint stereo and surround at > > the same time, based on ambisonic encoding. You should check the list > > archives, but there hasn't been much discussion; I've no idea what Monty > > intends as far as compression of the channels. > > > > David Carter's original proposal is at: > > > > http://www.xiph.org/archives/vorbis-dev/0844.html > > > > And there's some good background at the ambisonics website: > > > > http://ambisonic.net/ > > > > > I would like to know whether I can be useful. I have some experience in audio > > > processing/coding (mostly speech). > > > > Hope that helps, > > -ralph > >-- Regards Marshall Eubanks T.M. Eubanks Multicast Technologies, Inc 10301 Democracy Lane, Suite 410 Fairfax, Virginia 22030 Phone : 703-293-9624 Fax : 703-293-9609 e-mail : tme@on-the-i.com tme@multicasttech.com http://www.on-the-i.com http://www.buzzwaves.com --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Sounds like a good start - but I hope you won't just restrict to the mere
four channels of first-order Ambisonics? I'm looking at encodings that will
use quite a few more channels (e.g. see
http://www.muse.demon.co.uk/3daudio.html for a second-order, nine channel
Ambisonic encoding) with significant redundancy. You probably want this
facility for 5.1/7.1 support anyway...
--Richard
-----Original Message-----
From: Marshall Eubanks [SMTP:tme@21rst-century.com]
Sent: Tuesday, January 23, 2001 8:34 PM
To: vorbis-dev@xiph.org
Subject: Re: [vorbis-dev] Multichannel Encoding
[...]
I believe that the current goal is to implement ambisonics. From August
[...]
begin 600 WINMAIL.DAT
M>)\^(@45`0:0" `$```````!``$``0>0!@`(````Y 0```````#H``$(@ <`
M& ```$E032Y-:6-R;W-O9G0@36%I;"Y.;W1E`#$(`0V ! `"`````@`"``$$
MD 8`H $```$````0`````P``, ,````+``\.``````(!_P\!````10``````
M``"!*Q^DOJ,0&9UN`-T!#U0"`````'9O<F)I<RUD979
>&EP:"YO<F<`4TU4
M4 !V;W)B:7,M9&5V0'AI<&@N;W)G`````!X``C !````!0```%--5% `````
M'@`#, $````4````=F]R8FES+61E=D!X:7!H+F]R9P`#`!4,`0````,`_@\&
M````'@`!, $````6````)W9O<F)I<RUD979
>&EP:"YO<F<G`````@$+, $`
M```9````4TU44#I63U)"25,M1$560%A)4$@N3U)'``````,``#D`````"P!
M.@$````>`/9?`0```!0```!V;W)B:7,M9&5V0'AI<&@N;W)G``(!]U\!````
M10````````"!*Q^DOJ,0&9UN`-T!#U0"`````'9O<F)I<RUD979
>&EP:"YO
M<F<`4TU44
!V;W)B:7,M9&5V0'AI<&@N;W)G``````,`_5\!`````P#_7P``
M```"`?8/`0````0````````#VU4!!(
!`"<```!213H@6W9O<F)I<RUD979M($UU;'1I8VAA;FYE;"!%;F-O9&EN9P#U#0$%@
,`#@```-$'`0`7`!4````!
M``(`" $!(( #``X```#1!P$`%P`5`````0`"``@!`0F `0`A````044V.#A%
M,4,W,#@U0S Q,3A$1$,X,D5%1D,V.$$P13(`9 <!`Y &`%
'```B````"P`"
M``$````+`",``0````,`)@``````"P`I``$````#`"X```````(!,0`!````
M\ ```%!#1$9%0C Y``$``@!,`````````#BANQ %Y1 :H;L(`"LJ5L(``$U3
M4%-4+D1,3 ``````3DE40?F_N $`J@`WV6X```!#.EQ724Y$3U=37&UA:6QB
M;W@N<'-T`!@`````````>IMT\UF=T!&U,T1%4U0``** ```8`````````'J;
M=/-9G= 1M3-$15-4``"D@RT`& ````````!ZFW3S69W0$;4S1$535 ``PH `
M`! ```"N:(X<<(7 $8W<@N[\:*#B)P```%)%.B!;=F]R8FES+61E=ET@375L
M=&EC:&%N;F5L($5N8V]D:6YG``,`-@``````0 `Y`$ NU'-_A< !'@!P``$`
M```G````4D4Z(%MV;W)B:7,M9&5V72!-=6QT:6-H86YN96P@16YC;V1I;F<`
M``(!<0`!````%@````' A7X*&5HXD6[Q8Q'4M32V5)IN0@P``!X`'@P!````
M!0```%--5% `````'@`?# $````9````<FEC:&%R9$!M=7-E+F1E;6]N+F-O
M+G5K``````,`!A"/C5U>`P`'$!H"```>``@0`0```&4```!33U5.1%-,24M%
M04=/3T135$%25"U"551)2$]015E/55=/3E1*55-44D535%))0U143U1(14U%
M4D5&3U520TA!3DY%3%-/1D9)4E-4+4]21$5204U"25-/3DE#4S])34Q/3TM)
M``````(!"1 !````(P,``!\#``!,! ``3%I&=1<!0%P#``H`<F-P9S$R-78R
M`/0!]R "I /C`@!C@F@*P'-E=# @!Q.'`H,`4 ^V<')Q,A"V9GT*@
C((#L)
M;PXP-;,"@ J!=6,`4
L#8P!!Q0M@;@X0,#,S"Z8&``D(8&YD!"!L:6ME("!A
M(&=O!' @<X,!D `@("T@8G4%0*!)(&AO<!? >0A@E"!W`B
G!4!J=1A0+B )
M<!A0!1!C!4!T;]T:\&@7P > "7 @`A
(<$(@$#%N;F5L!"!ORF8;L&D08'0M
M!; $@9\0L ;0! `"( W@<S\9`*PG;1> &!!K"X!G%]#W!4 )\
6@9![1!" ;
M,!\1'0/P; ,@&C 7P'%U:7IT%\)F!]$$8!N1'!<H<&4N9RX80
G@`S!U!0,@
M: ) <#HO+W=I([ N;2"!+@$`!&!N`BX%H"YU:R\S9-QA=1^ )+
C0&T)4",0
M_Q"@&[$%P!?@$' %H!=0'21\+" #`!Q0'
8=B!\W*6<@(1LP&$!I9P,`'-!C
MFP!P&E)D%T$`<&-Y(J F61F1$@!O8@&@;'G_&; J0ALP! `;L #0`Q
@X(,K
M\"8B-2XQ+S<MD#$80'5P< D1%]!N>3\L$"L0+N *H@J$"H
M+?Y2#> 000LQ
M+S@5P0% *X%G(/ :T!%$,388H#)R3ST%$&<+@ = !=
'D'-A_&=E,G,O%C&$
M,5$+$S&&X&DM,30T`4 7D#83HPS0-A-B($8#83H,@SIB$*!-$%$00"!1174)
M*[!N:P0@6U--5(10.B6094 R,1SR.F,)\'0(<"L0!:!M75\O%3= !F
",#>G
M5 I0<]4E$'DG0$H`<'4*P"OP+#(S)T !T# MX#@Z(#,T(%!-.P=4;RTWIW8%
ML!VQ+0$`=D#2> 4@:"X%L&<["#C0QFHQL3>G4F4Z.3 _N+9=!=
C$'0P`B?C
M11]5SS0?-2H6$@'0-3<+IS#6_Q5A154Y,"[A.O8Q.B\C&1!V8AQ@")!V%\
?
MXQLR8_\(< EP*E$8`#,A+($;`0=P]PM0)%!,(F$=IR*@-V(0L'QU9QHQ,&U(
M/R\C$L$``5&P``,`$! ``````P`1$ 4````#`( 0_____T `!S @B2A5?87
M`4 `"#! +M1S?X7 `0L``( (( 8``````, ```````!&``````.%````````
M`P`"@ @@!@``````P ```````$8`````$(4````````#``6 "" &``````#
M````````1@````!2A0``MPT``!X`)8 (( 8``````, ```````!&`````%2%
M```!````! ```#@N, `#`": "" &``````# ````````1@`````!A0``````
M``L`+X (( 8``````, ```````!&``````Z%`````````P`P@ @@!@``````
MP ```````$8`````$84````````#`#* "" &``````# ````````1@`````8
MA0```````!X`08 (( 8``````, ```````!&`````#:%```!`````0``````
M```>`$* "" &``````#
````````1@`````WA0```0````$`````````'@!#
M@ @@!@``````P ```````$8`````.(4```$````!`````````!X`/0`!````
6!0```%)%.B ``````P`--/TW```YJ(4`
`
end
--- >8 ----
List archives: http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to
'vorbis-dev-request@xiph.org'
containing only the word 'unsubscribe' in the body. No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.
At 08:12 PM 1/23/01 -0000, you wrote:>I'm a mathematician and programmer working on experimental surround sound >techniques. Some of the ideas I'm working on require dozens of channels. >These channels are often highly correlated and are very well suited to >compression. > >I'm new to Vorbis. Does it attempt to address such issues?Vorbis supports up to 255 channels per stream. Right now, it treats them seperately, so the fact that they're highly correlated doesn't help. That's one of the big features remaining for 1.0 (though how it'll extend to more than 2 channels I'm not sure). Michael --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Ahhhh, there's my "welcome to the list" message... ;-)
The quick answer is "yes, but not as much as I'd like" - which is
a large
part of why I'm here. I don't actually know too much about MP3 -
I've read
a few books but never got my hands dirty. As I understand it, less
directional information is stored in frequency bands in which direction
perception is weakest. I'm not quite clear about that and have less of a
clue about what happens for upper frequencies. Please would you explain?
I do have some ideas that I find quite exciting about ways to combine the
MDCT with Ambisonics, but I've not done the maths formally enough to know
that they will work - but now you know where I'm coming from! Do you know
if this kind of approach has been investigated?
Thanks,
--Richard
-----Original Message-----
From: Marshall Eubanks [SMTP:tme@21rst-century.com]
Sent: Tuesday, January 23, 2001 9:50 PM
To: vorbis-dev@xiph.org
Subject: Re: [vorbis-dev] Multichannel Encoding
[...]
Dear Richard;
Have you given consideration as to the difference in channel correlation
and human audio phase / amplitude sensitivity as a function
of frequency ? This is, of course, the basis of normal stereo in MP3...
[...]
begin 600 WINMAIL.DAT
M>)\^(@@7`0:0" `$```````!``$``0>0!@`(````Y 0```````#H``$(@ <`
M& ```$E032Y-:6-R;W-O9G0@36%I;"Y.;W1E`#$(`0V ! `"`````@`"``$$
MD 8`I ,```(````1`````P``, T````+``\.`0````(!_P\!````20``````
M``"!*Q^DOJ,0&9UN`-T!#U0"```!`'1M94
R,7)S="UC96YT=7)Y+F-O;0!3
M3510`'1M94 R,7)S="UC96YT=7)Y+F-O;0`````>``(P`0````4```!33510
M`````!X``S !````%@```'1M94 R,7)S="UC96YT=7)Y+F-O;0````,`%0P!
M`````@'Y#P$```!)`````````($K'Z2^HQ 9G6X`W0$/5 (```$`=&UE0#(Q
M<G-T+6-E;G1U<GDN8V]M`%--5% `=&UE0#(Q<G-T+6-E;G1U<GDN8V]M````
M``,`_@\&````'@`!, $````8````)W1M94 R,7)S="UC96YT=7)Y+F-O;2<`
M`@$+, $````;````4TU44#I4345 ,C%24U0M0T5.5%5262Y#3TT```,``#D`
M````"P! .@`````>`/9?`0```!8```!T;65 ,C%R<W0M8V5N='5R>2YC;VT`
M```"`?=?`0```$D`````````@2L?I+ZC$!F=;@#=`0]4`@```0!T;65 ,C%R
M<W0M8V5N='5R>2YC;VT`4TU44 !T;65
,C%R<W0M8V5N='5R>2YC;VT`````
M`P#]7P$````#`/]?``````(!]@\!````! ````````T0`````P``, X````+
M``\.``````(!_P\!````10````````"!*Q^DOJ,0&9UN`-T!#U0"`````'9O
M<F)I<RUD979 >&EP:"YO<F<`4TU44
!V;W)B:7,M9&5V0'AI<&@N;W)G````
M`!X``C !````!0```%--5% `````'@`#, $````4````=F]R8FES+61E=D!X
M:7!H+F]R9P`#`!4,`@````,`_@\&````'@`!, $````6````)W9O<F)I<RUD
M979 >&EP:"YO<F<G`````@$+,
$````9````4TU44#I63U)"25,M1$560%A)
M4$@N3U)'``````,``#D`````"P! .@$````>`/9?`0```!0```!V;W)B:7,M
M9&5V0'AI<&@N;W)G``(!]U\!````10````````"!*Q^DOJ,0&9UN`-T!#U0"
M`````'9O<F)I<RUD979 >&EP:"YO<F<`4TU44
!V;W)B:7,M9&5V0'AI<&@N
M;W)G``````,`_5\!`````P#_7P`````"`?8/`0````0````````.4<<!!( !
M`"<```!213H@6W9O<F)I<RUD979=($UU;'1I8VAA;FYE;"!%;F-O9&EN9P#U
M#0$%@ ,`#@```-$'`0`7`!<`- `&``(`0P$!(( #``X```#1!P$`%P`7`#0`
M!@`"`$,!`0F `0`A````04$P-$4R.3$X1C@U0S Q,3A$1$,X,D5%1D,V.$$P
M13(`70<!`Y &`,
(```B````"P`"``$````+`",``0````,`)@``````"P`I
M``$````#`"X```````(!,0`!````\ ```%!#1$9%0C Y``$``@!,````````
M`#BANQ %Y1 :H;L(`"LJ5L(``$U34%-4+D1,3 ``````3DE40?F_N $`J@`W
MV6X```!#.EQ724Y$3U=37&UA:6QB;W@N<'-T`!@`````````>IMT\UF=T!&U
M,T1%4U0``** ```8`````````'J;=/-9G= 1M3-$15-4``!$ABT`& ``````
M``!ZFW3S69W0$;4S1$535 ``PH ``! ```"J!.*1CX7 $8W<@N[\:*#B)P``
M`%)%.B!;=F]R8FES+61E=ET@375L=&EC:&%N;F5L($5N8V]D:6YG``,`-@``
M````0 `Y`*#O$GZ7A< !'@!P``$````G````4D4Z(%MV;W)B:7,M9&5V72!-
M=6QT:6-H86YN96P@16YC;V1I;F<```(!<0`!````%@````' A99^X<^#!B/Q
M@A'4M33%(8 ]<Z0``!X`'@P!````!0```%--5% `````'@`?# $````9````
M<FEC:&%R9$!M=7-E+F1E;6]N+F-O+G5K``````,`!A!:SJ@I`P`'$, #```>
M``@0`0```&4```!!2$A(2"Q42$5215--62)714Q#3TU%5$]42$5,25-4(DU%
M4U-!1T4[+2E42$5154E#2T%.4U=%4DE3(EE%4RQ"551.3U1!4TU50TA!4TE$
M3$E+12(M5TA)0TA)4T%,05)'15!!``````(!"1 !````E 0``) $``"5!@``
M3%I&=3#TQGX#``H`<F-P9S$R-78R`/0!]R "I /C`@!C@F@*P'-E=#
@!Q.'
M`H,`4 ^V<')Q,A"V9GT*@ C((#L);PXP-;,"@ J!=6,`4 L#8P!!Q0M@;@X0
M,#,S"Z80L$)H%S$L('1H!)!E`B<$(&UY(")W92YL!: '@!>
;Q>"(&P9! !T
M(A@`!Y!S86?,92X:$!.0+2D*H@J$)0J 5!D1<74-X&L@WP!Q&% %P 0`&#!Y
M!Y 7<*1B=05 ;F\%0&$7\4450&@=4DDG9!DQ:\)E&8 M('=H#>
=P-T<46$9
M, K &? @"K$%0)QO9A[!&"
>$&T@%Z)1&C!)(&0"("<=06/4='4'0&P8(&LM(
?@'QC0&. =E ;@'/%-4#/;'J$>$'88L
EP81XP'V!F9@?1!N!O:P0@'.-E
M.R00!<!G'3$8$1! ;F2I!"!D:0`@>1HP01WQ?"!U)H
$D!E@)G$<0'2=%W!L
M&;$FL@60=&D"(.L'0!Q ;@(0<@# *1(<0O\98 6P"8
I<22@"7 ;@ GP^F,8
M(&(F<RKA'M0HQQ_0N0208V4%,"H%&%!A'G!?&6
A(2"Q'2(;@708L&/_*' *
MP2,T%Y IX!O1'C 00/\D$2AS(#$?8"^0"E C)1[0^S!Q$$!P+4 &,2FA)W R
M\=<%P"L6")!S&C!0+Z$0<"L>P A@;!XP>0A@(&6Z>
M3/QJ*(4(PU',8DMYI
M`0`=83!3(4!F"X >,#\O-#70-& I$!8P,A9A>7<X01C@&(%B"X
8L1D13;A$
M0U0>P"@P'<!!.Q'[-[ #`&,<M2/S'2(A83M4?2G1:#,S`,
B$@GP"&!G_QW
M&-$B5#!B%Y$8( /P(A#]-2%R&\ >L!SD!^ UDB)3/Q[0%[$@DQB!.=(#4B$@
MOD08X$(W!I 7@1Q1:SCB_S&",O #8 #0'<
00"4Q">'G*7$D$!E@:6<IX F
M-DL7&T
`<"4@+!J*+2U2_Q[Q"Q,:J!7!`4!&$2]@(=!9$40Q-AZ@3')/!1!G
M^PN *5%-&;1,<QJ&2X1+48,+$TN&:2TQ-#0!0(\90% 3#-!0$V(@1@-AZCH,
M@V(0H$T041! 01$,174KL24A6U--5(A0.G0'@$ R,2?!OBTM< (P"' G`!B!
M71J%+U% !F ",%&G5 I0<V0[.H 7<$H`<"'P5) @,@8S%W !T#
Q(#DZDC40
MH%!-50=4;U&G%G8%L#QQ+0$`=D!X:04@:"X%L&=5"%+0:F,H\5&G4F4Z4S!9
MN%WO!= U4"D0$#%N)9!2H2MP/P1P.=%.'T\J%A(!T#4W_PNG2M8585]54S :
M$53V2SJ]&I-$+[))]1.@&IE(,/+_-9),\"00`Z %H "!!($IY/\X,ACD)L !
M(!>Q*W WX6<1OUWE!:$)< M@9\0PLG4#@KQA=2; &.!:@#3R+QO0?FT+4"@P
M:X WD3,1.;%V_R@P&" =8221)X I`QJ$(#']*Q@_5E!%,@0`%W @,06@_PAP
M$' 7<RNA`) Q8QT@/O+_*F$7L1C@*N$CD1H18:]BMPMCKQ+!`'9 `P`0$ ``
M```#`!$0" ````,`@!#_____0 `', "U/[21A< !0 `(,*#O$GZ7A<
!"P``
M@ @@!@``````P ```````$8``````X4````````#``* "" &``````# ````
M````1@`````0A0````````,`!8 (( 8``````, ```````!&`````%*%``"W
M#0``'@`E@ @@!@``````P ```````$8`````5(4```$````$````."XP``,`
M)H (( 8``````, ```````!&``````&%````````"P`O@ @@!@``````P ``
M`````$8`````#H4````````#`#" "" &``````# ````````1@`````1A0``
M``````,`,H (( 8``````, ```````!&`````!B%````````'@!!@ @@!@``
M````P ```````$8`````-H4```$````!`````````!X`0H (( 8``````, `
M``````!&`````#>%```!`````0`````````>`$. "" &``````#
````````
M1@`````XA0```0````$`````````'@`]``$````%````4D4Z( `````#``TT
&_3<``(HV
`
end
--- >8 ----
List archives: http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to
'vorbis-dev-request@xiph.org'
containing only the word 'unsubscribe' in the body. No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.
> -----Original Message----- > From: owner-vorbis-dev@xiph.org [mailto:owner-vorbis-dev@xiph.org]On > Behalf Of Gerry > Sent: Friday, January 26, 2001 12:41 PM > To: vorbis-dev@xiph.org > Subject: Re: [vorbis-dev] Multichannel Encoding > >...> > Now, here's the thought: Considering all of this, it should, > in theory, be > possible to create perfect surround sound in just two > channels, one for eachYes, that is correct. There has been a good bit of work done on directional audio. A few years ago I interviewed for a job on a project which was designed to make use of the "cocktail party effect" in noisy environments. At a cocktail party, there is a lot of distracting noise. Our brains have sound localization built into them, which makes it possible to concentrate on a particular speaker's voice. The project was for an aircraft pilot's headset. A pilot can have many people speaking to him at once. By using directional audio, each speaker would get their own direction. So the pilot would be able to hear his copilot from ahead and to the right, ground control from behind and to the left, etc. This would better enable the listener to focus in on the audio. The biggest difficulties lay in the modelling of the ear and the sensing of the headphones' axes, so that when the listener, turned his head, the sound stayed in the same place. But I digress. I must admit I am not terribly familiar with the concepts as I did not end up taking a job with that company. But I may be able to help point in the right direction, if you are interested in delving deeper. You may find some data if you search on "cocktail party effect" "binaural hearing" "directional audio" Mark Borgerding --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> -----Original Message----- > From: owner-vorbis-dev@xiph.org [mailto:owner-vorbis-dev@xiph.org]On > Behalf Of Marshall Eubanks > Sent: Friday, January 26, 2001 12:53 PM > To: vorbis-dev@xiph.org > Subject: Re: [vorbis-dev] Multichannel Encoding > For earphones, I have heard of VR work where the attitude of > the 'phones is measured and the > signal to each ear is changed in real time to replicate what > you would have heard in open air > if you turned your head the same way. My guess is you could > do a pretty good surround sound this way, but > it would take a fair amount of CPU on the client side.Depends on how many distinct directions you want to represent. The computational requirements would be proportional to the number of distinct directions you wish to implement. Excuse me while I think at the low level for a moment: I think it would work best if each direction had its own filter, which accounted for frequency response and delay. An efficient implementation would be to maintain a bank of FIR filters (length is TBD) that simulate the ears' perception of sounds from various directions. If a direction was not exactly at a point for a filter bank entyr, then the filter could be the interpolation of the filters whose directions are closest to the desired direction. This should give smooth transitions while turning one's head. It would also be pretty efficient. For example, if a sound is supposed to come from (83 degrees right, -4 degrees up). A linear 2-D interpolation might use the filters for (75,0), (90,0), (75,-15) (90,-15) The filter bank implementation would be pretty fast, but could be a space hog. It would probably end up needing hundreds of reference filters. You could space them in various manners, but that would trade space for computational speed. By the way. Is all this stuff off-topic? Or is it possible that multi-channel encoding might use this sort of thing? --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Mark Borgerding wrote:>The filter bank implementation would be pretty fast, but could be a space>hog. It would probably end up needing hundreds of reference filters. You>could space them in various manners, but that would trade space for >computational speed.I can't comment one the algorithm, but there is GPL'd code to do just this: http://www.cyberus.ca/~karen/spatial/ The example program is win32, however. THere is an X11 port, but it contains a bug that seriously harms the spatialization. http://www.ne.jp/asahi/linux/timecop/x11-reference-impl.tar.gz>By the way. Is all this stuff off-topic? Or is it possible that >multi-channel encoding might use this sort of thing?We've certainly talked about needing output tools for testing surround stuff. Using this with headphones might be a way of allowing folks with ordinary set-ups to participate. I would probably have used vorbis@ for the general discussion, though. FWIW, -r ================================================================Internet service provided by telus.net http://www.telus.net/ --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.