Onset Descriptors - in the land of the 'lib

After a lot of back and forth on the FluCoMa forum (not linked here as it’s not fully public and this may be, before that does) it came up by a few people (@james.bradbury and perhaps @a.harker as well) that having precise timing with regards to JIT “realtime” descriptor extraction from onsets would be a good use case for FrameLib-ing.

I have to say I’m still very green to the paradigm, but my understanding is enough to know that precision and timing is a big part of what comes with it.

Below is a patch (that does use a bunch of fluid.objects, but aren’t required) that shows what I am trying to do.

Basically there is a circular buffer recording all the time, and when an onset is detected I want to wait 512 samples, and then analyze that 512 sample window for various audio descriptors, and get the values back as quickly as possible. Depending on how fast/accurate things are, maybe even moving down to 256 samples too.

The FluCoMa version works well and offers good/usable values, but the delay/latency is way too high (random/variant but around 10ms turnaround from query to results). @a.harker’s descriptors~ performs significantly better with regards to timing, but the values are a bit funny/weird in places (test/compare the various sound examples at the top to see the results, particularly with loudness and sfm).


----------begin_max5_patcher----------
18638.3oc68s+iijibl+bO+Ujnrwgcs6QCe+37Yi914NumAl01GFfyvX2CMT
UJqpz1pTVmdz8zqg6+1OxjYJkOHyj4KoT0vZ2oZUJUpj7ijAiHXDw2+w28t6
tO4Wh2eWz+0n+Xz6d2+w28t2k9V523cY+86t6kk+xCaVtO8ic2Kw62u7o36d
u4ZGh+kCou+ljkqte41mxuv5Uoucx8+4uGRD4u61iurd6l3CoeWnr270kGd3
40ae5i6he3fosfYnEf2Gg.P8+PEo+g5sh9+d9KJ43g7uIX16ZdqCe80XyWyc
2E8+Uek+yu66z+58d1G2F+EU6tVW7vyq2m1Vi2YsWRs1Kgs2KwL8+v3szKQN
6kuOsmlc48K+b9E9q9m0WpT6N5t+N8u9q9eq+c71Ul2oWvjqoB6Orb2gHPDT
++sgTPVWmOvIxELHThEDNixPz2Gg4KT+lSurSMZqOiz+eq8Y9HzmYxqQe1wx
gW2r7qeKZax58wV6vhFWN7Xx1C6W+WRaYP8xftfCxzELbjmKXzOqsKeI8Yc2
+8cqWt4NWHz90OsUe4n6RElMl30ey2hnKrhTzQXpAzHDAS56bird9X1i2u7K
eKhystd.C5ZmVJPKjPtfvXbISvOMOfAtF8YWRBTMEvBZDUILfZqmK6bGGBfh
p8bJNcEvLQ5Wq8YXW6yBAew0qe5X97l0ai+l0IyvtOlRk1mM25JXzUP1k6dN
pyKiQhEy7d6ei8tJdDjXkIl1z4mIhoczcIcdMqZwZw93bZqnG95Cah+lRrh8
Min8sulsCDaNsCTa5eK5rpnLkx1Uz2XdsATapeKXiPWVBmQ6E0r12b9Ho7sE
XHa48bU66FzQAsHUEE6Fng67TDJgoMGilBDJSylOphAfM0Q67xeBEdtiJPyb
cwD8bbzr79FS8KYOGKmwcV6piP5rYSTPpj5LwUJiDoybMtHc1JBRpv67tHUL
i5hutd6mr0KcX2Or0NY1LVw0PSq6Od3Px16N0F2o1N6P7tOFuc48aRe..aim
nN6PXBZAUfkTrz7CQdxVAR9u6ZGezWf9vl0OXcrkvFkNbtqrjyH6GdMR0nS1
s5e1k6NQ121oIUqThjpnZEAJVPjRNUgITIlS0vg.2BNT5YDuKqyk06T5esdS
7mi2sesZF7YT6c2s70WK71uqvsngj+bR5Wj38mdq0aMuE3zasK9yqyue1o2c
4NUO8fpadbmQmuegkaJo9qIYU7tsGWm1TLuoZvIqIkNLn0Ub+qKevby5Qq7K
WP1d5bCHHcQAij5.Ph.bFhTCyOsI4gOEupXS9tjWi2td6q6h2Gu8vxCYs8SW
dU7iKOt4vGsqeb4q+XVKz5Espu66t6ocqWkrU2HJMRne67G2eLBZ11hVryj9
I1t7UK2rZ1mBVbbQkcQGNt+9k6zCTYhpP4W7PRxlxW5z8sI9wCYWVIHeaET7
Pxqtu3t0O8bC268IpK9RSe2oWY+GOt0b0OplSb3i4Gp04O2xMaxVqV9q+WVt
c8KKODeXsYH.ANcQi35m2+vtjMaJ0eMW4yVtxJ0b7Gh+x5UGdN8AUbxf5iu9
07IQ2cZTd05mh2en76cX4S6K+N6O7UCnW3sNde1Z3OdH9EkEdGp7AJcZsEWv
VTpVo2uQqFJKhSs79kkaTSCUc9hefhlUJJdAaZozf4kNcEJwLgGltJlWyUJt
k72r4kMZDx6x2YHS.TOwyR6VT+.w0vcDzEXJcBlvdClTfQvHvtI52nXYQkIq
4ZGvD.il8bg3TcNIfoEFKotxzClFEYJ66nZfJzInh6MnZbQBzbbcBxaJP89i
O9X7tLPUu9C3BYQSvz0bQnFEmyjj1B1h5O19nR11gJ19WAhOq5h12EN1yJsm
nutcreexwcOj+LyWpGUtOo1D9v5smzj6Odd5q9C50XeWaEZ41yfVgv2VAxYq
H6My0+NMxgV8QiBkeb4gC6VqL70L3UzfhNo3iRK76WtISslSywrnWz2ctw0Q
S1dH4kWTpyWylM3hn+wMG+wj+vxnujr6SpYseI52j23S1s+6+GTqZU27qI6i
+9+Aspx6+s+osnEQ+b7FEBr9ywsc+ObbmRGsUQ19dv19dRuniuj7ajrH5+iR
G1MaVFU3yk26zSQdH431ztHwlgn7tZDN03TEBGrPYANki3JywAlfrRJyivBW
FeNhCXZL7nwrrnT02s0mycG18GuurUKkQglCFOyMue8p3BaoYUQ.P5w7orHO
MxqRCQOLeTQC2gKDzpiF59wPa12UX7pxUIXYbMh+Os8P7tkObH5KqO7bzd8p
kjs6iRdL5vywQo8E0qVdHZ4t3n0aiteyw3EMNsnDTA65RAFLyYKFHynMMFeI
FueM5ocKWu8ecyxuZMHwjcsu.wo8EJPTaYskCq5gjMF267GU13RT5u.HB.Ah
IXlTejRB86foBjDxPL86HkBIVBoLJFgDBEbE7PkCOTQPFmWSgAOTE7PUvCUM
XiU4cJJKhTKaLciAMfaDWt3Os8Os82niv8Tcz1GcupiEs9wnulbToQURte6U
hF06enu4WT2V1VKIp+d2WzVro+36Vu+SQ53pP+gWFsa41UIujuoT9dRZYU+1
hstR6Bwsazme17YQFNBUzeNvBd2oAK6.inIukN4q1N8q5a.S7piqa6on3XtE
jyXm1HXN6fDqcpZs3EhlNcso2KC+sJa6Wvb4Pwy4UQytm0ptB7BnBQzQey3v
CAWJXY+xWdUsV+k8eyIxP66JvxNTk2QGqzPbSbYgHk1JeJxDkbOp9eMOMxsy
+Hds1hiWvXLr.lcJlZ7iMzoTiKTE+WVs7A2SW38cgjAAPJyEgkQ.iG5HzKlj
6llLjF6fGRcWzSNv.TuWwjAAvz.twqiywyULiun1z9jyIAtOrKuNDFywFfO8
6gJKsvmIWIohpdk172tJ9WJX2xz6RV8zDubCY5ZpJeRkdYFiAgShKR6XaCZu
sAlj1VppQd13lR2HmNK2y1AcRaGrNgGnIqcPqMUvU6f4dBM3p23ftmQCmnFG
zyQPD8s9IRrY8miWjtcaV+3yKO+jpcQaVTYym1PRMmFZLtcjh0eTlRZlS8lB
8L1f0MlLbubyoCA9e8LXe04g6FF3+7xMGiSdrt6AOio6eNY2AmfeE3+3KYMH
0Wrd5r9eQE16svGNu0irbsMIaexuG45sqOn.l5i7k9Tunlc45ZYeCY1wkupJ
aM0nddZVNMphtYz5T1NGsubR5IsHM5MwM1meYN9oG2bb8pE8nWB5ru4wBkIR
HFRnrURn9M5bWFBvSPe14gN7xwMGVueyZ0dC+eVG+EaG8.fz8b7LKBdwUMCR
JBm7vk7jGfbj4TfXjvQODN5gvQOzyid3Q0ySYz796pD0TU1Vntry95EEnRpo
oBH.SWHSqsS3D6+nlPiGTWXWx5UcDMj8ELjlXWHGJXyHn300pVaGwAQucFsT
t.cdNgwo8yDfXSxwU8XEBu2KPjnEmgBFbFAE0UgzWzn2NcExL62iUJXgPXFN
SoR9bRrgKEr8EcHCEcfRRIzwht1SL7TPc6tbnp6iOn5sYVA9GizoC6BSldWp
cW.qX8egEBxR24AqztqHZgEFCTjCva9kLzOuqcZ2d7MIRKFLRCAhe0izlfVO
Etchz8VABAJMq2Cyn8Bm4fAhy21yma3jSOt8OmrdaDwwYFxX8E3L9bkwKAaL
nGGdJsvUdd8pUkMx0MTV7+tTQgPavW+U+ghooRReSffWNAp8NvzHHVpkHAIp
9AzjgBzAUDTJiIZSWLpbXBfCSm8BkECCkusmK2XPCAP8MngLPiR+ISgPQgLD
8+hGiPHZDhdnoAsv8EsNsg+r.vPWpvsRqfoN1WZMFK30SU1Bp1LIA.RZaC6Q
aiItNsMeBbFF45z17ItYnWAbi547Ml75z17Z9F+5z17Z9F55z17Z91keLMcG
AuZbV5ESeiC4aii0wF2aj3dyYTlX77+8Ge7ythwDXWKeZ4wXBUTMFSDzPLlL
wwXBRjEGFYEbDhIHSvzPPlDBxjPPlzyJvlxhUScDJJMpB9XZEAA4x3odmreX
ooLVfVnFY.PDBvAXy4ghPyqzzpCnkKbBAFHNojwUEmjfaVXpkIUnABVpe+VZ
R0GUZd8hSvBOPvBhf0QKH+M2RvFx9VOAJB8MxZv7njxrLDeWWCTt1vKrKI6D
5sLbgcljlCCorHaGBw29SrbsRjML3xlz8a6IVMJfmOLzxp3cxMsXKWHkXfHk
E462nqCKnjkKg6CUgTnvFZwt8kZ4RH+.mbYCtvhaacR6dwP1u4VHfzxbKvMu
H965ZgYxqoVVQK7LStkO0nbvBoCDB26bUHqFZyH5HBTLylB83lDcx91u5ZEA
MTPQlRjyT6fxi5Bw+gRN0s6wBw3WrTZZU2m+KIIuD8W6pjBQ5skxY.FRsnp5
BMuhsu4w5rWizqzbANjdWmYPBWfCYL.mpO8ymsQYbv4Yb35bNpb61NuCGm4g
6y8n8y9v04eT5LP73bPZJeasxEMNOMDONQDeNUjVOYDOOcjFNgD+NkjVNojV
OsjVOwjVN0j1O4jVO8DONAEeNEktbRJMbZJsdhJMepJMexJMe5JMdBKtNkE6
mzhiSawqSbw9otTU.RUYz0tdaxpcnWTIUuW8b0avgQJMpVfSMJ4mToLsRLJs
I.uYg3stKWoc5Jua2kBQue8qIaVtKB1.VhvCELgbXw5lJ3MJXZT6pInTNTjr
T8u7lCHMO1lV1hGI.hw007RNjSQRTFYuaONTcGAIsEvoNhRWK3W4cjqE.p0g
UawBkEvsVLQUKDwZHpnf3x08qNMt22FHBO2agvwsAVdnOcW2VF6OqL3wUqS9
4T0u93eHd6wxS8xUfr4t28OorEXyo3s5cN+jJ4YYRCxhNq2W4xkhYKjj.gBc
fYggXtIDsvpWPoUhyxB2K7T.eQIR.ReKDFhSnouRfvTbszbnvsiN+rAPo4IB
jB.w7J0aAqUvoyrBY6SYpsxKqOkYdxtjWS1cRO+EXYs6+3gjm1sb05LY.fRW
uz3cYCWLtwPcW1U1RKcTmhHU+NaZV8oIG+zwGVdolPbpy2xbhNNtTdVg3zbA
AGPEByqDDbi29oYELp5gpuGl9Gt4UluwqxjhJxOV9ftHsTBuvnT9jU8BRdaW
uBJkxcsrE08OU5tk.BjaBVRIkvS6vBBRxSgPJq18WNlK4THN8ShjX0WP5qnP
.ywCOdS7KUa+bgBmMCX.nPXdkpqfP1+NJZnqR94+hxXyned418Q+b7KquOYy
p6pcK6ML1Tx1xOXjpclhWps4QFgFB..Rr+b05okd+eb8VsmOhOMgEKQPloqm
KQBUbF2UYg8mLYL1+l1wI8YAdAnxnzcweck5SJk.YU7XiucGKKbLhFNM3wyj
1z5hlJvgZUh+yULYfW1utt.Y0T3pOyLTR.T62hjEE5nm7Snvqc2qtesuh6sU
X2o7cUp+JO1chKz527dqux8sOa034xrpwtTigM6xlEn8X0SZepfdrpQEr7Zt
xw3hhe2liwUc8UuEyhgbFwnH0oNpT+VvYPGcWxW1NZ8zy8O5IYg3YRO8G+5x
wqiBPJcEMaZxwXjQJFBBg3qeG82uKNdD6ooilX8OFEE0lWb86j+6wa1j7k90
KKdZS+wzyahJA.QMcXrowjR2dtYOGn.lJOlplIPup.xwWtOd2uWo2+nMnqWx
BRmdSUJGRS6wvh6SdU5n6LJzq6o8Solt0GaxrOFgwvxrU87LiFUyCjW00+mw
mK2LglPIIfCAo2hDSX.Ylo0b3zsdov2Ywuv1R7zpm6euNevlSDUaIiZY2dMB
wpwed+w08M3e5cgULOD6LbhDljEmX0B+m0ahOgJkRb1E+482ciErhNh4EJbn
wonkXdQPtEC7N.xYTS0aViDJKD5cR7anPuCIFJn7qrPuC06UZY.1a4PuC16P
uSu1JD5cW8PuKKGQjzPr2Eh8tPr28FM16rPs.id32gx3tVgvn5DLD9cgvuKD
9cgvuKD9cgvuKD9cgvuKD9cgvuKD9cgvuKD9cgvuKD9cgvuKD9cgvuKD9cgv
uKD9cgvuKD9cgvuKD9cgvuKD9cSd320aJhLqrr8qjfuyFarW7bWFXoM0Vbu.
A2jkfx8up5a6VtY+yKeMt4xxrn2kKPNk5pvLyQ27nlS7hNP7JidQtMqYtVlX
4p74RDCDnrUXleCLwpwpyLQNPPyZ8Ylc6KDyEfQACEvrUlleyr3zkPrAOMyV
8GlQear3zUkSmOPPyZk0kAtMCveWkV2dWFYkkqst72PA3OTLTP4WaA3euUxR
91u15hA8M.+kgZqaH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.
+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3eH.+CA3
eH.+CA3eOqutr9F.P4gN1uRBw+RAimqXjENz3vyVf9OULb+ous5mnxYHz1w.
T.HqeBJfp9+2gu+0E14h99uwAxN2J7rQPmv1fv2FgdI3z0JnyhVAZVzJlEyM
Eyf4lb9bX7fOCPB1LnMPgyg4kTvbXNAYVH0jHlEsBxbXdAAOKZEduCBdJaEv
Yw7BOaDHwD1Hv9Nf.mzVwrPnk2qPPSXi.4qLKDbJaEyB8MQvYv15nYg5lnY.
R.8dp4j1JlC5aBYygIEP5rnUPlCCHyB0afnYQqXNH1b3sgyN1pd7hdtc4Qbh
Zw+4EZ4NBWhhNT1cLg123Ac.wB5.hCT2QDQyQCQCQBQQGf+9uqc2+W20+1lr
4H9NGuARaw6R+i1jADCm8O9MmjAyBKWGTrZNj3zr2wn4PiOyNFalCNtLGgXx
bRVzYO1Kadwmmwa4jzdqEgTM2T6crT1k3nrOQC1jANUhURuGI8O9HmrldsS0
6RrGPuh2wADqiCXmmIdSfQdVb8UpceFQ0XWrCyl8IdEmjYx1iKwNKlx6XQbJ
6DUi4vN2K7NNCmvdQ03IrycBuigvIrSTKVA6duvi3CbB6.0iCvl6AdD6eCHt
+llNp036qyCTdGSeSRmvZr601Fw8MRz5Yr5MIcb6wj23O54p26cL3MI89CKe
sKqLKX4zOs7PRzOoKYI2MFswrui7uf1hGvr1ky3.zVL.1R7+UM1+Lt8H82FH
L8YVHR0LsgFJfBmCNsia+yIq2FkWVfJUH..4khIqwzmsPPiYJ9ObV5+TufcT
IfynYu6yqWspTkuwdQ7pz+YFVlbPPJ6JFPgp0WHFR.4Xg5253ua1BHkJ7aME
ym2s9osIpN4l0O7ohsqhPEYLfJAlZuB4UuPvYuHv0ZAf6ZBQzQAh.r4EDkWb
NpttJuHAE8aTqtV869s+os+osutV0GSei+v+z+i+I8tU6Se+Mp2Q+W6RVuxx
kebyxCE+pxeV5CE3gjiF6H4VPbViHdoJHFN2FVqUQaJx0.ALqnGCDNWXCJ.M
+4i6Or9w0Ob5vHPW6Yk7QYVoD8FdgqXTfHWU+y2BPjPzSsEdSKwGB6KpDDx6
uPdzHIjmCfFrm6qb8q7zKBsuSudSKsF2WT4FV.cgIEnt184.vhyloTuPLes5
9NrW6k3kasXqFA229Mxv0GxVLOyYuTMLoPGUWc81CiZO8urI5ocIGeUIDxVG
1thanV6vLyLblnuc3KhEmE3AAn+B3QMKf2z+IHSYIt2SzcV0fuPKAdM5gWOp
q+s6rLufhZbdQ2gKfw8EzVlt.6N3XV4b56qRg0N2YZNJn11Jl149EzRQz1RA
z1dwyt4Bmsshl8I2K1RwxtXgxFkw0DHSMv2rmClWvkrVKR1sTfrsWbrgKnRJ
Tvj4+X6SWsLY6QIx1Q4wt8RicCkE6FKI1MVNranTX2bYvtwRfcKk+51J809V
1qcTxqarbW6tTW6tLW6tDW2MeVWqrVaojV2Z4rtdortqYdcspIjqRPc2ndNR
5LVtIsoo4+tgjmFT5XTNWrZUZDDu6yK2DoLo4k8Qq2GsKVGZDwqhTnR4r.uV
crcZ4gg6ibxPEdgadIgwA.CAox63lcfMx9.dmc5MVy9quIy3y9GMfpp8kePK
01E+VbofVHfblP1lNrMeC7KC398JwUtJbBdU2DFBvlITvTkJvo1O+VAWcKJf
N4yWMZg+qOIAroGYMtw5lUPP0xrYI9JquzDTl3QB26s1gdLixwt+5qFsWoc4
9Hk1apFRzxsqxz.PoLPltAVUB.cg.RQeARy51KJNtTiZY.YifG7RUqZlEUhC
eyLxYPMvPmtdvop5S3YaXNjW8v4PpjOCRYXbCYf2zG2U94ijqWzYcHZSz81b
vW2C0FrvvGaFJFNkvE6qieGcmYdHZs8tIsmcSJ0vbTHTe6mZG4OM9s8ebywe
L4Or7eN42c7weL4kWS1Gaqqy56HrwWsPZMtjtbzcRAPIkwY.g.iwoAlrdNOC
IQTIDXx0M.FBvR02ERRvxx4chCe+5l4D+0jycMmuBFjQByDh43VnShycQK.A
24Fbm6aB241j4zaV+pp6kxe4c1uk3lskgBkKzIhBm.L+nyLHpZ4KSmFsLbV.
bw32Ljr6ipIw8wYYvd.TDVUfhPuY.JSsY8a13WdrKSl6M7AA.wBp.Ko3LkLS
mnIgKT6ByfBHi.XFlS2mJ1JpcGx79KrSuxgyzP.qEXj1WXTnVH5EJBuwQw7f
qqEfjMp.IgypAjX1sMPVd0sx3wWbAlLwHu3lpdSkbRDGoyhJLOMxB4uEVc2H
LJG2E21PQf3Mxp6l.RNXbWbSv0ARL7MzpaW.IALxKrQTSUV4jVO273X5xZW3
GdjWQWC9Dv2FKmcBfhwckLDTc9m4j9loHX4DHvYD8VS9WuUtgR3KT3BlByrU
QWPjz4wUYaUxB62lOXTvkYpzgHfqIO81lCtfr.ADpduAEf5TZGSkUfADZnlr
cN3mul.ETLp11hofaZfxCul.g81uIXsG1qa+.910uIkxD75oDu68.IsL4Bje
vDkgFshs2JXyqwwepSNJQhGUGxgk3p63MbzaBhakl2oa0Zm60IQ8OfUYKT5x
KwfrEhl85pNaKKiGtb600HTr7WV+xobAolSM38dCO1BITfnhrLuEQztnTWCF
ABBBInPZpSMxJGCyhM9a1cs7d6oL6Bnw3aW+0dHh5BkviJJc6qhDyULc1aKg
M6hQTHS0ES2VXi6cxbBZDYe2KiwSWcUeUGjMG28pUXyCW6SDipFkXIqJzAj2
hPme9yGS6sZSl0mJ3p55SHFcaAXm34wTn5aQevDiYU70m5s0W19R4OjFwnOp
SwzHPzGTc0zWuOh657i689rL.qjQ1.cAXGoFGnX0jV.lCy70JkfFpOZRcMy6
ujmkrigBi6ByGBJJVnDziJA8o+wCOqqr0vrOl9uJaOeoJ9.1gwXxQc6dDs9f
ExnWDQ9FXv5j2IyGupHKpgUKi2PFaXJdfXx5iQlDaX9MF0joOmJBJ+fRx0t3
kou3SweIs5m7CQe53tCI6Wqe4tjMIO9n5EmJNJ+PzC6zguiKrG4x2Ef9KbCT
W3l9vNvUNkXStLqIa1Yf4kmJLM+PzgcGieMdoqLwQx6OxjlgMDRMnHKxClGV
ZapGOpINIaeb8p3sO3h8gk82PSLegBGDBDDCDTIRvsCL.zEFXZjalOKx6uF9
9Hif.65jw5KuLaDcoV9k9OP7sSHjcZmjhJWcdGkCoQz+SQe3w3k5foceMcyd
7wC6Usc0GZuVNcDiTbeEZY+NWDrwipdXPdMQUT9MrZX4R1rLT7k0aWk7Ecr5
lB3e34jWS+CMxeZXpfp.e5Kw5vgUeup87ykSVd++JiSk1rg3X.DNrM5sLhgx
pgp2jiXoBfssxwxBjSCS4pWubySI6Ve34WJqIcSCKNz+BCFUUlsMJwAySswZ
JEHw8MEHyL2V86zeRcNAQpSTBLvj..C0cXNxGxIurQzXY0fNrE2LHr5Qv6CZ
AFJVLII.pbNjscbeS8QHbJSBTt2XgXNPB4X5bfDxSOS4H70lEx4z4.aSOkYE
quonLt9r3jcqL4sF3J2zftaZnqbSC3toAmllFsSqvlr4U9N4lLosB7bfdjw9
R7pP1bfP2gxYPV3iYyApT2xr3IVnm2MMKL+8DKYA0IAKnohDi8UCBNdNv65R
7bfxyyU9ah315tMyXpJzIm1PKPv0ABtNPv0ABtNPv0ABt9cABtNPv0ABtNPv
0ABttehoBDbcffqCDbcffqCDbcffqCDb8MGAW2KhkpyryHAwdCvTZn91suQI
JMbW4Isr96MGOoMRzjVd2+sMMoAAiDMokCWAZRKPSZg5pant5FnIs.MoEnIs
.MoEnIs.MoEnIs.MoEnIs.MoEnIs.MoEnIs.MoEnIsKGMoA6retQlR00sEMo
Aw8radCPSZ+q6hcSRZPVeGeyHIMLOPRZWOW6RLqxvbYfkzBdyM3M2wsjS7fQ
roiBmGJ5CEzzvB83f5Q8.iZ9RO+FvtVjp5csaWpq3D0JoaXTUhLhXnnCL8Fp
JhDH9t2EH9tKJw2gBDeWugyBkASTf365KJVt3ThBDe23r5Ncm8.w2MFDeGJP
7cCe0cS.Yf3653paTf36FzxZTf36F1xYTf36rgfWbhuiYBACkhLAptKP0cdW
lbu0Ap.U20M7JP0c8fffPAptKP0ccGJBTcWfp65wN+AptKP0cS1NYQn.U24O
U2gBTc2.n5NTfp6FHU2U3T4sy2cntdX6W.1tCaNGpapiauQ1tq7nfKwCdwGW
iL80IsQecvEbhjfxFkfuQFIJe7A14yttOdLPtoqtfajI7BuMg7qP3CApF9Pc
meAQiqDs5wODVedvE9Apq7j2l7CYsQXygfTdjspPuNxomCdDkMtBIqOfhzdv
Co9CY1HptJH8FY.8z4xTdL0h3yIagJYjy37rJtikwwLtEBEnAz.MfFnAz.Mf
FnAzQWK7dQCnnNRCndwigSN+fhLQzxsoh.iE+fhto3GTNntBe+ZisPQ2XrEp
w0lA1BMvVnA1BMvVnWC1BkMG3Jt.kkFnrz.kk9t.kkFnrTKsBwEexsusLK7A
3TOCBNGpbNfYPIDJvupA9U8cA9U8ME+phlCjX8aIVdcDZECmQuCTMafpYCTM
afpYCTMafpYCTMafpYub6ADnZ1IZVbfpYCTMafpYCTMafpYCTMafpYCTM6ut
nZVmEs77FZxt8VJW4DRWKW4BBrP4JmGpV4WwpUNDXpcgJ4XFJBvDMWBPnZkG
pV4gpUdOSU7Hc0.wUIHh1+5VHcQV1GCgJgepsbID0N1ZW3RHiPj2W8wG6zKE
NDE5RbXka2lXQGhFcKdrcQjtDSV0MBsItrhHSBxTzGxnQZSIf.CXU88gMold
H4zGomsJA0SonMHI0OoosHQsUopsJYsEoqsKgsUordHo0GoscQhaCRcaUxay
ReaVBbyRgaTRrKow1kH6PprWRlsKctpDjpRoqc81jVWVhsxffU2W8pNh16Fk
a6R1sRwzEbJPqmZdF4pyxP7BJmIYTv4ZY.wZYSsYo4slKUkOKzRVVMTjsdp4
UoJFkDoDp9oe3A8jQ2HLDLTHla30JSoMmaszf0Tx3cEgnG2k7hEqrrAShwAk
HofjjeCAR+Xh1E+GLYDTJSZdJOiebW7+uiwae3qZdgLI5k0qVa9.5Rmo5MNt
WKPLZ2xsOEun5CvUNC0TN62OQ.hT6cMquUlplSfly.ruEoi0Jol0p4NMhMj1
wFSI7MKEq04OUMYk2NBDaALOtsU3DNzoZF3TZxFWN1THNyyZMIxKzj1JZ99J
+2b.cqY5UMrUPFJ5BYrEp+iclsHYFV2CBwBIL+MwtrJqmScs0vJagVcrqQK0
ZxZMKeUtrZqAK2Z15M+rfqIq3pYImmVyUwhtLBizTt0IByfF2x1htMpySC67
03NuLvqCF40hgd9armGF74kQedY3mGF+4mAfdYDnmFB5qwfc0fvVLJzKCCa2
3v1MPrciDa0PwlLVzsAiMXzn2FN513QaBzrs2i0OmWppZWc0y5ldb65CETPc
axg3z268QK2G8k3Maz+qVk0WSY37C6iT6B75wC10asUcWaS+0V25ysdrY7vf
Q.A2kdrsoKaMkDpqnvkdP6AyfV7pSlO7p505gqVQeTynOYrPeiQDPIagjR4f
SJ0oTQVHsUfhtQGJpZS74ERCcn.MtCEl0ARwaHrO2t4cwK2mr8rsyCE4g7Qa
Y.DQVfvDAKuh2jFPrDc.byXn7RCJPbplL8VYnIaakh6hre3iKfQabQg10FWD
hEBLmRI4Bqz0tekM9LJDWzFpafwHqleZY0ScBawJvy3sB7X+.dCA..0rHQ0U
.LoKqPa2RTubjx7aLnNaLXG9oi07dC7K0UDtJSzI72Vne0hHjcmr.aEXAcAX
MaxZuD4LPnzxGqXk0oa5gTu16bEFgpU2hrODgF249PkJPjppj1oQLPKCEcS6
7NMTT2GG0JORtGkrk7oMLVYIYoqkNpEMy1ZZoVI4XGzrrA2906h0s1u.MqZ+
zNi+bua+U9PUcWYaQKpK+5NH+uzbzjZwMMem6N4r0O+3A6mePZ89CSJJQK3m
+fe9C94O3m+fe9C94O3m+fe9C94O3m+fe9C94O3m+fe9C94O3m+fe9C94O3m
+fe9C94O3m+fe9a2O+11lwEuzz5lKtyEBSV23VLUW8S+l06OT2W8MsgRs9nU
4VSDJacqhl7aQOS4jLBFyabFz..1XhC5C3UNKksJnuLl5R.SEjsNQmXSrRCL
qAtmYJSeae3Nz9f0E3M4sOTGZejqP6KMmg5zHL5h2BwcrEBuzsPy7duZezqw
XbGwuqQ6C46ZXA4J0Bgd2Bws1BKbwRE8SOTj5+XnUzf1UbxlRSEq13SIkJpS
fTWz3GpuL4JBKzbHJAepXDnq8nRjN4Qgnba8D37jG+lfkW+9HMO6Ac.VH53B
VDkkwJygoTAlJwbhNSaMFMODv5wMIKObwPr8Ke408GRdY+2bU+h7iOocGzCV
wS.0BdZXUXtzCDD4nNoWKDIpiu6W+zV8G4Rizut7SQKUO4ut+LYnB0NTtj+L
Kojt6E2zdC85vgQRq.8Zpnuz59TBENil58etbGGItXPeAyxbNMGLASysi0bh
Ncxwh7HJinDavYSKTmY+5EhbYIv9xtrVELP.tMx7MB8xRv8uJwYYNFlxpOG6
xv3rSvxWkX5UMvl8DxTrCksc7QLYcwjrIVJYZOeTW5ViECJdBaOd30k6V9RN
2b+9RGRRNmpq13R8MlUad.KT+U7p0K2FoLi3imJSOHPT5wUptn5R6e7kxuiS
h7F4XXV5k1aivvLFx0CyHJBvApQahdXFaTCFSur6FNID3bG3H2ojbZmED3IA
MKvBuYnL9TxFk9hEDzDRcx9SGkSIsEmBzPeaEvKIoX5cSyBRNwjhIz2wNxjR
Dj9RsqPZfu+B78myA0.e+E36u.e+E36u.e+E36uK2d.A99ahlEG36u.e+E36
u.e+E36u.e+Ma36u21rI2gnMQ2aiE4.8kE4LzvCxP9CHmz9FxIsu89RGoRG6
lOtIQ8bt6TybmBLODu6ilTuuvAnUJHc5Lk4IIDc+iyRSE05mCuRVztWVdnPo
23xz8cLJW33bJ0uo8seiLYGTsHAw2doIxNde9Q9Nd8z+xlnm1kb70nSwKPoN
LyZGF0ZGlIRylLC2R02Y0WxYzhFGYseFbM0+yHWKQumn67v1tPKAdM5gWOpq
dG6rMuP137htCWFpzjQaY5Br6fSdLQ8qOpwjZD5jQJIvLxxAymDlwDpzDgBO
ynDx.OYF3Iy2D7jYig4DruQ4DNs1z.4FJtgl+6dDHSJMBh284kahVuM5k8Qq
2GsKV6op3UQJTI9ZFmS2G4Lxk6dDN4RBiyvHKUdG2rCrUhBaNGRSMfpp8keP
K01UDzeof1r5PGkMsX6kMnl+9Fhd4tGtVcDXyDJPL0qGTpEguQvU2hBnS97U
iV3+5SR.a5QVtdN5sqfflh4adeC46LwiY0SEe1ZG5wLJG69quZzdk1k6iTZu
oSACcMmvnAv9nbcCrpD.5BAjh9Bjl0sWTbboF0x.xFAO3kJzVEdF4WzIL7y3
d1FlxvWj4ab3UN2fG01fuwBHaJCrVeiDvoLrZ6vXwT0F7MxdwMDPjS+AU3mO
RtlGmwZqGmAF2yiyflUipQn9ddFZObeENRCXW6vPF7MvQZfj8seeidjFHQWO
RirN7s1QZ.kizQZj2+eaejFH3HcjF4vU3HMBGoQ3HMBGoQ3HMBGoQ3HMBGoQ
3HMBGoQ3HMBGoQ3HMBGoQ3HMBGoQ3HMBGoQ3HMtbYnAryAuOULCRPiNdvMPR
+5kyfyswoWa+G2b7GS9CKszaY8bHMy2r0X7lxIKEE.kTFmADBcQRKM8P0D.E
RhnRHvT7D.XH.KgTDRRvxxIxrCW85lcW90jubwFlnSZR4YE.ZFRHShubQK.A
u2F7d6aBu21j0y1ozM+bSItESWvXcQDDyI4DlmTer2zELccYggQBHhZSnZWs
p4h4qg5jule9FC1CfhvpBTCuvZe47z0wGeLd22h1+ppSta4l8Ou703OpkKs2
kAx8F8jTplz5kzb54hjV9WEF4orNVJscG0AWTuZkCfo0WzlANZeANAj3F3.7
aafKuNs1L1wFUrCYbZMDIuswtxqZU1.9hKmqJF2EsHhttqI.Rf.oK6rZ+pRg
uEVA2DHJG2Ev1.Q.7Mxp4FvQNXbWLCj0wQD4MzJaWpw.F0E0HITqbCCJfLBf
8l.FSWS65nmG2kyVfON8swhYW.nXTAPHmUE.kyYcapyU5VCI2ZR+5spLTyQL
oPnJ1ajEotsvVBWl4NGh.tltzaCGXxzJeO.kwn6Pbp9uxJv.BMTytNGuxWSf
B160UT0bfh.EHkM.twAJO77AD1aeefg35hhHb7squOzTvVDwI2A4dWORKSt.
4mlPYngRE2LXyqwwepKN6Ph667J6KEUanAKAdYNGelQBaMu0ll9TbgWndGho
X9BLmJwfb9pAaSzUVNJb41pqQnX4ur9kitLyQv6KVnTDRBEHpfiE59Mhn20G
TUjDb9rqey9ak2aOeYW5LFe65v0CQTWnDdTQoae8iXthAyda2qYKracjw4lX
NQrF3ip11FSX3n9ZK4fLzsE6gZPs1cSOQLppRhkrpHG3lD47xO8XZumnYVZV
GsP2XrT6iaNtd0h6O9XJN8snOXBFLK9pWcozUssnNpODNGhNplLhTCCTrZP.
f4JKFvZZAAQjdPpbnVCp02eIOFXGiEmc0e9XfC4BkvdrCSpji5911vdnIC70
i.27XeoSHHG9cKbwqQ.1vTHvJjaRD0aSHWYzxqI6icH7I5CEB9TWxk1c3Qct
nGAh9fpKm958Q7NOxfFW4RXzBQYWDiHHKVH71XTyvYokFspH4JeTBMnQI13J
.CKpMHAdKsX5DyxVZjotbsQZIz.ksYYICjLOGMZxWKmnp2ePIwZW7xzW7o3u
jNP7CQe53tCI6Wqe4tjMIO9n5EOtY4grK+vNcH+0Qkpff9anIVTW5EDW+.VM
6xnrtZN3OqSSr+gnC6NF+Z7RWIqmr291hARSBOBoFT.wyHW6YD8pl3jr8w0q
h29fqCgW1eeagDKT3fPffXffJQBtcfIydwKGvzDoaWPt1eM78QMv04XlWofr
SIWpkel3GGe6D1omz4snRUm243P59EOE8gGiWpC.+8UUL6wGOrW0zUelbRMu
3dGzxpS4ytGX7np.FjWSD1aHEvxk+YY.6Kq2tJ4K5r.Hcb4COm7Z5enGfNMX
dVwfO8kXcb1quU0t+4BSKqIPSilDGilCTY.aCeY4MC5Vb.KUJsskWVVFcZTJ
Sq5kadJY25CO+RYkm6ynBXT0e1xfDTNSGjZJSpw8MSpMyUIpem9Spm.IRcBX
gMNIDLT+z6Hspm7pOSiUmG5vVayf0BKHePKvPwhIIOxkygj1k6aFTmF9ISWy
varPLCxncs2PmrFg9j+7KOpmxBL.Q5KVnC1lH7T0L7sXKvoyfL7eJWgf8cxI
t9RojcqLIkK3J2zftaZnqbSC3toAmllFsSqvlr4U9N4lLosBek+Oo6BYY94U
nTd.lA0zDeGNvSYi.46JDKygmXQdd2zfhKsbETmDqflnwNnu5OvwS3LHnupO
HmzVguaBkq52zzJ7chAZJUkBB617yopvQcZS0qZY7x68aXMTyjJUMGbXa54h
AxwUqS94zRxwG+CwaO5OCd+35MaNUeYdWsOw6tK2CDYUgl2W3RkpKMHIABE5
hOCFh4nTd.GqdAklNbW69f42HgRj.j9iSXHNIs90PDHLMkt5AVtUz4moldwS
eR.o.PLuxJii+t6Vt8orxTB+bs1vL9tKQWF+xqYKKvxR22wCIOsa4p0Y9W3z
VOmGulFFY+mN9vxoXf7TmogwxNfqkGMEmFCEb.UHLuRPvNu0SilLp5go+7L8
ObyqLeaWjAyBqWW9f9PaKgIXDkjxQ8XRd6TO6N0smUbS08OU5Nk.Bjan2dIk
jQu8DjjmBSTVEm5U3NIbJDm9oPRrvTZmPTHfY4gFuI9kpsYtPgilAC.THLuR
07Qn52ewhKjR9z+xqwai94ka2G8ywur99jMqJ67w8pm2CZTt7CDoZaoXSZth
jt.V..PR8mm1k3o26GWuUWtnhOM4CKQPloqlKY.UbFzjtn6S62rVoH2+ltRS
46huBPPZvtU5WSb68kk+BqPorpslpkI1lknm.cd1J9FmTWoaplE62Xa5mM+W
WFvohe98djTs5SsGDRVbgtdhHgBuTM8ZGl8kXOfBRxykBW+UsHImKz6e+dqu
x9sNa1Qe5lEWekZ2mQT8TX5vr4z1aAcsTnKVdIlIaNMze2liwECHrNKlBC4L
hQofScBo9sfWvNwtjurcP8hysc5IYJ3Kbu3G+5xg0I.HkdLlMK3XLxHY.AgP
7kqS762EGOvdQ5nftlngMJrnUc8x0A92i2rI4K92CJV4H+io0NRpD.Dk1201
t6J8E4F4uPQZ0BUcezS5TOscziube7teuR2wAMPoWd.RmtQUJoPS6Mvh6ILo
chcFEB08B+2HtaseWp+yHLFVls5hmY3fZrSdQVmcteOsidt58R.GBR+3RLgA
jYlMwgWj4tGV9ZWVYVvxoeZ4gjneRGCU2MFswKQ0y95UWreMJ4wG2eRklRIg
tc19rIRtihW.QLj.lk8457qFw09S.qj9gol7igH7j86BkDYcuFgXF1xI8eHY
nGADJIxgRhbnjH2yXQ2H0Khg6Zlcz6HSGAxXBO1YlAFSmKQltZOrCIkJMKMP
nr0x689F9gvRbdCI+2Cj4atP44w+ys6SCJ3COu7PjVjz9mSN7snCK+T5aFGo
WKGseoZtcrZW1z25yptdxtHkBJK+rNijxYfm86OFueQWiP+dmgjHnHO2gzQH
LHm9gt9UCF0tUJw5qJqURsxcRuSaZH.mSnEb0FpPhfkEpzuORb6TIX9dmjLm
.zW4SYPCmqLwgiAbBlXJJp1X4j4bMNY4Jih.Q6W+zmcVUng8EmvzyKcDCtHS
VlpyuDwJNdnqcX7ETNTpV9vQBJjpW6bCGu35YBkOfbGQKfXJCiDgugmFeJiW
PtrCAIG7RFYXd2xvW5.Cy2.8T.upDz00zECOu90kmJEvkTfg2UpZRZpWs3zi
8jg6qiDTaMrUYp5uFcmfjXzylQM0jSlwPDZvcBA2IDbmP+T4bSxxUZuMDI.c
laXa0DYC61HxTNeVYhRSlGqjcdHdW7pnk5XiLR0R5pqV5sUdYaRH0GsKPxUZ
rJYBsmmYyoRlgAWVusqvRuqeFnL820FyYpJXyDjvnePzCGOnL6uqvQuy8UTF
E9PNS9xyD7fb34nTsViztkKd2WR1odmmUaNTPOJuAn9WWkElxDrNSy09LQbg
gnFqvq2G83nSV.YxZEfETg.KXTLkQHTsKP3Cltmxyh9KbA3LyeZOtK9+2w3s
O7UWfV+8kHvb7TpIFDfDnCZOAVJHXgtNrHucJ.Ko6WoTLQYZPzGLK+HQePot
qR4v8wmV9E8AkJbIaRd5qEWc5pl4yG3bQnfsPInhwoHIJaWLLavk0lUqSic3
k695kiKFVt+gkqh+lKjh1WGxYLjAoYxDJ.ivDQg5kKXnH0IKEuTk9BZeq8EY
3vIOwMrpcQEKjq4zNaZub4K7E8mGqxPKITmmDE0QbPtxrYkduBN2rio54UNk
rq+AKGQRoDsMTPw.DGSxodarDHH.0al9N.BDfwX.Eg3bZ8.JbrSxaebcLF0R
GSvPDFfHIpVOwzM.bHBvEB0jSrHMhUUcZBl.EBAGQ4rIum40PFu4dVSgp8X1
dQiT6c1MRXoA2vhG3sTOyWwBDPHMMCooYHMMCooYHMMCooYHMMCooYHMMCoo
YHMMCooYHMMCooYHMMCooYHMMCoo4a7XnbW7CI6VYIFJw1igRTqwPIGZB.vP
LT14XnTIZI8PhLgDgvjYWLINDCkgXnLDCk8LkLg.fSZ.Wza9.xjnNTCS.QA2
LgihNuA2qlot2U.SL.hqzvHnl.LgO3TXJaefoIzlZZBCg.cOio2gSBVHLzVM
eNNiYP4oauCnVDMkI.wDLEPEYkwADiMWSa2lYgrjWuqq4wbKyYHFUpvlHJ0j
bJ2FRYh2DAblLon9xH2DAqPZ5jkvIMiGXehcxKOSDA6azXQjYo6OdbhGqNR8
PnKy7GchOz.O9Q5Moioi6wzfwxvdUCVJ73KioAT4uQY3sq4S.rWBYbXAfK4y
5bxUgVDYpXZHRJxq+NcN2sca9fUf8xFNyq1+p1jB2QQpm7tZGgWaa+kO0zmX
wkze7Umb3mTs58m96KYB0+Zxmh+VNC40ctLEO.bmrfKgLBTGlKYoZNgS0bbG
RfHb.GfSmka3B2VFE.MOJbY.y+1uE0ndqxIQ9fNJxU.IhJ4LZJu.RTFFQ0Tj
GfB4TtQXQmCr7NNYdJBl5lB4bclw7MkBNMh3BwDf346eIgogoCkozG.lQIiR
5aJT1GwCT1DHd.Z7c4sf.gFi3eRe0wLC.flchl9X9+5Fq+3APW1jph+nDV5j
FJTRw+pgxKkzIj6W7mZCmzBoRW4lnBevWVu50DklUYCKPoohKpyaVgV6uL1r
Ve98HkPbdkcTq+4tj0gDu63zFJdKnqaSyRlhLw0UFluE+mIbgCwW9JRHlArw
4j1H7NuSjW7YJd2znrqA2zAu1CcnY.sKm5cE+k9CaR5OxbxHXkljPsF7zLdV
WH0IAurpzeKetKoHV+64Bgaw+3qbaicoWT6eSi5F0fSTSC34x5SchqJY5gFq
jzJDp9gP0ODp9gP0e9Fp9gfdbbC5wunZ1Olr6k7S04t6O93iEdXk8sXoSnN+
3dRiPsxIBQV5BndkNo3RW2QJkprk8gD0c.SRwFGrgXUoxBrRIQJ37OP8yQPc
FLkLWAS4oy2ogWrY89CEN86wIxRMP8I+2pCWMfMflzXk5ztiacgk5hsgNeGy
fLc0HSnV9J4TZFxlltnv1BMUTgmtU+i2.NW3.gGqPzUcgHcoB5Pzp3Clr9zB
RJ6ZIOEpsrnBqpjE7tPTH5c6dz6BMDpBxT5SElE2bczLEhd2Pz6Fhd2oI3C6
ccmWscPpa3ky0nMb+CpUNthHETuIZ.jpGmVzRyVRyMATVsSnZmpCk0499TqF
h3fKVMZrw9Nr2jkCwDIkTLy+tNMy.lYQOu+UyNSNkb61y6cPzRyhmXTWmt+8
vKWe2jqlk5711NuJnLEA0CGapbsLntjvP4XNFPL0vNJyZzOnUep5yzcf9qzP
r7dV8OTTllbCP8UpZ88b6HXuCMUpgnfvYEPe6glp1b1kGJoaa2IFlweqpAAY
8N8AHlCRGaJ+12VHVSrvjKYfTvTDQYfzH4EqVtKjXjx74yUDVHj7lJjxRyo5
uEQU8H.0UrkymhPJKCj0eRM+fwkPoDASAYA3sHFqCMxze4BlESQf8BHNgYH.
8lBmWdHY08tjTPX8mMtLA1uBBQp89QjL3ivliU73FvG0ma8KGe4aZs4bgR39
SpcFTRsyCDSkmqLzBwMFLUgEk5Pl13GDAUSjzZNhGMbo5iM14Y34vIitbzXk
a2lCGc3zQ2NdrcmO5xAjUOjs1bDYEmQ5AcL4zgjd3TRebLYqNmzSGT1fSJ8y
Qks3rxVcXYqNsrEGW1tyKa0Ald3DSebjYWblYCNzrUmZ1riMa14lM6fyFcxo
KGcZ2YmNb3oWN8ztiOqJAop74ZWuUxaorv5FKC7mh9aaD2RS1b0nNBsTrW7k
KjZx8Ik1jq7FcSMfZgsA5DLxGJL1N2IMSQNWbqTmfO5PgOng+bxvO9MD94MW
L0I.EMT.0KtY5JAo0zhsLhVk6lZLen52rstxiSM6polqIAUcr40CXahqmpAx
CdFXO38olMkn0Rgw0CYGMtgplWn4izb8NvST8YTvJeQc8FOrvkTMdBbMZTry
Shqy7JUeP1ZtQXBQ0pYgZiYXYiYiZKXVCbP0fvnJevV3iJGYi40DfECEfyoC
1QFgaFaaTGx5kblIDfqkfv07CFXnBT6Hwf0bRBWCJa2xPulvV1kI0Rk35vrs
r9vBXWOq2pk5lMk9lDqe3QmFh5zDrd2yocnmCodzyGN6kcY530X0rlFxQukF
xIcYHm2dO2cZUbY5OnQt+byLR5ny3KyocK2y6hXqpLql0cZrmJQsjpIUOhmd
4I3lS8DaoeR4NvkgZkWcb2xBw5uyTToyGwKBXNSHFTYgMWH4RR9gjKuc3U4W
1eHIsX15LhAk8mijM0nUHbg1r2bGPnAnYdYssA75o3seK5C4SpzSB.ctbk5W
rCvT3FRvnLv4pZGDQBG.7Db.vq1+5B0HqiCAla7lqw4jRSvbQIfvY.GNC3vY
.OaOC3UwUBqiFhF7pxtYcxMEMrn04AtYbKLLUjBKySvXZmcUTok5+29CpEbo
B0V8ObWe7ojI9yu6+0ZkRA69ZjYc5uQMi42928m18m19upCF4SJT8aLksQUC
WAM+8PyGQ+eq29vtTV7L5uOUC6e3zsb9i7Xzuw7sG8e6uWoh6+k+KQO7rNyA
9Mq2B+sQ+8pa729ej9gOj8wTui4tO7ep+m3M6iiJ7s7ODA9sQUuirW78Q+lS
Moeawuj+zA8uUHCT8gy+lRe1+c2c88h95skp720zKav9pLaNH5ruJeybhPpV
Uif2vOCSC3oyLZM5w8D8.yAmFx5hInsZA5DXibGZeLOrPtzVkqVqyi9OVkbi
eegeU1q.ip8p0CmghQ7cxWhh2943MJklJ9IbE0BklP2+puYt..1BZ0f8lZB2
d7EK+sZBedToYU+wGR+4ikaC3Qa6gVr2gjnkeNY8pnG2rT8oWEs7vgkO7o8c
Ewv8NifQFwgL.bACIvTlDmWPNxxzn4Ahs+gmeY8AEhoLF3Ik0ucDfP7dCPYo
NLgBU+KEhX4U41rIayC7Y486S1b7Pbz1j0JssdbSRR2AIZ+cuFLM2.TPhDoz
zDwEXTV4ilMmlFkc7qp0cJULS9bbz+ierynT+8wFSrvDxC7ETJHqPwmwpIWN
7oAGlorLVI9I5wRAUVoNOaB3Dhbbg.UKw3PIAAwBF2H2lySW90A2O1KV23h6
axCOuKd+yeSMYHBRcRAGxoKw7rkryPLaZo3jKbl9rKQKRz.0Q4a5lJaLKEfh
Ryi6tRTdrdi8XpHMHSUXOEHnjSmBAxqzNk0ejOsZWU8+tXYbkwmRK28jNt9L
CGoS7WD8Ayvw2yAQeHaLAptzGzCLlR.kqZXObBDEkO9TRBMlYOC2pVxKqb9z
SgvpweTqocK+rRkhWhiVsVWWyTeHUqKpyaWNEjHDxnSlRc0E3pIvM2K0xlC7
WiRmMWGjmTNE6ylkdhLzBLWvXH0ajU58t8RS3FYZNduoAjr8FA+ZfEPfhgdp
mRRMqFuk49CeocCBaBKRz9y8GxIjHEzwpHxGnndweezB1G2ET9LB2EQvKvHD
JqtwCyxD8hu6DBOXefG3UAdLkUGnjWQKSIpw22XEOzj35o+0DBedUHzqWE68
WoqJfRlVbXo5eTHIHmbA35HCMyJSKWbBQ.uJH8xlPfIaBDfapSfjEBFfouqy
g0B0j4g0u1DAUTfuyVfWZNOv6llQX8EkISHdShQfKMeL3cSCdwolFBuKn1kc
.02lF7hy8GDluZNMkzqAwWBHwx7up7UFLKW+U6Tx0YY.DQ.HbZFYl6pl5W6R
xRM92Y4tmMfutMMH4huFx6lF9huFxW1gI2LkqJ0kAmTpKC1ApjajTOjCSU4A
RTpNijLFFbhcpMJNV6JSUm2WB9CNkLPFaVXEJh4MuMIMFrNMMCP2HjsoPY8o
gPpjdhuVPfYdOyWkkHS4xHHrSKiPSFW9IGLiB9Fgebbv3Hwqd5TIPnzoWf5J
EiXRmHBhWkeaHHOofEmkGkQmmUzg8+9mSNXqmygV64nFniFgoFZKp1ysTRc8
kZUJdl7iSe1D5x15wLb+FqwBY0dLc3bIy30iWEuY4W+VDEhz+msI43tNTm0w
o3pcbIbF0w2u7YqqoY8r6Bg0lYCrvmPW3Ne7eY0xGr0OIcsapoSpZcFi3LZ9
uczy.8ngmmoI4MyOuN9KoYQz5MqO70R30iOpiC.CQEjJK46Kb7t2YxjqU6Ta
ObJqZxoMqG1r9gOc34cIGe54huesjzo7E9b8Kb+S407+SuyYKBydGc.NjMj9
Ach3kJ0d2w8OGuewx0Op+yOrII4UsVW5W+4jME2X1VY1y5wW5dRJCaJkwo+g
RvzfIypSaCuJ9kDklEaWsPMJpd920ig7WWtMdikopvSklbuk8hU6mpSRUJNy
U8ZdNiq8IiR7aVIFTWJ2bMgM6cqNlZi0HuqR.nSzbsFQnzzESvrzHQWy.aDL
UfzLYLyTOJDRrDRYTLBIDkr+rFCTRGSrr65rvxqwXDy4+X1HGcCBdoenxmIr
AKqpxbFhZ+LfOcvmt7.SI+Nb2yqWsJKyIsNN1omMtsmMe5d1f1d1zo6YCa6Y
SllmMzi9MDWfQl6+iB30iBNFOJOdR49VYPOHA2imDmOBcIA6h8jv97jfi.5w
8A8fnxsmJ9sjQ3KfPnDKH773ffaNGefIVF4mNG+92P8YUBk1X6zryLgjxNoB
FleNQJHtt1ourxGKvz1Ov3d1ODTf28Cvz2OJe5JmezjKvituPHRP7FBQWf9Q
4vunxAm0+mcWV1OnmDyGYYnwX2MeDOmeb9C5AQ84IQIiwSh5yShMVOo1z.hN
FCSThuOInqkuTi8ALCqBxMb4T4+hZBnLtgBNR+qg1tQ9LVLF5NQ8Qv.YLTzv
moWigP.hOxaHixShMbEZvT0rkpF1O9pzPH9.Jfw.T7QFLYLVbS.dr3FKaZws
fAKrbFBvzz.8t7eRyiWT.XrVeS7x3HPYoWELBzQuwnQVdyGZpBbU9SrfalaA
MLol9OGaE17r6AbYi63pFL1G6CwhQXFY4ZdqqmzXrEJl6qMuC9I4iPN7XnpC
1qcHFCUcvdIiZzdRdIiZvOIuVyUNrqFsUX98rA1e1CSViWp7OB3KxKoHigJY
HejhfFi4LHejhfGCk+PLe8xjSMkxzFReRHUNWRiE10tvXOK2q9PEs8FoI4He
lkyFCcZQnNLPMrmjWtjXTll6yCZDdNPubD8XnnAzOKOFimjOaJW43TlLM4f9
33.DrIM94EslGQJpue1ekotOxDgrig19dggFs8A9qsuQp2Ys6YkU12TGZDox
LkFCYR+qwVzDjH7Zy+wPCQHwmkxD1n7nfC23dBBtXRNgBneVX0bqiZr.bJZc
9ra0nLH4q7fA+j7xHRHYLNmLHV3s5zC+Y4k24GEKTfdYJIbTrDG5kwjvQwZR
HlL7EiBFdwzbZlPur1sklmTsMHAxED1IVMZJZoHuGzF9De+NYO7n7r.d+rF9
jQedTrw5IAaUqlw3I4k4OROLdDYJJtPot9OVI1OgECetwR4HuZ4jo3TW8+QC
G6Ca1uGcEmOLVlHf7cVoSSDPxh9MmZnd3x+UlMBDAZzrQ.50l9ixYJCgdsoO
bT1H1OSUGomEc36zwoxoOBifdohbptYtmlBSMpCwLMNB4zeMzFW5AyzZTghm
jHTz7vaaArPLgO71hIUBcBe3suRYhd39DvA0ViZB.4JDsi9YTgXcpPnN0IRG
2DnSUhyIMk7LD5Qkfd9bIZSyPl+bJmI7w+P71il.DOmCZJiD2+ziq2r4TLaW
LY9xSSfrb+LOu7JEe2HIQI6RGD2XkoTlv4Fi0EIlhYZp4dfmBJbJQBRSQTBC
wIzzWIPXJNMsQqbanyOK.TZdB.ovv5iX8aAKmt2VCL85ARNVd5ye7PxS6VtZ
cVgICTJgCee1DIcIQrJOpb2KJ.cc98Tbh2oAie53CKG5.voFmiw.OwkxiBhS
XufCnBg4UBB15scZTfQUOD8mko+gadk4aZRFDxS1fGdPcgR8aLhRRyn.LIuM
om4QjkSffp7I.PYGmIQDjTBOsCHHHIOEJnrS2W47XfSUlT+9TzFqtwzWQg.V
kGV7lTtLoz8pK9hlucA.JDlWoZxHT46sXYVTIw4e403sQ+7xs6i943WVeexl
UmSSm8pmSJuhW9AgTsmTb.x3HyhIA..IkeN5LlK89935sZ4VwmlHgkHHyz0x
WghJNiXTVP7o8aVqjn8uoE20zBip01gR+ZjZKur7WXmnhH6MCKS7LKUNAV7r
UcNmzUo4CJ3DZmiE1I.0QoCWnVZ1Jxql4qjKijEWbomTPnvwtY8R798KeJdZ
jUVPpWtTq5upAodbgd+o2a8U0usq1NVi6rpxqN7aTrXIH0iYWoskB5CnPIrb
LmYYRV3e2liw4zxj2KywPNiX1f6TCTpeK3Dz.2k7ksctEdtcQOs1DOQsve7q
K6dCDfT6wZDRxwXjY0FRYdId7af+9cww8nElhdX8OlMN0pEM9Mt+83MaR9Ry
sthk24+XZAdlJAfSkYLq6DozAgajEAEvz00TEhSG0NvwWtOd2uWo+QmAW8TQ
P5vOUsYIMskBKJ+aTZf6LJTnagMu4Q2Za1TCjQXLrLaVLOS4QEdKG04ym6Oi
OhaqWIAbHH8iJwDFPloRLGNpyiNr701l8WP62eZ4gjneRSEn20mmuwvbKzIq
d+ru6+769+C.Ds3kJ
-----------end_max5_patcher-----------
1 Like

I’m going to have a punt coding this tomorrow (as its late here). This seems like a very good application for FrameLib.

1 Like

Right now I don’t have a FrameLib loudness or pitch descriptor put together but I have many of the other features you are interested in such as the spectral moments, rolloff and crest.

What this patch does it keeps a running tab on the last 512 samples as a moving window. When your onset detector triggers we get a single sample impulse which fl.audiotrigger~ interprets as a scheduler message. The onset is delayed by 512 samples to compensate for the size of our moving window. This goes to fl.source~ (where the moving window is made) which sends out the 512 samples as a frame when it receives a trigger message in its second inlet. One large FFT rather than many small STFTs are computed on this window (it could be made to do STFTs) which then goes to the processors.

I realise you are quite ‘green’ as you say with the paradigm so some of this will pretty new and won’t make sense yet.


----------begin_max5_patcher----------
9367.3oc6ck2aiibk+u69SAgyhfYR51Scer4.ytYAxBrI6AxBDrHInAsDsMy
HQpPQ0GYP7m8sXUTx7nHYwSI41Fy31hTRrd+pW8pW8N+w29latK9yA6uw6e1
6O48l27iu8MuQeorK7l7W+la15+4Ua72qea2DE7o369q27NysRC9bp9x2u41
OEFsN9SO48neTzw6GtVeW0m383iWK5v1vnMAo5uOT9E24mt5wvnG9PRvpTy3
ASo2RemGFhuE7NOIM62Hzs.u+xyeQwGRO9MAyup4RoeYWf4q4la79KY24e71
2l8q24HctMX+d+GBpQnah8Wmcu24k8W24G8fEZEBE8kZgbM8AzDKDxVVpsgY
0zGC2qGqAIVoRpUpD1IUh0jGi2AQhZjHemlPyu8d+Od7F+j+yraUZX6cyuH6
W+j+mreGDs1bkAgR2uIVM.shDjggDPjH6enfZHw8wIa80uWV+wFMa4oupc9I
9aCRCR9PPj+cazuQvzwjbbEgGzJxfFHOhLawOGbQrN3uofujXOH.3889qRC+
XPCDKbXK6YHCyvfW0mOgOcTr+58o9oG16sZ2AqTJXXTpjnmVG0J+vnzgQqqh
2tMP8gqRr+KQ9a9x9f0d+Q8VX1nWv.WfiY5Mwf4qvAMRwfBCpUwahSLeKfak
RFfKHpsA4HJQ9NGuDL6IMc7C67VGrw+KdJHbWPzd+zv3Hq3j8k6flwIJ.ZXF
.l0.csOeoupfjbRImVTxMC2D7wfj8YiummfdyM961U3xuovGIC.9qF7V7tSW
JLxbIvoKkD7wvied1oq5mnHnTE0bHQyidymYGYUx9ZhWGjDcHTOTLWTMUjOj
zfdjRzrZypUlOb1bywaWPSHtV9.xrgITh0uThKvMolVeXS7peHXcwA8MwpIq
vncIA6U78lIsh2dcv89G1j9g6iiR2G920iAX1Wtk6ee9Xz5MynByRojP+MmH
gGRBWGGkMHJMWjc4iON0BE8Ld1uelXzuiH+cV9vJ1LEvzvMMBstyOIapJeqN
zwalFGuo7sN841DbeZ9s2EFEUAESi207MSBe3wV9r2Eqt4119t02Y+GNDYt6
GTbEoe3nNMO+972rIesY4u9O6GEpzTHHMzLEf.mtoY69G2uJIdylRzq4Nezx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.kAqYfATLa5PHWYXh97eO4Qgn
r+uArQLT9EHVaXfbFFI7RDbt6fRU9nFnb9PYJD4msSaGPwweOT59hRh6pMgq
9glVFQGGfYVEgkWhLJaUGKNb+lv0mLat43N4G4Us7wNjPFIOjYGHrzH6kO98
ftLXiteyspiP6+4l3jviD1n3mW6MFVoESTrBPzmtJMI7gGBRZBWF9IXyEFy0
qtDWU.y93CIqBdx661DD8fRcul2qB1H7fchsABMFrjLOvyyFCK6nvMXED8nK
691wMCXj+XNpPmW4A55f8ogQmrM3e5D33AbbZq2CBjqCB7LNJn8CIJNAlr1X
2Y3rLvDNNvPW.finYvALKCLtiCL5bx+53XXN4dcccLYo4dQ8XYE7LCNrkl6k
4J3PZj0I+hG8hjN3GV+AiSQ9fepRk.0wTLaWTzsX8x38OrI9N+M4ll+jqcrX
a+297faZbyn0CYV1Ai81wyDfnfCFqezRW8wdoSJzWx8v16dVQMsaAKNps3es
hDsj2WZV.z5nPTzrRQNFzR3kXyieM3EdEkWxWPMhMuSayxLyFV74Tw4fSni6
2EnVeEkFtIvBqhr2QmDy34FLTqRGizkK6mgXnAb6yC6hzRuihFpL2n0hLN.A
eYiilllv9ONjjFuObuERTv5KIRLyPTM+6Ql7oc5pI53O7CAeRoTtU5PLP5fn
ljxhMFxRRG6RB7WaiJnCkJLd7bQmM9MpKlDGZkNHCkNvhBhNWDg.q8W8jRmH
aa6IP8cGfbiAAo5ixS6LpZlvoi3n.u6uOU6P4L2Lm4.ycIwqTqWBrthA124H
rTuTASLV5RPmgIo5FwqtA7JY.X4nhPLLPdLLmeNvV0O8RQCFHKXuDPJSvPbB
EHE.hN9uPEuVd3e05l528P4u3BQOF.8NmtRwGRGwn6LGCpUMQXoIFw.mXvMX
WvyRHnpnP0ZJazGt2LdGirdyxGLcjgg8zQj68+zSdbNvlh+39Gwo.j3V0rnP
hXbISvoTnZ9jokqCNGGCnojJPupl5QUm4iZi3kCf1UaiUi1gXsFFWHoRQmTc
uilZoZIKjw.DFv7y4inaRQ+vnfmrxd2+PGGBnMwdiICcM8ymjaxio7lo8dqc
iTokbsI5KXR+mYmtwSiHMolvkryhkM5EAS5sgLjx5Sz4zK4BhdW8kUaBdxSH
ru2EcRn67stXWXacgtUKCGYmzY8kz4JcfkvpBwQZRWdAsyE.1FY2aS1wAflH
6bGNd4t2kXRliovqvstjSx77ELoaWRNo2ZjxojaIPqRxYZqTeIuyU+EfyXvl
HWiIYuPH2cgQ+fUJFzaqmaihyYrEmisrJEudk4dGf4z0xfM6.mGxGjAmmlKi
QQJE9ckhyL1.IeC+KVdAo30O2JyKZfEafbykZR93KDJrgn8xVX25tQty8IJE
YOBuNWl3p4n2xVn5aMpsZY1EIzpSYIjsNKz6NO0MJkLT1SpVz.mcgl40tSo1
t4l+5KmaOlL9T8+HL6fwyN60qob6qob6qob6vxRhAGp1Xi8aMAnQc8ut7SSh
8qTqbJEo1IpwP9348l5bBuZpEVbS.3fSEKEnw0tx1rp2DfAP7xkGicP6TiWM
ajzgCO07PZ9EpodCcER4zgR4r7BKzkJkq42eOrMZmLTZmdzBomK98xQ6msH9
qon9a5R.SNNObbf2xn.JGywpSlKxDDPYMjwuV0Ho8LGpZnA1uvCrsPDbQxgZ
3fSMbSLmjE7FVqeWWy4PcqPlbvYctw+TXSsth90RVm+9Fy+QvfyUVf17DX0R
agDiD..Rhjl01PHYbKtcKYIWrkusfr5ni5IOZVjcSa.jI7gljWGA4LvBpjhx
kPoDA0fr.7UFFSHPf9WMAyhgByH.oQXFBPesfy9owquqIIEjAW.SxYhQJTBo
TA.QxQVB6Jq3CndegaOr8oLMGaBkviTdJRsyDDSkBlRlJDQmhLecggocdOFt
y+TXoWChfiDhfF+QaRX5IAWp9Xe13hkgfFMxXSFZrxG2lAGavniMa3wtM9XS
FfrjQHcvPjULFoLuZgjqsoQA9LiZWZBvtAIcvnjtXXxNMNoiFnrEiT5lgJ6v
XkcZvxNMZYGFtraiW1oALcvHltXHy9XLyVLnYmF0rcCa1twMa2.msZjylLzo
cic1fAOcxnm1M7YUIHUkOW69cVUypVm21nNQRv57JkmRFY02ZoZfGpxMa3LY
spiPigSowrpRvsXNPxg4wV3wBSbVfvWZ8WKlOozlbk2natATCNFFMTXjOVXD
YJd6ZmjSfWQHmQqAuUGRiu+9gBezwBe40ajb7ieEgejzG8zoNtWl2OBR9Tbh
5JOpjvWQardAnnwBn37JSmIB3yq07WHPZMsXqz4A7tyqFmXQGIQFM2l4PPpy
xSEBrfQwTFgPwv24wsVUoZ2TTs3moBl877CrJc.UJDs169jf+1gfnUeoMPdz
bfHfw+CJlNBPBHHLUfkBBNKE0.tByvtf4KBj0retRaF0Ya79di3.h22qz4dW
VLdbRbf22qzMLdS7CeonzhVlFf7IhWGJX2hPRFm9rACwL7TMKrNTGAK9Ie4h
X9Xk+9U9qK4CpV8uVqGJtQ+rYhslrPhiBvHL4YqGnqmvSCxVyLByHppI+1DJ
zNlAbFyLIe3oeOYXTk23w8taSCwvn0Aetxw0Oe.rXr.b9x8oFgaGaaUGxS.L
ZA.Xy.uEDl.Fq.0bVXI7VQ4CL4LZCZAJ69jgNwvV1jI0phb0gYakyHKfc8Bh
UsxEVaUjJP4ZaTuXDF7Hj2iQnlbfVX1K1UVDLDgAHRBf.I5TvG.3PDfKDJFB
rvjv7HNASfBgfinbVgtxxhQ4XTenb6vzUJkSlXJ2DqN4+5LQOPGomFl1uRmI
o8c0aWTd1OLnfhUTG1T.MT+fk.AAntnAKHP.FiATDhyo7yBgK5Ag2.+9RLkW
3lE8xQWEmtpt3YPVBt8hUmsBVWYBXl8HW9wsWeHoXOyplodFdmLAX7ITVxWo
lEkbI4n+ykxqlB2618owYc7j8MVz8kCuc7vMUIL3sYG68nAH3VaCc8MbipGX
aKCd8PPzSde+QlpLlfFcK9Hic.lB2PBFkA.TAUpwsInlG+pCfs3.3062cqZl
sAm.yMVy0XbRoINunDvq9.9Ue.+pOfuX8A75fJg0QKQ9cS443H7nlwtuPsLC
VtodwTKVmnkk7ShohJIM3W96UqI0x8V+quol9plHk+l+8PkRAIewyrN8aTbL
e6u3Om7mi9uyBl2SJT8MYgkn9DYJn4WAMukr+OLZURPlIU79UZ8W+tSejmeK
268M4s41e4uRo57O8m5s5wrrT3aBifeq2uR8A+1eT+lSyeapqX9zo+ir+IHq
ijU3a4W6A9Vupeh7+38deyogz2V7K4Oml8aEDBUu4ieS5m8u3lyuUzCiJ0wr
poW1nsUYNKJ5YaU9hwiPpQUqfGYhVeqNplF83NhdfKBiF1ii1xNClLj0GaZ1
43qLrFrNLMalrRM3zzImy+U4ybOomWss9s39MwexKH5iAaTJM0XC4roFXzfa
pS3iB.X2RqFG3TSj3iuL5NmJMqFN9L3SiQAWGvS1YOxD6kF64+w3v0d2uwW8
tW64ml5u5G12WDCO3LFFYDGx.vaYHAlxjXb94WyyDoKCDa+pG2FlpPLS48nu
.DZvMdti4ELgBuMqNBhXRJV645blsKC7w+t8waNjl2jWU7Sww8GjnC27Z5BX
DhlUhkUZZh3BLBXfI1kDaTt6WUq6TpXF+w.u+seSuQogaiMl3VSHOvukRAbS
.PvwKL9zhAyTmLVI9w69RAUVIhmMVhm.Uqi3PIAAwBF2Hbly0qwlljsslVtF
iNdNRq1Vf5zGSB1+3SJlAOXS4kGDHGYh4YKwlgX1szuNRXrj3LQhFn163ltZ
Yi4o.jmNiy6aWZk0gVZlRWGTg8TffRN4pAzXS6TVmcixp++kw7PtMk7SdHKt
9LSGZF+a89dyzw64.uuOeNApt02mMwXp5SE+FKbP.nEOM2f1QCN+0ONSVRfM
lIltJCfC8Tz4cVqscK+nRkhsAdqCu+9fjrNHj+Fuduc4f0o.YT7RoS5s3pIv
M2ttWNOE.tHRm06ZzQdR4HcHEjgtEyELFRcAs571s6wUqT+pwGXigBbi1Zqs
sMaIl.GKrTLD1JM9qErZS.JUKH+JAShw50SIo1oFcBz.iEOlm19Zl5V8oeQN
O8MSWabllgKZtFFt1CQIyZSvUVMZtZpMuBpMbm7P5YZ6.nfA2aRuvILliSYy
K+Kh0qkQm6lrJdNWEggNNkHZiWqCEt2FtdWbXTZtvXto8PAITEymjwvfiwES
t9y0tyLQ7DW4CDy3L.wUg5R4bNJv8rkhub87W2GZKduZ14gFg27PCelGZ0WY
WY8ZdzOfx5Yf7rH9EhH.DVmChGsaQ86YmXQyCw5pDcxb1RwItpWDbw6a1NOz
nfEeMjvUTit3nlnOnFbIQMJnW5v.tHGaH3ROiJbUAT47oZcyBZI.toFMStUv
.rrO0ygLroyYX4dyjzLW48snM8f0LLW+OrT8OJTCPkle3YIORtmfrbyYDAvt
dLW3hyrXJfpvrcdK4JASmFowqa7GKzTBxzuZFgOjq6Iu7vWdJ9Rv2hQHDixz
9eMuh0U7pyD7HuHNDpzY6ZwZTyo7KdL576Jok9wgTa9aOIkplfRlAWOaxHA+
8r9GtstcDq+cVTtoibAzqxHzi+to1qc9U2DuxeyoMl6aSRoTQytZAy1VHSas
Zg22N9CElKWodSRwtCDZnkrUsPX6dQvtI2Irn3mbfvGQvM9Q7qa3iOT3igek
664nHu2vGV7Bg6qoND013LWHs2VSvhMPPCKzphm0VDJ7CDRZsGQQ5FYZ4u9Z
rCRYZoKlMOUJ9oeoDSdsARMtFHE30FH0kXCjJJNCQlt1HUqQAAcvQAgoFHNU
QAQOi+AzxD8M2u41H+HSQD5IuuSGuYlfBt+sAEjSok.N2uWhqlBNP+wnQztN
P4aEfewiRCteUwAvuVvnA2WqHl.c6qALZv8xAJ4RFhZMwxxZQQdw26cWXzop
yZXv9FkGM7X+0vGAoPS+QCcrXAdIjMTG1lgAa8eHJL8fR8kuyWojg4OaDIFb
wyIqfnlWfjOKHQqAwnbDAwnYGGAF4rdN8JvEIKFFL3raKeCkYCBvKFDL3DSJ
e+hYCBPKFDL3z2yrcvrg.vEa6yzD0Q4e5llbxDSJ.XIBRYBFhqcxDz7mTf4V
VuTKoVhDOZ8cElRTD5JqIJoP6fOuSomRVU.w6m4U42eq9ePMkZUrgCaTSBKa
x.DnfdMod25vOFVtVaWBU3ilYxDaAT10DnXXirxE0DPMh87oF2zk2weanAad
ghTJ09Zj2QNV0fnllPv004kxLMXyIQ0nMFAgmW+3tlvDEYdnYLYDM2NyRGcn
4nvD9UEexRqXffLVa77RSuf10HfOb7xXV0iJDPEubTHfiGKSzKF8AZBgFb5E
yMZ.bTQ.D5Ehh.LwHsEvKO8.XxQtL5kmZ.bvvE2xtd0BnMtDxfOSJwTBexEk
bcwkXYCnYVyHBXrd14ZbSsEWAzg2EqOByWuJflUi72Dj3mtrH9f0EgXboByD
vEBx0sL04FlwijwlwuBkSu+ukj1zNWX4v88gYWbI84HX+kwQmPiH9ALXhwzf
WW6xzx4BP3wxj7h6bAH3H4Qd4ct.DXjJAeUdtfMAQOj93SSeYRRjm0LFQIhW
OXPaarC4iLbld8bAtfxrQhxudrfdB3CtZlKL8iDL40SEzMJOxXgL28AW8Fuw
RxJNzvU2rqkkrF6Z0E4Cdabbd7Oht9T.tE7fOttr.hAdo4mfwsfIm+35RDRh
ZOwiwZdXztCode29U9aB7L0029VSe6JFBHxBgcyw3J9JApd7K6hazFLiL2fx
QD4E1wlZK8xXCN8xLz5YJ6xVrpIqz0ZylbNK7fRWK7fhYcT3ZwhRLmk+PmKF
KyYs3T35nPvmyQgqE9IIXFqKMBmqkdf4DKX8YFY1vBWqSoxYkuv0AQ6UyoVK
GXybMwzURPvFAMTs3YRz0LBTd05BgQmd4RW36bgzyB1t4gzElplQKTN77VG6
lMJmfLsM1RuhRx8YBQWDkxd0RVwT4tJSgimQIabWUDhNiB1xBTPmFDr4rPrw
cUOniC24YTzCg7y0Xf4pZPr4TuTlnOKPlsQAsmkn7kaGEWUMZN0IwUzQLmck
CmK9zH3LJ+v8588bpvLA0mBwM7rWR8YyIV.t.ZYLjKgV0BV1GEu6nrspKWhL
joBL.XrSubIE94LEYQOqVoHdgWsjZI6LAYQNSCDjIxEypytGe0RdVO2IHT+l
gHBXGyPnyZ6hANmxzPtJOAOmZHhv84Lcy1nv0cYPyoFht2rnXKce.neCM3RJ
+1U0TrXi7pG0OuOfZBtM0FRjSu7RjfrbLyVIHtwRFsPOvyJ8TmyoUxQ10rC9
rRMndN6PDnNnm4YuHnnOG1n84GooeDfolHzynWWoWwA4QWAVW1cAnkcokyDK
j2OhMO+RK9hbRMKrPamTgm2N63btiFz0ySKmSqBBI8ognMWmjEh6Cu2YuuiJ
gcsBfJMUPaS8hCBPmd4htjF1msKchfxi.V7oWsnqag8YCF2oGJBzA8fNuzCr
yiWanGFRms1RwoWrja+6N4fcjbxqzU3SuvN4PlGxwU85ky51D8ILEb.SOoAO
RdVzf24nwoejSm5uiOqTiEm8zF4vgfyxwQf833HtPM4YG1yuXQW+N71yWayM
4lIawOLh69HaoM.BrOiL3E4HC0O2J9BoUf0X2jY+NEgj3u4CqRxFW0apLTQe
apLByQ9xapLRbsFwi0.dskjl5quFEC0D53.RgRk.Q.dsQw7Zih40FEynhjez
f6SLL3q8IloNUPvlSwdk03JpmGyEKYKhQBI4MgxqvbNdUbjZG2zm772t68+5
020DBwGYtmYxUXH7ETIbASFIlX5bGzWNIYFd3UadDqP0Q8JKEE21XFahGbI2
JGPNlSuWUKa1DG+CG10XlH17xFbORiU1UUkQ2O4A06t+YBOzsjyzTggoWOIm
IE9hI4LmkdkBkNNVBRtOsVjdkx7jdp8z4KySfqx6mU8lmXty4vmcNc+J1Um.
imyTCA6ZPrfIyoCwI8YTLa4c.re4ISWVP2DprO+2Ko4yo8xWqN36XHgTH5VJ
8BiiBDfSuXIs4syDJ1QuvlSaYAvywWcIFxeXbmgkDG7LWXgWbl8bKYNyuKhq
Q0Es4nL5kt2NxL+Y782awMGRhU2bfa1MGTlnfaNpaAqWcyQmt4fYbvAzj4vY
YJhNDcnu5miW8ywq94nyS8oVfs2+g.q1IXWfZ2A0ZiGB79mfSeO3yTmFMlaT
bEYqfAap.CQheg67mVKYbvQZiZHVd8YP11KDZ3wgHxqPKT2NKBZjtwHWEfqG
DYWVyY+Op2VoAHYvMkcnwHrPIaxrHa0G3ypYVl7aTcylT4rxG2lpmMn9Yypf
1sZnMoJZI0QcPkzdG8MMpYpCZm5hFpcpkpiZp1h1ptowZGZs1olqcp8ZGZv1
sVrcpIqCZy5hFs8Qq1VzrsSsaaWC210xscMcaUa2lz30tVuMn4aOz90tFvUE
iTU.cs62kePZS+uVEH2w9TfV7YT6tEoMU9Zv8Hk1Eq7NYiEvpsylCwsPadbr
0886xyiszDZZeutN0A3rhf1KE+0NL1zvYl2+MrWA8ulQw58sjpHHerHHj.5r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.goXcGBwxGE87yDnNJg9Ik09RIl+R
cIXs9JxatwO5gbO.xq3ZpcIw6hSN4tzawxRetCowOj3uNLWOJf0Cz+tBrKIp
2oMeTcyV0DQXwuCaramlT+cGV465D49fMJMFUTPo4DNBvjBSWdkiL3o..fj5
sckReNBmBw5IBjDKjl+hBUeY0+fYGdT+g+PXVR3tO3zjSVumkY9nGmgQbAAW
exoW7gmlKZgUrGrEkYFEmXAEb.UHL+k0AcUlQVVKzM68yx9ga9Ky21hvKV.N
UGhW8FJgI3rIAMPPNNNyVbRj1lLJ8Ik.Bjq+nJNABWSXBBRx0vDkU5ypXC2V
8QyEJ3vfo.nPX9K0n.YoYCWLhBTRI+u1ED48G7i168GB1FdW7l0kNv+Lrn6G
1uITsQ4eLK1ObcwWApEl2Rid9Wy73UcDTVo.Jo8gZiqVasmLYgexrx3zRbd9
BsV4kp.NJlmE.bpX0BmmIURAU6AgjEWojI.iPgK07ZsvrcNlamFQuEDfdT3W
8+pCAnbQ1t9uy5eY+idUoGP8Up8minpMl5A2rd9uftV5189RvIar27+5lCAE
CBl1G411.CxYDytpmHBY1kfKHQjD+onQQEOO1omjofWXp327E+wQD.jRQ.iX
eNFiLqwQPHDubDwuMIHXjTgdV.m8iQC4Lc+VNB3+KXyl3O4NETLfM+S5P1jJ
A.Qm6SqzamajjBEP8d.T0rGcQHzCauKH42pzZaTSTYKO.Z1MpRcCplZfEktO
qDQhQgvLpXPxcYDFCKyWhvyUeVMAHwcp484kveltce1qxAG3PfdPKwDFPlev
ANj5jFlmWpO0eWeVYV3LK+N+zXueWVX+NImU4ERlwd+lXk3.K4+pPZM+Wgsj
+qRtwI6lTAqVIzQMgjr0OsPH8WwAYnlxH1adm2M2ozOrXNwlnfizfjOXh84S
yM8Mwf0hBOFuA5TGn3HwRj1WJdDf8sTnlmivDH4YWkUAirE++M.T6CenTDc2
XfJ7N8DcF1W74TIEAVb3ih6K7wDXkBALfRrk9GTlAPHlR3uobLdMClcVRdue
ieZj5vm1pJunIAKySdcHP9Z1qOIEoWFf8Zxq+Zxq+ZxqOtrzdjoo8K9hz61.
+nFyAWvvS3z7ZIpIsjuppSjNV6YoiLhR4vqvRO6xVRmwu7JoyiLNaIX90ZbH
2dUKNqMmNpz82bDxWPUsXBcr0HCx0WIg3gfX0gySBW011RDz3B817ckf.wUT
koE7Zkos0JSKbjUlVA3JObrIrKfhBKA0yNt8YMzEmydjJw0tHEAszQ1oykOX
5bxpPuDJevv91F1Ou7JVFtybvZScFfnWBAqsdT.eIWPROZRkiFs8igAeRaxq
vMgoeojAVu+98A4ax8dsVfuuPdybiwbhqSTioSlI5nArWsIb0Oj9XR7gGdr3
0qYwox23i0uwcObzz9mtxybrGsO7QfZs5vjp49n02pHYEcdjF8SdHeSRzogu
kShaUsHaa4m2XbLMyTDTa70rNAyfsXcg4Q82Q480MSqU44ymbqWKwKxp2TYK
tPF1Wi+w0GDykGTo2SYYMC9AKIN7jYvofDynQTGOoRUdgpYlFA.Mg+a9ObBJ
yuQljajX5iWkekIU0vlhrFwjpZimHfcPD71HhrdHNt3OYAmTqDgwgDGIBpbZ
HBbGDQ4zjp1TQVt.V3GXVTu2FUPA7BTg9USAUzEmKr0ohbHEBKBvP3nGZTjC
CMJtswFShx30q3iTfoqZ238F43V3hrfm2TevOGLxwGjEwcv4+IilCAsPLvE9
ULdBD0BwPmeVvQ+rPN8rHSxyh57yZ7XnSxWvhI4Ywc5YImjmkKqxgDvjLeIc
QuF5T7nHtvFVzRGCWFBvY.bxkh3zytxpjIRzozM4WSgphR2DeglfGEwIMumh
mjSxtljmTUdu47QgIK2iBuTOp9LSMwqukNb9BJpzap1ALLpyZ5RNbgEcEwE6
cSSlpUtHquhlw1UHOenaUM2y2Pux4MZXnyJdhS8qpOXwyOKRkyo1JKh0yx0.
NOtgN2EUO3Kk8TpLZlHdIWjcHJu.doexVnYz7yRJItyRRknSuZVFr.WFrNJm
hx45Fo4bLXktnUqXJVwHbFVPiUQHm1nP39Df.BaYBXI1PV3NqsPPmsAK0E4q
R9TnLFvEULmhyF6j0E.foflfN8rfvI4Y4zQbfSCc4zI2fShUSfNYcAHs1yx3
qlJINP1ioRBCTIYApmn.Mmj.USP.s6GqTBfLjaGk9mJYr1O91lx0+iNtzdY9
YHk3mAVdeFXo8wd572bp72PZ7ez4pu6sMmgdkyNuh7dVJYO1m.FSY5YPknmQ
WddbhmoZcgXXUmgAVJdFVY3Yx3aN5z4gVxcFZ41YLkZGmKyNidAQ8xoi8EFN
TBcF8XoTA3v9vn2kPkgUZbbqr3LIDbgnDsSj2sRdyjLrJUZalp4hgKpp2kul
AV5ZFnfwkXOtxqNbaVrXL15.2UWkflQyYUuTyXeTMzxKyTM.KVFYbdD5ToiY
hFgEKQLNO.cprvLQCvRk+E2GgcTxWlnAW4R6h8QWGkykAVJWFOATqjs3L35T
85XzCvZkiEGj63XIXYPkekQSP0KyJNPQNVZUNeSSmJeJMy82VISoeOeyAyaH
84e6+3s++Ort+rC
-----------end_max5_patcher-----------
1 Like

Thanks for taking a look at this James. Really nice work taking all this on from first princples! My build isn’t present as I type, so I’m just working from the code.

Mostly this looks good, although I think there are possibly a few small errors - most notably the weirdness with the spectral crest taking the maximum real amplitude, rather than the maximum amplitude. It would be more efficient todo fl.hypot~ once before all of the stats (which are all on the amplitude spectrum anyway.

However, my main efficiency comment is that you’ve made your life difficult for yourself by not using:

fl.crest~
fl.flatness~
fl.accumpoint~ [which you did use]
fl.centroid~
fl.spread~
fl.skew~
fl.kurtosis~

You can run these on the amplitude spectrum to do most of the work, and then you simply need to convert into relevant units in all but two of the cases.

Finally, you haven’t taken into account the natural delay of fl.source~ - which is the length of the window, so in this case you get the 512 samples starting 512 samples after the onset, which is not quite what Rod needs - to correct simply remove the delay~ object.

1 Like

Sorry - I’m totally wrong about the last thing (the delay) - James has it correct. We get the samples when we ask for them, but they are for the previous 512 samples (that was what I was getting confused about), so a delay of 512 samples before we request the samples is indeed necessary.

Also worth noting that James caches the spectral moments by doing them manually which might save CPU (although it takes a lot more objects). It’s possible I might add fl.moments~ at some point in order to do this. The descriptors~ object does a lot of caching of calculations which is why it is as efficient as it is, but on the other hand you do end up with a rather monolithic object.

1 Like

Most definitely the case!

That is really handy. I was going to compare this fl.source~ approach to the JIT double-buffered count~ + poke~ combo I nabbed from the @tremblapenator but the fl.objects~ don’t appear to work via Max CPU stuff, making View/Show CPU Usage useless.

I imagine that the fl.stuff~ is faster and more efficient since it uses less objects, but that’s a superficial reading. If it is (much) cheaper, that’s worth using on its own, as I use that JIT bit of code a bunch.

Hmm. Is a single large FFT desirable here? Like if I have a sharp onset, that is at the very start of the window, won’t hann-ing it swallow most of the energy of that? I guess the window could get offset by half of its size so the “onset” would be in the center of the actual analysis window, but then you end up with the problem of having “bollocks” (to quote @a.harker) past the desired analysis window.

If you do the many/small STFTs here, how are edge cases handled in the fl.universe~? Is the first frame the first actual frame of analysis (with part of it being swallowed by window-ing, or does it go a bit before and zeropad etc…).

Or rather, what is better for analyzing the information at the very start of the analysis window?

I assume this is possible from the underlying fl. lego blocks or would it require having some other stuff involved?

I picture this every time I read “spectral moments”.

edit:

Oh, on the CPU usage, even though the objects don’t light up in the Max interface, they still appear to report back to Max?

average audio CPU usage was 4.33%
msp64plus: 0.08% of total CPU
fl.trace~: 0.12% of total CPU
msp64plus: 0.12% of total CPU
msp64plus: 0.13% of total CPU
msp64plus: 0.13% of total CPU
fl.trace~: 0.13% of total CPU
fl.trace~: 0.13% of total CPU

unsynced.fl.audiotrigger~: 4.22% of total CPU
record~: 4.96% of total CPU
unsynced.fl.source~: 19.58% of total CPU

I imagine they don’t highlight because what you see with FrameLib as a patching box is somewhat magical. I can’t comment on anything specific and @a.harker would know better as he understands the complexities of this aspect more than I do. The CPU readings are fairly accurate though as far as I can tell.


----------begin_max5_patcher----------
9353.3oc68s2iiibjm+ca.+cfn1CKl1t6ZxWLeb97hY2YwsGv4cuGyBXrv9P
CVRrphdjHKKQ0O7fs9reIyjRhuyfTjRTcqFyTUIRIwL9kQFYDQFO9ke8u5M2
8PxmC2dm2+Uu+j2adyunuxaLWK6JuY+EdycqC97hUAaMuw6VjrdcXb5cuK+l
ogeN0biebS31iWNZo4hIO7WdOCe3puDjt34n3m9vlvEo1mK2mbO5cd9T089u
ySvxdAAcOx6+29OT7t0QwqBSMOebgqlrKc+kQlq9e9q+UY+V+q2c5zy+8UAo
wga258cKe3sGo.MUpe6AoQIweXUTb3hjcwlO.oIRm.gzw9Y+RRtWg4HFGY+m
Odl.D+eSVk73iMNyRAPdLEJalkiO2yrwgeROHqSOu3snDqp8gk9kWBsi661F
8Tbvp6NNNKRxntHYgO6dV0IQFUlA.LyrLgzW..W54EtImj2Syu4tGiVE9wvM
a0bjEe+u4tfWdov0eSwOTFT8WRLeWx2c7ZQw1qgNdsMgeLZ+WA+3kC1no8TM
guaiA1t6yb1cE9lRVFtIdWjY.ke0rIu8CMyrTbv5vsuDrv9EjMad39GgTTFp
Y9AmY9ESVjGRyE7zpjE+b3xRi66RdILNJt3p0x2eY3iA6Vk9gGShS2F82LCA
bF+ZSugGyGiMe2LxvPA+iahzrMGdOOsIZYRb1.o7zR102+H+Sd1E+Y+r.MYd
KwAuzzGWyEpAm1t6VMwta6CAaxl0dXUXQIS50.IIqJeuiexUgOlle+WhhiqB
noIuzwc2D8zyc8oeHQe20c90at01OrK1d6On4PR+v1fOVA4SCVsJekckmvmC
hiVGjFlFYmOHni2MLNPSvOucwljUqJS21a8wlt0RMu+hvOEsL8YyCqD6g9CD
8xdFq6NNsuL5oLoKkuXZvSaqbosoewNET7Z6dHeU9GRCW+hdGnpuiR6XWZEc
Q4gkuQmxEKKa7wU2aDN9ZoaWSB4QYiUkOhok9jMHiz2rJFiP189pITrcAiMI
b7vcx2gXufloBchCh0xaSC27p22mIjyaS35DMO5PgKAP3hKM+hIAfWjYDdkt
QK6zA2T08aqsmqDFHQPjrMZspcbEgQKRh06Sm9pWv5Wd++vxGFLyDPbBS4le
g8ma.k4wNTxmBbsjUMB5ge1Awi5f3Kdq85NWRVsY3EuL7ySCVYGK20FX.cMi
jkslADZzgbXz3fFETAJy1p11nyPMYugVvosI61rXOCy9ck7pPZ5cvSihOnZ3
e53Znr2I7oq9OXDPGLryvfQBFYDS9fgBFY1OgNgCFFXjg18zz9abz9m6xTqc
4GrZx+gfzzMQOrK0xtWxztdpko1RnGBVkqA4AyRZRKzCinIv55GycbxvLvlx
6zmBBVcujjafMGcy.6aFXey.6aFXOuMvdu7wAaiME2Kar47qLarGM6gPW61C
clcFA4lyHb6LBwUtyHNMarc5dOl5dk9ebk8eBp+2JVbyFBz7Mh82HvF1QldS
dIfs+9LXxKFLxfl9ACX2jreXOk1eC1Y.Yq791y96cqePuwuSireW1NeIAoEr
1NSUgRQrPSlfUTzFQ1kM49V0AxifCN1gU3kdpMXYXEIakd6MaIXFola9v0Ab
pf.mVsQuVgyV7NjVIrm+xKIoc.zsD2EcAYTeeylnRtQkKFTHqIUsFYp8S58b
S9zqdOGDG2WZlBglwzLpU4eto40ZKXCdJrNQqWcrL6luyK6udHH9o9R2Xbmx
avBq90TqQa74xrc5yQa26Bw1o320NU6CfpslpxECyUmjha.lOh9692xFSkF7
d286x9we2+mreFFuzdk78QGBjoEYpGHciJFNkijyiIaVGjVzonkAKlavBSrR
RQmHX8RvFsfRsY1ev5JqoIrz1utwC2+0KD.bN13qbn9HezI2+pFQ2jj4iJue
HXQZzGC6hvqvbTh3wtIdNwxOLajUDrz5GYuEurq6kEQwosQ3H2DtxXcoPLL5
lL9AS5+nVIpurMbo2ezrwXSQUJFAXsMka19Cu2YKEIsrG1hjU1SZI6.Nz1bi
DRlduRAwmodGvKgmCwp5xvUAewSimuDFu0Z4y92UIZjojHphf0pOxIB+2A7R
knwRSBcJSwGkqWp0gfDWpezh1nnusO5MMilA2ramiUTyKUzam+1sye614uMX
+zVYymxRTyzjvacXadwk37LPvLk4TxsJVPqs8S+7a6n6j5NI9+csV9d5+Kva
6ZMaldyk0IwaS2Xje3oU4164jO48oPuEAwd61FpeiqS9nl98r1w5EDuT++dQ
qeY2J8sSSxxkB8jX12ndA61mSR8RdzaaxZ8k1sLJw66dIbyy5638PP1195Gi
9yqengo5mdlPV8CU+euDtLX0aKMpKkcMh1lubdFeDgwJ.kxJFxZ0XlwiWEyX
OFsYap21.8ZkPunXuzmCKOkb2vCFTlk4056.rhcUwJGmDoY.+sGXE+tsIdAq
9TvWzFP48aeukEz7ldaqPDGHDQIVEn8wyMHprg0PMttECrqgOtOQKI1hOVeU
hNs.MfT9o0lQ2mqCL9usxKN8YOrazs8yKkCDAI1UeLwocZwjyJ.8adMSiA7o
cd5tSXjbaLX7w.gNuwbfQ9yIFvAtiJCK.Y2fiwtlB1Giwsu54iIY++ogSN2u
CSMNCImQRgmaLROrSaBQb2fPCxwK.ANCMEYt8lF2hJ2+yAxoLujVuXUzhe9D
Wp4CC+rqznp41Js0ZC1i1tJZ4wCfXH6Yw.xEY2xhprhmES319EeC6crfVdwY
O7wRzu6gG0l.gUqEJ9xqrf1zXsW5lnmdpvwn2WLh.UFtvr3SdsAR1fb4Uuue
UX7SZcKctoWGpE.jcBisdnEhdAz9CUSUDiwAGVTroOrnfNVZXPmrYo0s1nIb
.BNf1xhZK77.qXcfU3oL5+.GUYSNeE3.byexGJ9PGJxKDKN3jwiLevJ7khEm
LiRixdj5s2xhRGFF2uLoDi5L.AXHYgivUgO2gFQqgHvKgZNn3znUgMEc.pNi
RJt8vYxiDPNatTyoNDbOn6KP..iuGU2GEuxFPD1LMQJN2A3Sqyi+O2sIMYaz
1llEkclkuL6Dmu8Wm+JGVqjzO8ygepTpMWjjjPHoYWYt6mdYSXvxFIHeHDTo
vTdNPP+n9paRhZljX.HoYWAqaYvhW0psSZjh5T1PtOZv9Fql8GXPBOASRIwg
dO9Xp226sN3o3nzcKyNy4UK1spbHNUjR6LN+XXaFTYmBwR+y8jWCNeqOQ+qP
ANl2nH09P91Ew8vSkhQrhg2Fh7NPWoT.hYouxemYeHYd.kw7QJIhYhlNRwqU
OX5.GnsGyJgxN9ajCT+xN2C5VyB.QpNk1lG8tPgdaVNn+XmofQqroTEjbRvl
umNyCC7XG0osQvaC9zqdBAZf5QS6NPaQD4854XohvEJtT3mUSR74fJIImuzx
vrN02yWahkOp2Zf5.AzRcqg.XpQSG+uRPfN28QoWhi4UJKMWJ.nUaPzVsCJA
25HX6ocGr8XjeaKEnrycLnOo3Pmpcoz52WicXtAC+lWGp3PZ+EGpLDuZ1j+E
Cm36zBBoRUehOm1mMYk3hurXU3qdR4f2Ozu2XP91g74zlAj6M6EP58dA3tqQ
YBsl40pQYXS0.gqlW6FhvCFBDcBAHTaPfj7U09gxdyI3i+Jb6PUu4Flavvf2
QX.cD.0diC7ut2Mr6MB3bbajtubtP5uDE+yCk5Q8l5yY5O6dBnbnE1qzqk43
vGrsvDyrZdLlvNgBv44HQqKGof8r8evAfFVdbpZt3ume6PYvI.b1StuksoW9
LonazXrnAU0FIjCMM+v1lStzqiXKCJoyALeSjF03ZJfxZNXxldQ3d56TJkxh
5RVWWb7D.y8X6rt6bd9V4FtobddeQXv27KocuPQlkg2R44ao77sTd9Rk6HNC
Ocp0S11.roAU8l6IOx1E5UVkiL8MZDIGcdusx3HpkamE2cvYLWSznivD6.VI
.137.6ntdRNmYLpKXvOubv2dfm6NmFIJartv+5EDblpP77hO00KHXVP7dbmv
fybExeuOemsKHpT4G6Y0erwJ.YqUAx9m7qBqGxXb78bejufJnHV1opXL5tmY
vXaJF4Hmr5PgnFJWjWOo2t6xnuMLhxh6llqiby8zaeRwOk65cM2BejqT3qqr
458mZtl5rvkfQFepP0K8kJJQhPDEQYW6iwroawO1wh+yJNaB4rW87yBTc+ot
z8uGxyvGrVlqPgUJB1.4RTeyftqaDmwvHyOl5l2HAwZEzwHx2HndPZxxGN0d
SAGHCNQCLDsxDDVNNytp5QL52Xz5cqeMSI0SDwnPQL8lXXpuRx0RhwDefIe7
roRj7h2yQuDbLyEFHbgABWX6g8ayk8S.ip8zN5uzJPQ69MsMemV8KnQen1le
T6vWp.7mZq9TsheUg3a0J9WUkWAXx0Y0ZjPlu6KOEzhOVg3mUP9Z0s+Vg5y0
t76JPeu5x+qt8Aqa+v5xWr.7GqaexBwurf7Mau7OaW9n0seZc3qVG9q0gOa6
1uss561V7eaa9vEleba1Wt0jxTUbd82f6xfW0ZDXVO.KbYdEWTK+r16sTsTj
T8tsYTWqw2p0GvJz8TARIv4A5495rMsZwUraUI5zKOU1erxdjmAv0hoQwCGR
EPfThstTZhb.F9qNTzp7g2hcoIO93vgReHPYdwgIGKEe0gkrzm8L0Q.urS8I
bymR1nuxy5cEppfW+.WBDvklWuBsI2PdGV3ZBdqqjbkt3g2Cd04Pa9vzZ1cv
0NYMFHtVqIXRz89RIUx8o9bFymhemmn4JMVuAYxrAj05Vp0vZo2iaB+q6BiW
7E2.tCLFDyKAYOLFM+JCoPLB0WRURFMK+DQiDJimMnrUu.s1QZq579AqHCl2
On0q+krnk4fHCuePqzYxpjm9RQIJtmRVFYB0lfMeo6IGrnGK.vR98DhhK7O5
sSJm901byhfsKBVV932faSd2GGYqGIoMhkxBLQeDkPYGclQdqBF.BSlKHbs9
X5vgOZefut5ko0gKjC3p5sK16NaU0yF5imyDz1ANK6gffyHP2s1oGP6y29m0
6zr07DGpGrrJ78xxFpAFVco2VWvJ.6RAyIWwYN0p0fMf4MUppZB4qWC6pWrw
5pNwwZ9cWtCDI4DFGwTLDCyLEHADRfIHgTpmXnRasWfHXTFVJkBhufmWzDFN
W2vQ.+9f.XeHHfo0Gik9TMURskHB8+nJjjgzWzhILLhRoHeBQH7EWP.fYqTh
.YAHeExBv5EKf..BXiop7ebwnKxnSWWYyrsQScr3F+UFBzKwarpMd611g53E
JeDNtJTg0NIqA5PaGEtvFKdgUokyxYQla8+xcaJUsmfnRYO68OH6ofkkKdZd
QkPw1GYAJ0U0IduMMIqI.s8jhSwhVLpb2mqD1B7F99LSw26pDQy8Sx9EJWmU
r6ov3W89g8LaY7DmZPC.7Pv4ZniH49bDxW5qLPGrBwM91Qh22iDe41WtWOS2
1whKr9s15JUkM547YnamJ9sSE+1ohecdp3KCqFFLPcIUqq7dSGcuq1OtQqeq
wFQK7bWUS86uqU5RpA.OZURZweX2hnkAd+XR71jUg2UW6caFKb2+iHslEa9h
mcU72oYhd6u6Ou4OG++NK.qOng12kEGnFy1zi+eO19Vx9+n3EaByb3i2u2nM
+2e3ib7s7n22k2rq+u8605Q+2+26s34r7L46hhwu062q+fu8WLu4z72l9J1O
c5+Y1uByZueE9V9G7Pu0q5mH+Odu22cXH81heI+4zrepgOr9Mu+ax7r+c2Ma
NTfn3xcatg3iUbe3YIGcwJ.W+itVNbE8fpAfrMM3b.RZCVMnjXbNBJzLxQm7
dYFLHqfmDy06yvjCzX8x6ztLJMaxuR8x01V3y+QEmULEFM2YCOc6pjO4EF+w
vUZ8vZuQ4RZynEmYrHcu3.989UiCee09p0z0QWyUqw1ofUtSqQzWKPUl4NYx
KSS7B9XRzRuGWEne6K8BRSCV7ya6O5Qc2QOsBS4H78bhj5yUTZt8w4oL10B5
scwyqiR0nmsXyzevhHflS4Le78YESSBW4SMGZeNS30BVE7fVUzco4ckYMeVR
xP.Le2N8yTksH9YEgbsdqDgjRPVHiecwdkevz50lZUZS9Xn2+7ON.Dysq93x
6sQAh3deejvFSHB5bCq5xKeZq00hq7drbv6UW6YqeQc6fTBGJpwv5ElBrhQv
TIWX2EPHLKZ6ie9Fury9rB7oOuIb6yupYd7vmXBZhQJftVso7gGS426+MQtB
tIISLpE381uAtQdZdVd4YpHAt6ArU++g2bko1R5HVOu3ij9rCmyBATlJyG1r
BeNMqj6GrfMOkEQk1IGyhh689A6jy6EHueHeFBqu0OjMMYqbY8tg8VvPEbCm
rda5lggNQVZq.Ju2cZ0YrzrN2x8iZcTVG5sL5wGC2j0FxBV4Mf8bcpjBwpUm
V426oUyteQuUrCNZilUI87CaOwTdVAcGCLmbOUH4bh9BFSMnhoim9rucQs3y
b.69Bs7Hv5JxLa2kgcAQECavRihFBPvwBxpGjkkPCIzS4UwpYEKHrACbQ5vw
lop4LmsnCXW9kO4MXTviEi1kSb6ONKxQIf.FZsvpZ7BopR0lpnkujDEmlO4y
xCtZF8dJgP39bige4EyfhWcxwIJHbBegvIaoeBqDUzm0Vhma85VeOfsE9.yq
lbbDVaQtgdSdOTgrB5jqOHUo+kFSQ9J6+DYgratouMbyIGJvfjCn5DJlNVJj
vV.MY2K4Hd1G6XXAYqf5MbuosUVi.y+fuPM9a3CQq78KQC2lAdYngZvy5gXC
AG64YllI5EJdQlnAODw7KEJxAqa1z272Yfa96MweVQ9YdD1Px15UjkFIXBCQ
nlr2cuWlpeuVlBHyCpVzAWBcVLDwrK0ZM3CQ5kZsFF5PbuESS4fALbIm7wBE
CUYL4XpVpv1vcwLst6DEmSQ60uJWg0Z2YZQAnyHT7jOiP3yIyjMiFL7QCYZG
MfUwjMgVNLgDHVAEtaBJtJnPvJlwl90ZXb+VqMsb2MZVDrP7qp2Dsu3XxX3J
w19kA1YMbjHa0Rhs8iygzUgB+aKCVbrIQ0XiNnw1bp1Z9ZM0Ps9lFy3scJCe
2MCqUIKBV0QyB5D59YkKq88HUwpUN6aLd4atM.zQiixmaq8+HvMD6VNtkVN9
vNJS8UKQ8yc7TAANo1i05Fb5DNEPfy7Lj5Fb5d0N.3joPW0bms1p4VmjcVba
g.zc7WM288fz98x2gIqOpT3eXLaHshN121shN+75rYtQi1i0VUNrKt0J5NoV
QG5Vqn6JnUzEmjgMiZCoqqn94wU2GGDaqNTu588lffyFeyCsRP.nU9XiHCZt
aftppc+S.dAn4yXq+ZZ63ugXUq7vM2j6w2vqdTiLX1scugW.a+l9r4Mb0c93
k0st7Rdz6gn3CE82nvssF7cH2wobtZaVk3vBx9ZF40QdisacFbrN3o3nzc5c
6+9.89w1+rcPwsTaIee04ddBJcGykJIzskjTB3XPc7hyR14FOD.204x.GzyM
bvAtoxkANHma3vG1dFWFz3rugb5FsYzCIaBJc9VbkDQUDrOWxIBy4ags+oOx
dqFuTW4bihBVYaostTQtxzEJ7yunUCJqdt38a7p7y2Z9EYn5EIAzR67sYStM
qYvR+qLzaYzGiV5hysCDR.l8xF.p8uuAOKXuZj6ZvfF.MM7smhWdC79Taesm
cTSqh4v4oTP4o7ssTCl7JCcx7h21AiOJv9OgIxK0fWYq4zCqcm.9fgt7xjk.
Z7QbkgOyVsMjLndp5qJkMNQ0LDtQMqqi2qkgu7aKsLDTnrUeMojwfQKha1Id
QsKHjugztfKAxK8soxEbEP34aScKDHfqstZUs3j3db2C3Y1B6TtnmqNtmgrQ
1TqzECA83ttR2cb1ptq6NR+dj+JUc2rt9vpvMAoyMVdha4LFrlaiqJI6q+Mw
mbLmBjYmKtNksu8utIcv67QUtQGqFApCwm+UEG4IZBGgAEer957paSpSwlDB
EJ37MoMIDLP34aSaRb2IpxU59Z0ljUgwOk97fwG.ESKYd8ywJ5Q9Mfn4oVaA
20ys7XJ6lMIiLvyAB72LIYbwcmJ3Hs8nGJ6lEIiDjCLtUyO1juIb1TSIxoiz
RvtqWSYK2WuZb6Vk.ZdvrRtJU39TvFAL9FBG8s34i.Dbr7MWchb1n2cceNED
E+xtTuue6hfUgd1hN8PsPwc6dUUHNm1GH4yHXCPENtCCPfwyj2pulGU13op5
8Bt5jPbVR1r85HL0eeLQr+UsTLtXSYcNou0+JmzkzlgIJ79WbAJTavqvXHny
VVWPfQjiu7BT6z.SXxdRWLeWjk+bfr3p9QWBLxEcgmEzEPxxl2VG+6VnI9bf
lZp3210bEiIbMWQlzJ9DXgFz4TwvhBDk44A7O8vKtDa2fOkJnYyzEwrlPIO7
hKwFNvWUzPsLtQ5ht2OWnCu5Rrp.LgI5IgIjTWDFdVPXJLXByl.oZkDN7xKg
NBvKArhoWPFqWBxl1BnIFb47T4O8HC3hu3YX6FYuXY5dwPtp+412jGptEeg.
c3Tq7ye0kP..XZlg5IM6yEEH57WkS0LqXutn5oT3.oW6S4nzhma5Zt5aRxkT
8MReT0F2C5RwbQVz4.YI567kvsIe34.gov8jvvHN6RZidOJLx7N5qEyng3Eg
+.tF0S+9iDvpTQXS+fArJUzou94S6aq8XJGKvKYytcRat2H3Da6GRhO7pKvl
a8fxnfoLafFm0Zy1+pKv9avoLAqmyYhBu5BH.CNkI48jxz6vwcQZn4PSy3LT
13YnwoNsel6rISuzPFoWxLvykVZDZ5QFdubg8zhL9fkRL87ufwko2uQbvVmK
ld0+3fEkym9tDDX2FM8J+I.K6UPl9ACoWGh3ztpV.Vd2z6nQAqWKkvyjVuJy
QKrsW83p7y2lXKyULhwQ8kekOKOEULNOv7pKfl9xdoN7DgORI+vqt.ZNK6kh
ySDFfIThKP.MG.A0IgAmEi7jfk.oldMKjf0xQJldwgvaMpSudwRYuBapIFYT
8ZZZZQFviE7zavqBrRWxoWCPE7.Ra50LVAVoKkzQOaqbCgnsfH8XSFY2xnje
xzkO9v+ZX7tBgr59tRRKDwCO8XzpUGDX+l5uk2b29X0MWt96JduRh5IJFFKy
D0SwTAga9KZVWQu5VD4eP79OIymoPlDYgYRiEyeIITepYiil9rjiOUDVYeVY
Y+By9W5KgqueyatKH9o7FfhnPq6vNguI4kjMGZKL2SUk+j6RSdZSvxn7.C93
tuEm9J1UaLLQax57P0aqGY6+qmVhJ8M0Li3go4+vtEAvmZKM0vD9XpAVIJpT
Y+KeLhqZXS4fEKzC5RedJQOCYPVSNFY9Kg8Z093gqBWW8yKjbtv7TkHrTZ+K
8WEooTU5gmJ28JQLrv7v0CblPX+rDk86ymW9CuU+3WjMIV94SzjpzloTBhkq
RhPYcfoZO9r3527g+PTbV+RJ7.mVVdXwsP2d9UhPxn3FIhdrx5.eUWKt5Cad
4kWxCKpjBjuTZ+qlG3UWdw2Oeyy9mv9W1uty2pqh.aw1qjVd5+qWBi89of3s
d+T35nGRVs7txBVmpUi+71UQZkW+iYs8pAspDmqQ7webdF3qC9LuXy0x4XtA
4AVVnCqGD4rjcuTtBAqW6BbgWcaHNW3T07WoWxb06eQTEEUlItf4iOyzv5vs
aCdpG7n8XdXJj3UPr0dIN0+KWhsDxL0GdWi+UKe1qREJZXw7fXRp00B5Cmtg
knfNbljP9LxkaSNu+oU6BKkyX8WrFEK3L6dbGnFU1kvmepYSxmhOQx4HQ3eP
BD8xPN+3WBNUpAQzp+Y2tQPoDq3.BFiomcp4eYSX3ISNl4EZ1+NlA+mcJ4+H
b0pjOchjhTaJgvJgEKwl8G70yUMHpsTGi8OY5tl9JDRdlHYSC28eQqi3IRvY
qgPFVQ+rRyfg1wE2h3bPMarJflQNCTtMmw4TU95HQtkU54ME0sUYyBH3HBzm
IzJVWJvHynWQYbjJ25RA1GlkwyBbHM3E3jeECn9CAoId+grd05HZ3z9uoieM
2k01VW9Aamp8CAooahdXWp02VEGg8sKp9zpjGBVk2gTO.HM0kUOLhlfNLcVe
hM4wGc1goarGRqXc0Co8413a21CoanrQzRu4t4za+aztFMGwr4YLeen4YxsM
+acM5acM5acM5SqqQW2+BEEMFtIaKS8889ufmvtkks5fYKGdWcMSrSqz53tB
pkWmTnp4Ys04zJRLTfkqqYTQhYtxHQ.Vjlx297ppHM8RVaL9OZD.OY8qdL2t
ErxHFhepkp5ZOsv1crZapu0lJbU+BZTUt1TmqCU5.nVWqp1U+vdbphWE077s
kCKqxd40vdlr9IH2nldPz1CjFet05Cpleco8GPM.coEnaMAcqMnKMBAnUnaM
CgncHHMD6kVhcoonasEcnwnCsFcn4X2ZO1pFjsnEYaZR1GsIaVixZhZpJOu9
avYKUtlr4Z2r8cuPnNTEnaIzMHkFU81soVPapFT87dJuI3oCd02TrlZCMTvT
gr4XM+X.eR.2Y2Jna0H5TUhY.Z1RkVdHHprGHJpqN1RuYqmUHZCEx9gflBHn
IlgbWS6mwvYMioFBRgQvY7XrdINE4.nlFwoUzGpVo4rAbuo3+rIzGRkbyctr
UaBqdP92aNmyDEHcPA3K.EH5CATMerNeCSYeGlnKxvT0mgIFzvr3anR.943f
hpY56.UA1wAG03gG0Pr4bNbqyiazBAC979BL8KAaBVucvNSk.zYp4kyF+qst
b6fawHEKW4HfdUDitJ6FHAKVrasI6yd0KqXyMLVIefk79bGhgQzupb8rB5Jo
KgumIiNVU2J7RfAT+vaKY8fPiNjpfl4kqcZeSYK2koXRdWG2tjp7qxK8mBak
Ga9W4OgVccya2wG96KQInseTEF9TYdo8ivuLo4L3b+TN8UOAEYFktk8nJQSl
9hFK75JAZFUNc8m9oIeQulll3Zzxbp7sKI8ZZBeKSTukIp2xD0aYh5sLQ8Vl
ndKSTukIp2xD0aYh5sLQ8VlndKSTukIp2xD0aYh5sLQ8VlndKSTmSYh5iqRz
hPZOcSu6cd28fVgvBIb5iIaVGjVLZ8KEU.pNyAUkIiOn1ByZdbz1dNnhaNGT
IGeBazvSZ3lOXC94iyYCJqbMxRe0Yt29NCrkAAGF0YgSQoQVSQqeo3B.CIUc
UlNKVCghYKopaKIRPavW6ILPFolGk3CFN2Go2G9N+XT3mLQ0ezpnzuTZ9M4w
G2FlS7u2Ptuu3QOdmMKKVtQy4eH71OPcKVEs3mSedSxtmdtzMpEv7UtyGa3N
O7z94xiW53gdBOsrC17TtStIMlq19cM+S8skvWyg7QvlC5VpZc9G4HUsOLGu
Lbcx1jcwKuWOSoet2UYx098W9Drymrqd1A6mxczTMbzcDNr9o7oaR4Eqn0BN
8dEliz6kY+Wl+FL263WPkyktQ9U3j.BBMPjcRCMOpE1yg12RekdkVwYyuDlV
jh4U464bhTCDhgx6wDBWKQhAcBAOBjP0QWKSHpNoglG0cMgXDgdXBASNmSHJ
VOlPxGhsMEPFio.PbQECszS74Utdf2QuvAOdOODjm2YjIfCaUIFaaJoGeUCL
AzQXPiqzrQaMTj7GMbBiQvdjrQ7Qhg8HQi3ij.6QhGsGIrGHZDoQDnYR4n8.
kJvBZwsuqAxXpR9ZrhuZTFiJPSCxwStpOnmWkQ0ntSB7Q.dhTm.j1DBYKXvX
LBDfjiJNyS6z1l1GCkpwTZeFCixSjctehvrbXDehT+y7CjwOyOPErMiGymHL
MNni2dwTXrM9SnDIXCgZL3iqHBFncCY7QTDgBLxiGomnDNPOVORALFZ0Hhqb
XOR4H9H8g8HGQCEnD3OxwZtDn3P5H9HQvejij7PP35TpsBrQ.eBEHKgN.FmY
YNrErf7GAmVzeD7wxQllgH0sjyNGirrDcpv+vYALKS3aSoLr0KfEekusVPkS
MlWMdTiSGbI5Fvkz6oE+m.6fXJ4pH+QyLVCwPbZpc2yLHrMhby+mfQbPLT+h
DC0errIGlu45dsvdWxxwEcPKez.bPps.T.Y9w9To5tYd1UqlaUqhaMT815np
sUqZskeX5URJjbvvUxfTIn.9kCmfWsf17vIw2RheLnj9XnI7wPS1iVhKyNhI
y1hGyBALv61y20XzPTIRHJyq1TRbzxTxfRbiSHoMNoD1XvIqwoknFmdRZ.iy
uVXJOvPEdnIjw.SFiQk6uo.fnyDuXjVozPBV3dESqIUwHMnJGu0sLdFVBS.N
YI5chRLhzdwDh.j7KHIAwHN9JmrCsLB6eBNbRRK5eBMLzjYXnRmNiaVVYAD3
I1RIn.DNOmIkvHw00PxG.Wr.rDNXbGokRr.3CUXISvnNTKkz.vGovRTfQcjV
Ng.5wP0UR.LpixJA6eO18BT.96H39GKRodP7CmPfEs1izHsd.5CQtEzfxeXA
j+HQZMD38PnMnAa+Eep6X.02BY4LH5Gv.I2yBsT70+0+J8a4+ObSsTRK
-----------end_max5_patcher-----------

This is an updated version of the initial patch I sent. I optimised a bit of the calculations, in particular the numerous and unnecessarily repeated fl.hypot~. I also corrected flatness and crest to use fl.flatness~ and fl.crest~.

Not really, and the window will hide a lot of the transient if its at the very start of the frame. It is probably worth doing multiple small FFTs and I will attempt an example in the near future. It shouldn’t be too difficult with fl.chop~ and some ninja patching. If you end up delving into the tutorials (with some new ones to come in the future) you will come across the concept of multi-streaming FrameLib processes, where one code path, or set of connected objects can process numerous ‘streams’ simultaneously and together. What you would in effect have is a stream for each division of your window that is processed simultaneously and then averaged/medianed at the end (depending on what you are interested in). You might also consider using a delay that is 1.5x the fl.source~ delay so that the transient is in the middle of your window. Cheap and nasty, but interesting to test.

All the blocks are there for these kinds of things to be built. To be honest, I’m not that familiar with the literature and how to code yin or EBU loudness stuff so I haven’t, but its definitely possible.

1 - Yes yin~ could probably be done if I could be bothered to read the paper and loudness deinfitely, but I’m not sure how it is being done in flucoma/fluid right now.

2 - you will only see objects which do something in their audio perform in CPU usage. It’s not really possible to figure out what goes on across the network. However, all the CPU for the network will be attributed to input/scheduler/output points (trace/source/interval/sink etc.). The total value will be right for the network, but the individual values won’t be that meaningful. It is useful (and a bit of a shame) that the unsynced versions turn up in the values, as you’re not supposed to ever know about them - I’ll see what I can do. Out of interest can you double click it to show the hidden patch?

3 - I’d argue that although it is possible to take several frames simultaneously and average, for these descriptors that will just be a more expensive way of doing it. The question of how to deal with a sharp transient is worth answering though. The obvious thing to try is centering on the transient (half the delay). The alternative would be to investigate non-symmetric windows, which must be aching, but I’ve not researched it, so I don’t know pros and cons/details. There are ways to window in frame lib with custom buffers so you can design whatever you want, which is sort of what it is for…

It does show the “hidden patch” if I double-click on it:
32%20pm

So from what you’re explaining, the input/output/scheduler points will show the CPU usage AND (at the moment anyways) the ‘unsynced’ stuff does as well? So it will be the i/o/s + the unsynced as the total CPU usage (as it is reported) or is the unsynced stuff getting double counted?

I guess it doesn’t matter too much at this point, but just to know what to use to test the JIT vs fl.stuff against.

In terms of the transient, the onset detection should theoretically be bang-on the start of the transient, so want every bit of energy from that point, and ideally starting the analysis asap, since if I have to wait another 1/2 window to analyze, that adds more latency than is desired (512 samples is already at the edge of what would be musically useful).

I don’t really understand enough about windowing (in the land of FFTs) to know what having non-symmetrical windows would do, but that could potentially be something to explore/test. (In the time-domain, it seems obvious that having an envelope that takes all of the start of the window into account and then fades from there would be ideal, but I don’t know what that would mean in the frequency domain)

(so is it just us 3 on the forum so far?)

I was reading about this yesterday:

https://community.plm.automation.siemens.com/t5/Testing-Knowledge-Base/Window-Types-Hanning-Flattop-Uniform-Tukey-and-Exponential/ta-p/445063

Seems like its a thing (aching), it works in this patch but I can’t speak to the usefulness of the numbers or the correctedness of my window function.


----------begin_max5_patcher----------
9844.3oc6cs2iiibb+uWC3uCDSBBt0dVs8axNN13Rtfj.DmmW.LB7ErfiDmY
XVIRYIp8gOj4ydZ1MkDe2EoHkn1QCta1QjRhc8qqt5pptd7y+xewat6g3uDr
8Nm+Zm+nyadyOqtxazWK8JuY+Edycq7+x7k9a0uw6hB9b7C+u2ce18RB9Rh9
5qcB9x53nfnjP+ku+w3MyCb9bXzh3Oe38FuKYYPRxWWGXdj2cmy+y96EtP+s
n9peGAc3Sr1OY9ygQO8gMAySLeHJmOieuCEilgt2Ayoo+CgLCc76JZ2pvH0S
ROfw4tpY.T7x5GQvlLRdOM+l6dLbYvmB1rMLNJ+6+M24udctq+l7enTn5+MV
+c4c+wqEFYtF5301D7ov8eEhiW1eihbSTz5tMZL5tuHX2k6aJdQvlncg5AT1
U++9k+hCCM8rTj+pfsq8ma9BRmMOb+inHWpQOja5+HXZrj4gxghJFgmVFO+i
AKJLzuKdcPTXz5MAaUS09IYzvw6uH3Q+cKS9viwQIaC+y5QAVM8T6a3wrgY8
2MkRzDwe6FEO0Qx3oMgKhiRGHEmYRu99G4eTwZXXPJPS52Rj+5593JdOE9zz
c2pH1caeveS5D2CK0ODxg6lDGur38N9IWF7XR18WGFEUFPShW2xc2D9zys8o
eHVc2Us90qu01OrKxb6OnXRR9vV+OUB4S7WtLawcomvW7iBW4mDjDZlOHni2
MHxWQvOuc9l3kKKR2la8o5t0BE6+7fOGtH4Y8Cq.6g5CDtdOi0cGm1WD9Tv1
jRWLw+osktz1juZlBxescOjsP+CIAqVuTQMkdGEDEVXQcdQhEuQIQiqTLtga
WFpVjdbQasx8tOuruxx+DE9v0HBDKzqdIRV5+P4BMiNVlmSuQ4fF3HHQcK0f
27EZDkVyGOm.SRwA0F0ZyDEZZljKJcK8AjsLjiIGtpRR09+NUnUug4h6.UbW
nGWNSsbv+KuXYBnQvmCE74oxzb3dU19wBzWZanAGclGuZkRxb8vyOraqBdxu
Sc1dzy9oneJ5Gd1O5o.GEFtN9yu372nFpODr424jD63Gc3C4703cNKC+XfS7
FmOGu4iacdPstbVgm3xvnf4w6hRJtAWoM5sg0LpFjwDrQTtlO0S1WvFctYE2
D7T3V0pjdyMhcsAQd7bbiBN.tQx4jaz1ZU0NJa6O5XcwpgiISPImckgNwQyC
5O3Xc0ElpWVI0xy33olfLKfiVDEqunCEH3PLlYvcuxXc1nsc4D3dP.AHrKbI
OSI1Gk9Kqew48KChdJ44T0Tbd+tnvjsNaU2XYvVm2uct+x.Gkoaq7WFtUoVc
egRLTnzraGQRlZLa5GauoeOXRoYZxlh1+6VHdTKDe9asWWnBZ5qGdQKB9x3f
UlwxcMAFRXfA0U.FMZSimKMZr1OJXYSfAy5BChFEjYbFFkAYtCBZj5KkJWTo
+alk8t4rI03Lj4wKMtwI0dIx8E9UoAz5Mwqi2bv2Hy3Mfq4bLQpByMY9oFYR
eCMf3ai2sY99EhYFw3TZHoLqNIL5f+Z9iGL0L8MBeZuyCEJzgRMi43MKLNlC
MhCvTwyvFgo6INxn09Gg8ACc7GKDv.Cd5vFkZwjCtA9H7XBWtfGhbyPbLGLd
cZvT5ctJbw53vnjsG7UL+dGWsFdG+6KAD60It.HTEgp0rWYf39WbIDAAVdcp
5CsrRa+MNdl.2k5m2Eev3Z6O3mjrI7gcIlcZJbbGczsqOsL9A+kYtT8fe5qy
srGFQ5+H6ey.NXm3TIOZkyaVaRGX0bBRLbamfjHy0cTCurwq.ntd.Rngmd9G
V5mnzBXqy2s3g2VKcQfPWlC9viLShEHl.Y9gimHT4+oROpGerVxiBf7XRT5z
l.etm1Z9fOmWfOrhoQaCeJxeY8m0Iq0y5zkylwJOIxndo..ie6fO6wAepMnB
c6TOucpm2N0yg4TOs3nMsvw96GRJL+kfQnqRuXG4Goj2ldDQNuOUHmylfUwJ
dzQ67hxfKgmwWJdWaGXzFeqmIR48aqrmKPGRRPjzMZ4jqLLZdbjZe5jWb7Ws
9c+tEOL1NtESyB.A92T9slBbsDC8Jvu0TnqY7zVROQ7a8X4eUFXGQP8FeWiA
2wSmAmF1YWzLIbIGa7mlvvcgo2qPm7zr00Ol4Uj9YfMUzpOEbYU8RRlA1BzM
CruYf8MCruYf8z1.68xG6sM1TbmrwVHtxrwdvrGBcsaOzY1YDjaNivtyHbux
cFwoYisU26wjyjpeDRyOtT9qEKtY8AZdkX+MXC6RQww19avF1QOCwIyjJZl.
GAQL2ITnUseX+5x9act97hUiruOcmuX+jbVauORJ2usyc0YBVg79wqMax4F0
AxBOCA1hU3Edp0XYXCow1cMaIXJolY9v0AbJg.mFsQuVgyF7NjRIrm+553jW
5ZNlyfjh4dBsJWLnPVcpZ0GpckxZN+mBpRtJNkEo27dmz+5A+nm5Jciwst1C
axyED0X.i3bS4MMOm7b318tSqYJ99loZN.p1X1lvset8ijeyfrQzew+Z5Xpv
f24teS5u9K9OR+cPzByUx1SoOPlR7gZfzNpn4TNRNOllSKExLyhfEyNXgIFo
JnSDrpMQhGb1m8qabvce8BA.miIPB6q+hGbx8OoPzMwo9qw468mmD9of1H7R
LGEHdrchWjkhtSFYE9KL9T0Y95csurHLJoIBGYmvMwLrqa+naxvGXk+sJEJ9
51fEN+ghkckBtvBvZapPu8GduiGxSZoOrBIBix9SjqGihotDNSdOvKgmBws4
hfk9e0QgmqCh1ZrBX+6p.MxjdHpjfUpRIHt76AdoBzXgIgVkovQY5nYbNFwV
bd1flYnW2GCE0jWpDy14XoI2JkzamE0syh51YQMNUWjTMIbVEzjGMI1qHALc
sHfYTrfVY6mKb09nUh++Rokui5+7c1tRwlo1bYUbz1jMZ4GNJUtcdN9yNeNv
YtejytsAp23p3Oon+rJvhiezhzxrR3p06VptcRbZdEnlDS+FUKX29bbhS7iN
aiWotztEgwNe25fMOqtiyC9oa6qdLpOu5gFjnd5oBYUOT0+sNXg+x21b4Yws
o4Kqm2Ew06XEjHqTrgSMd7pXF6wvMaSxx2emvHmjmCJNkbW+CLRlg4UZDPKY
WUrxQwgJFve8AVwuaari+xO6+UkATN+52YXA0uo21HDI.BQThQAZNdpAQEMr
Fpw0MXfcE7w9o63gM3iwucnS6P2sW8tNqGd5edoSTxyN3QsLokgfYU+G1IVB
WNu0wre0KoZLfOsyV1dxSjYiASLDHz48720xeNwCe2dDJX.HyFbL10TfunMt
8EcksgWhN6LN4YuB1ncFRFijDO0XjdXmxDhn1AgZjimCBrWH1xr27XUAL828
jSYZIsd9xv4e7DWpwggelUZT4Tak1.UZOY.4hxprmRzgRr0Xss+2DEsSJPX0
XgxzqlcZAbzV6krI7omNgRIIApLbW8hOuqMPxDvGEJoa8NvEAxNgMksMBD8B
ncGpFqnmpCYhxTpHCM9IuDui3xYu.3.N4pHSGrx6BUWub6XgGaLYxmNE0KvK
8YWHVbRWVChmFXk3BwhK.iUraYTnECi6VVEhQsFf.LjWtivUhO2gFQigHv5.
EGTTR3xf5hN.YqQIkvb3LYQEmfMUp+RGBtGzrbD.v36Q19QwKMADgIqK7bO2
A3Siyi+y61jDuMbacyhdslwqLyDG27Om+pnUijzO9wfOWHMeySRdPHoIW8b6
GWuIveQsDDGBAUHjcmBDzOnt5l3v5IIF.RZxU71V3O+EkZDjZonVkMj4iFLW
a0Lm2OY7ivjTbTfyiOl37dmU9OEElraQ5YNub9tkECwo7TZqw4GyTC7oloPb
VGm3LN4Uiy25Rz+5JAGyaTjdONI2NwUpfPmO71PlZBs0qTH.wLzWwuyzOjWV
.kw3HoGhoilNR9qUMX5.GnsGiP+hN9afCQ+hN2C5Vyt.hTcJsIO5cgB81z7w
9wVSGgFYSoRH4jfI2GslSB3gNpSahf25+4WbbcQ8TOZZ6AZKh3MSMG6IIBWo
vykmVeN3BPkmiyWZYnWmxc3JSr3nNqApEDPI0sBBfoGqHyeCf.st6iTsDGKJ
UhVtT.Pi1fnrZGTxd0Rv1SaOX6wHdSKEnrycLnOp3PqpcIU52WgcXpAC+pW5
q3PZ2EGJ0Dubxj+E8m3a0BBOor5DeFsylJz97uNeYvKNdd8d+PdmwfrsCESo
MCHyz6EP57dA31qWWtJMyqTutv5JigPNs1MDg6MD31JDfPMAAdjuo1OzqybB
b72faGJ6L2vTCF58NB8n53K2ab.+5d2v12HPHvMQ5buoBouNL5i8k5Qcl5yX
5O6dBnXnE1ozqkY4vGLsyC8rZVLlvNghQ44HQqKFofcrUXH.fFFdbpbp3ume
ceYvI.b1SlukMoW9TXEcSwhFTUa7fbnoYG11TxkdsDaYPIcAf4ahmVMt5Bnr
5ClrwWDti5NERorv1j00FGOAvbO1LqaOmmuU5cqKmm2WDF35+wyrWnapkg2R
44ao77sTd9Rk6HVCOcpwS1l.roFU8l5IOx14pUVEiL8MJDICcdmox33VI2Ny
u6f0XtlnPGWcrCXj.XhyCrkZbI4blwn1fAdVoQu4.O2dNMRjlXcQ7sKHXMUg
DYEepqWPPuf3c3VgAq4JDeuOemrKHJUED6XkPr1pgXiUDwtm7qtFOjwD3YBN
h6RcoHV5opnM5tiYvXSJFYImrZQgnZJchWOo2t8RJuILhRi6l5qibS8zaeTw
Oo8Z+rv.ejqT3qsr45cmZtlZsvkfQZepPUK88jThGBQjDoYsOFyFuE+XKK9O
q3rNjydwgmFn57wtL1uGxSwGrRlqqDKkDrFx8PcMC5ttQbFCiz+ZrajgDDqQ
PGiHuRPc+j3EObp8oAAPFbhBXHJkIHrLblcU0uTTuwvU6V8RpRpmHhQghXpM
wvTtzSnjDiIbfIe7joRjr144v09GybgdBWXfvE1bX+lbY+DvnJOsi9KsDTzr
eSax2ok+Bp0GpM4G0V7kJ.+o1nOUK4WUH9Vsj+UkYU.lLcVMFIj569hSAM3i
UH9YEjuVs6uUn9bsM+tBz2q17+pcevZ2Or17EK.+wZ2mrP7KKHey1I+y1lOZ
s6mVK9p0h+Zs3y118aai9tsA+21jObg4G258kaEoLkEmW8MXuL3UtFAl1OrB
VjUwEUxOq7dKTKEIkuaSF00X7sZ7ArDMi5hjt3r.8bec1lVt3J1tpDs5kmR6
OVZOxy.3Zvzvn9CotPfThotTpib.F9aNTzn7gy7cIwO9X+gRNDnLq3vjgkte
ygkrjmczotuS5o9Dr4ywaTW4Y0tBkUvqafKAB3RypWgljaHqCKbMAuUURtTW
7v4Ampbn0eXZ06N3JmrFCDWqwDLOzLtmG0SvobAiwo36cbquRi0YPlLY.Ykt
kJMrV373lf+ztfn4e0NfaAiAw7RPlCiQwuxPRDiP4dToGillehnABkwSFT1n
WfR6HkUcNeuQjAy46U50uNMZYNHxv46UJcFuL9oulWhh8ojEg5PsweyWaexA
61gE.XOwLBQJb4G81IUP+VatYt+149KJd7avsIu8iirwijzDwRoAlHGQIT1Q
mYj01bAfvjoBBWomd1e3i1E3qs95YU3BYAtJe678wxFU8rldZ4DAssfydcPP
vYDnaW6zCn84a+ypcc0JdhC0AVVIdlWQC0.Cq1zaqMXEfcof4jK4LmJ0ZvZv
75JUU0g7UKEZUqabsUDxJ2iR6LyQ+GntcYfpIKb02cwVkjmfvDHljgXXltRN
fPtXBx0ySwAQ8LEIBhKixvdddtDtqHq5NbAP.JoSHPC30UMBvFbDvD6SY+5h
RWXnzUS7AW0yr7Nu51JBn6E2XONUQkTScZQ8CUh7XH0EMXBCinTJhSHttb2K
H.30E.noEBmKVf7ughGgisBUXkSxpmNz1RgKr1hWXYZ4rbVjYV+uX2lBU6IH
pT1wd+CxbJXo4hmZJV5JY6ir.o7p5Du2lDm1Df1dRwoXdKFk16yUtlB7FdVp
o36cUha88SxtEJWmUr6ofnWb998Lao7DmZPC.7PvEJni3I3BDh6wkZnCVg3F
e6Hw65QhuX65YpY5lNVbWieqMtRUZhdNNCc6Twucp32NU7qySEeQP4vfApKo
Zbk2aZo6c07wMZ7aMVKZQj4pZJu6tVoMoF.7nUAoE+9cyCW367CwQaiWFbWU
khMYrvc+SgJMK17UGyp3uSwD81eyOs4mh92SCv5CZn8cowAp1bP03+2hMukz
+OLZ9lfTG937a0JI+9CejiukGc9trlc8eyuUof9e0eky7mSyyjuKLB+Vmeq5
C91eV+lSxdapqX9zI+eo+SPZ68K22xuyA8Vmxehr+3cNe2ggzay+k7SIo+VA
eX0ad+2j9Y+ataxbn.gQE61b8wGq3tvyRN5hU.t9Gcsb3JpAUM.YSZvYAjT1
ApQI2g4HnPSIGc1E6qEWL2bJ5j6XAMLKgxAKBSRm7KUubMsE9reUxG.igQys
1vS2tL9yNAQeJXoROrlaTtjlLZwZFKR2KNPLiWNN74x8UqoqitlqRisSAqrm
VinuUfpTycRkWlD63+o3vENOtzW81W33mj3O+ia6N5Qs2QOMBSEH7LAwixER
JMy93rTF6ZA81N+4UgIJzyTrY5NXQbglS4LNdVZwzjHjbp9P6yXBuVvJ+GTp
htKIqqLq3yhi6Cfws6zOcU1hvSKB4J8VItdTBx.YhqK1qrClVs1ToRa7mBb9
6+gdfX1c0mvalIJPbmw4HWSLg3RmZXUad4SYstRbkyiECdupZOa7KpcGjRDP
QMFVsvzEKYDL0S3Z1Ev0UunsK94a3xN6yJvm77lfsO+hh4wAehInIFIA5Z05
xGdLULi+pHWA2DmJF0.7N62.WKOMKKubzUj.68.1x+e+atxTSIcDqlW3HON6
v4rP.koxh9MqHlRyJY9AyeySoQToYxQunXly2albdmKx46ylgvpa88oSSlJW
VmaXu4LTAWyI12jtYXnSjE1JfJ5bmVcBKMq0sb+jRGkUANKBe7wfMosgL+kN
8XOWqJoPLZ0oT9cFsb1861YE6fi1nIURO+v1SLkmkP2w.KHyntdBAQcAsoFT
2wim9rucQk3yrG69Bs7HvZKxLa1kgsAQ4CavBihZBPvgBxpFjkEPCOnmxqjU
wJVPXCF3hz9iMiVyYNUOsN0hVGyFWK3NWqYXSF2QC3V5K6Lz3pkUhgtl5Eyn
pC6gOvoFgNwK5D5UvWEDn.5T34f6lH51ZsIRuOlN9K0nXnySdsxHZSW9UgKV
GGFkjIK20zh4vLth0TJDTz9.MJSk8J2YTQAFX1CuQeFgAdeAob7GLf4UwzKT
i5tCCwKUeWG9Pj41xPjNMFh0HHnzp6rXJgHU1h6lFy1XBCQn57UcueUpduFn
ZxXR0f2WXO9LlCFvpfgEWp0ZfGhbzkZslGXTjeoPQuNgh3K.JxQcSgIzjdLR
vWnYZOvJAKGQ07aQRMC4ZJN5rYdBjH8icLjuMcGmZt2nJFD7hi5Tpu+Zjlo2
IUp9GE9g3RyOtoIRT1ohUyMGcnfB1lb7Ef8wTngwoalW3zSLMTnFut4jtwlx
rm9UiNNR.uK+k.GyRcbFcFkPHBtPer1Ykpw7WcTwI4TxJYIbe0IZU2r823Xx
XXKw194d1YMrjHaURhs8iy9zUgB9yK7merIQUaiNn11bpRheklZnR6asndSm
xfauYXsLdt+xVZVPmP2OqXYsuCoJVkxYeswKe8sAfVZbTbgo1+i.2Pra33VZ
33CaoL0WtD0O0wSID3jZNVqavoU3zEBblkgT2fS6q1A.mLI5pl6rwVM2p3zy
haKDftk+p9tuGj1uW1NLo8Qkb+fwr9zJ5XutaEc7r5rYlKzMGqsrXXWbqUzc
RshNzsVQ2UPqnKJNEaFzFRWaQ8yiKmE4GYpNTu37dcPvYhu49VIH.zJeLQjA
M6Pwtppc+i.dAn4yXp+ZJCvugXkq7v02j6w2vqNTiLXlscugW.a+lb1zFtZO
e7R6VWNwO57PXzgh9aXv1FC9Nj83TNSsMiRbXWx9ZF40QdisaUJbrx+onvjc
pc6euuZ+Xye1LnXWpsmXe04dZBJsGykROnaK4QIfiA0gKNKYma7vE3tNWF3f
dtgCAvMUtLvA4bCGbX6YbYPiy9FxIaTlQ2mrInvASIjdHpjf4BOAwUevTXye
xQlaU6kZKmajTvJa6YpKUjqLcgB9xZkZPo0yEmekSoe+V8+P5qdQd.ZocbS1
jaxZFrG+JC8VD9ovE13baAgbAydYBRgt22fmDrW0xc0aPCflFbyo3k0.uO01
W6YG0TpX1edJITdJtokZv7txPmTu3ss23iDr+SXtYkZvqr0bpg0tS.evPWdo
CEHE93dkgOSVsM7XP8T02TJabhpY3ZG0LtNduVFbuWWZY3RgxV8sjRF8FsH1
YmD40tfPdEocgvCHuzqSkKDRfvyqScKbQ.Wac0pZwIw8XuGvyLE1oLQOWcbO
8YirwVoKFB5wcckt63jUcW6cj98H+Up5toc8gkAa7SlZr7D6xYzXsvDWUdru
82DezwbJPlcg60or8s+oMI8dmOpzN5XzHPdH97up3HOQS3HLn3iwWmWcaRcJ
1jPnPAmWk1jPv.gmWm1jXuSTkoz80pMIKChdJ44diO.JlVdYYznQzi2q.Qyi
s1B1qmaYwT1MaRFXfW.D3uYRxvh6VUvwyzidnraVjLPPNv3VM6XSdU3ro5Rj
SKokfYWu5xVtuc031tJAzrfYkbUpv8ofMtv3aHBzqwyGAH3X3at5D4rQs659
bJHLZ8tDm2uct+x.GSQmtuVnXucuJyEmS6Cj7IDrAnBG2hAHv3YxZ0WSiJa7
XU8dkfKQcxwuJMJAWkF8NCCFvUBKuwujQBuHxL9U2TOvCFO2wev.txVIQid0
0wCdwFDM9HinSSSiLx.tXvJOC7LfGKVJbUsWKzNKkXTvzhm3jHlxEiTlt.TP
xJUYDJ4vKuPUDPPfPZj8MVffmmvFFfmDU3uQDCXDSCOtvq3rrCIgoqsNou5B
TUZcAKAxkN5hCcAq0EezkFlFdhvFKhwu9z4BVoq8C6wbvzk8IF2gh.rNWhwW
yXgWmVJMxCFvikQenvAuW732HTfWDtI3QeUcGpC5iudwLRmJO43oReIPL9HC
Z5zHeXSnNmCU1IkrsUUZc0MJJhohUfDhCu7Bn4LbRqN8mZkzby8pKf9vvor5
jN0.kYBhwzBK79WcAL4qCTFoiyYLOrs4LxTnG9fGeggDvRfniu5eDZmrtajG
Lf2yhL9p+0gd9k3B0RE53PDeA1F.rJQ04g9xdDHqmwZBtM0FbrCubJSY0Y4Y
qTlqwyGsQX3o.gUCOUqzkz57EcJPVjtNew7H1HrwbuMrWmrAxxLlzzqGnbSv
5YzkrvqbQYQJAUWhfQjKxxO3TM1siTcVxxl+EYzbZbiZglwSht943u8HFra.
jiumFwrN026FWKuwzNwYNUZasRr0kIbcu.f5YJ6eXD4vKuDB.vcZCXXTVV.y
RO7pKwpbbm1opCDFmfrQXjIAggs6R.CgIH5TTW5c3EW.MK5.cQgRWYEaL5gW
z.cwFS5Br4Exyv9MmR2iqVLdu1aF+TbwzdqKZaCgrLUp1i+cCzjXRPSxtMW4
hQWRKifGDYcirv1M3iOEHKIpizEhHujlni6Zeh1Jg4YrGxXtfm6kY6FvRBgN
aQ46K1T6eUG2sIWLjpaHDMEDoGaxH6VDF+i5t7wG9WBh1kKjU22URZ.Nd3oG
CWt7PXk7lpuk2b29X0MK5StO+8JDPJDICi8RCHEJl5RD5+hl14LKqWU1GDu+
Sx3LIRmHKLcZrn+KOBkS0g2RceVxwmJBKMOqzregY9K0kvUiJl2bmezSYM.E
2bstCCqyl30waNzVXlQkE+j6RheZi+hvr.C93Rs7Se46pMZ1wMocdnps0iTt
G0zRXguo5YoOLM+62M2G9TagoFlKWwIduFtndRyewwHgrlPGxe9b0ftvmmRT
yPZjUmiQ5+x0bsJe7fkAqJ+4c8DBW8S0Cg87L+k5qhTWpJ8vSEa6jHF1U+vU
CbVZKIM8yRjluOtn3Gdq5wOOcRr3ymnHUOSlR4RLbUdHTZGXpxiOMt90e3OD
Fk1ujBNvoklGVBCzsmek35wn3ZIhNrx5.eUaKt5BadwkWdGVT44h3ddl+p9A
d4kWh8y2hzebM+k4q67s5JOvlu8JoDM+usNHx4G8i157iAqBeHd4h6Jp47Xs
Z7iaWFpDj+GRa6U8ZUINKt8N9qyy.ek+WD4atVVGy0HOvvBcX8faFKY6KkKQ
vnzFeKnEdUizwyENUN+U5jLW09WDYdQkohKXb7YlFVErcq+ScfGsCyCigDub
hs1Kwo5eYSrkqWp5C2W6e0vm8pTghZVL2KljJcsftvoqYIxoCmNIjOib4ljy
6ua4tfB4LV2EqQwtBlYOtCTiL8R3yO0rI9yQmH4bjH3Gj.QuLjyO7U+SkZTl
+xYlsaboThQb.AiwzyN07OtIH3jIG87BM8miYv+YmR9uCVtL9ymHo3oLkv0H
gE6g06OvUyU0HpsPGi8Op6tlbIB4clHYcC28eToi3IRvoqgPZVQdZoYPS637
aQbNnlMFEPSImdJ2VvDBpLacjalkUp4MI0tUYSBH3HBzkIzRVW5hQ5QujxDH
Yl0ktXNLKimD3Ph+Z3jeICn989IwN+9zd05.Z3z9uoieM2k11VW7ASmp8C9I
IaBeXWhw2V4GgcsKp9zx3G7Wl0gTO.H00kUOLhFgNLcZehM9wGs1goqsGRKY
s0Co4ByAqX5gz0T1HZn2bWe5s+JsqQKPLyQCH1Gxy5CvkeqqQeqqQeqqQeZc
M5p9WHunwfMoaYptuyeIdD6VVlpClob3c00LwNsRqi8JnVVcRgJml0VmSqHw
PAVttlPEIloJiDAXQZJa6yqphzz5z1X7ePK.dz5W8XgYKXoVLj3TKU0UdZAM
6X0lTeqIU3J+ETqpbMoNWKpzAPstFUsq5g8XUEuRp4wMkCKixdY0vdlW0SPt
VM8fnsGHM9rq0GTM+ZS6OfZ.ZSKP6ZBZWaPaZDBPqP6ZFBQ6PPZH1IsDaSSQ
6ZKZQiQKZMZQyw10drQMHaPKxlzjrKZSVuFkUD0TVdd02f0VpbEYyUtYy6dg
PsnJP6RnqQJMp7saRsflTMn748TbSvSG7ptoXE0FpofoBYywJ9w.9j.t0tUP
6pQzppDS.zrgJsbePTuNfnn15XKclsdRgn0TH66CZ5BAMwLj8ZZ+DFNqXLUe
PJLBNiGi0IwoHK.03HNsj9PUJMm0f60EIo0g9.pAjsEqrkB0+Nydz6goWWFl
xK1vzsqCS7EYXVS1czoHZuovtepRAtVn.j0EjMFfeVNnnJl91SUfsbvQ0d3Q
0DaNmC2573FkPP+uru.Su1ei+ps81YpDfNSMKqC3Wac41d2hQxWtxQ.8pHFc
U1MP7mOe2JctK7hSZFE2OVINvRdelCwvH52TtdVBckzkv2yjAGqpZEdAv.pe
3MkrdPnQKRUPS6x0tG35iAe7KpyhoTlmxgW6sISnJc.+LTbAfWpXPSnJcfdX
OtLvRxTpR1ClA16Lv.S6TQ..R0M.SD4JuPDwkI2foCdwMHqyoe3uuDYmNsqs
GBnyVXhQQqhuJqzh3ZlJ6QkE4VlndKSTukIp2xD0aYh5sLQ8VlndKSTukIp2
xD0aYh5sLQ8VlndKSTukIp2xD0aYh5sLQ8VlndAxD0GWFqDgzb5ld28N28fR
gvbIb5iwaV4mjOZ8KDU.xVyAUoNiOnlVmXVbz1bNnhqOGTIGeBaTvSRvlOXB
94iyY8Jqb0xRewZt2duF1RgfCi5zvonvHqtn0uPbAfgjptRcc0rlPwrgT0sg
DInI3q4DFHkTyhR7dCm6iz6CemeJL3y5n5ObYXxWKL+F+3iaCxH92oI22k2I
y2YxxhEaTb9GBu8CT27kgy+Xxyah28zyEtQk.luzc9TM24gm1OWd7RGcuM7z
x1eySYN4lTatZyaa9mxMMYSsm4IX8Ac6Iab9GYIUsOLGuHXU713cQKloloTO
26JM4Z99KdB1YS1kOih8S4V5OevZlCkO.gZY2.OBJ+kCnK9bZOPICxCTT7nX
pbLLt5BdLSXJTv4d0fLFEoyKDaiQYqCQFBa7YZ1OtLkkXLWSCYfoG2EeEeey
Zfe3UCGwfsQLtsRLoEHeZ9eRMgrUhAiySL3AclgZiXJc7jUlZROotb+fSOGf
VoFSJtsmZ3xgjZrtVn8kBYCJlflaHpe0fLD8.sZMWvHbZONJA5yqNAh3y1Hf
LhhjwTPxjwT5fITFSwvej3A5QRf8HYC3ijC+QNT.Krk3TuA7Q5B6QJGvGIHw
D38MuoA4QJAo3Be3dhLPLrLwfAq.kCvFS8CgMFJuvZPkHKAJNDOb5nBTZHYv
dhLPhIFvGHLQgC3CrBa53+Dory9SjdlehcZVbTDP3BRzu6vsYCnUJkGUCp.I
PftGonYWWnQPcX.YHDQBZyPugaZ2Cz98xhAbaYikDHsuTYFqSy+pgY8OBjLt
gSWOXpPiPCmDNLrGIFOfORX6FiGRpDlJG3AzPALLUnw7gSmCPlZK.48gBt6I
aAUE4NzgPuEPhAFt4EpDh2yFNaTYfkxYeRgPy6cTBc1XsY.LibIs6vTgGaFC
KPLQla43ZeLZRf.tghJ7Jt9Ta1Se5Wc9lBnhNLETOo0vDxPne.LcsIdsRCtB
5LYmlPn4Y3zu57MgTljacBodRqgIDT8G7So56ldzWtdtUtNtUS8aqk51Vk50
V1woWJsPxfSaoCRovB3mObFdUBayCmEeCo9QuR6i9lxG8McOZHxLaIpLaJhL
yEx.2umys13gnTrPTjautz3ngojdk5FmPZabRorQuSWiSKUMN8zz.Fmek.Ut
mAKbeSIidlNFCJ2ecg.Qqodw.sRolTrv9JlFSqhAZPULhqaX7zuTl.b5Rz4T
kX.o87oDAH4WPRChAb7ULcGZXD18Tb3jjVz8TZnuoyPekNcF2rrzBHvSrERQ
AHbdVSKgAhqqlzO.tXAXobvvNRKjZAvGpvRmfAcnVHsAfORgkp.C5HsXJAzg
gpsz.XPGkkB2+Nr6EnP72R38OTjR0v3GNg.KdsGnQZ0PzGhbKngke+BI+Ahz
pIz6gPaPC29K9T2wPpuAxxZXz2iARlmEZn7q+K+Ep2x+O0d2kDC
-----------end_max5_patcher-----------

Hi @james.bradbury - I’m not sure I’m seeing the stuff about non-symmetric windows? I understand standard windowing techniques pretty well, but I’m talking here specifically about windows that are not symmetrical in time…

@Rodrigo - yes that’s the sort of thing I am describing and yes - just us three for now but simply because I haven’t sent anything out to anyone else and James sorted this out.

@Rodrigo - shouldn’t that look like this?

2 Likes

There is a ‘exponential’ window as referenced in those papers put in there instead of fl.window~ hann. As far as I can tell, this is a non-symmetric window. I think I misunderstood what you meant though. Did you mean that in an implementation with multiple windows you change the window as you slide along the overall frame?

Yes - the exponential window is exactly the sort of thing you might do. It’s also possible not to window, but then the leakage/edge effects may cause issues. The exponential window starts “on” though, which may work well for strong attack, or is may be wise to add a small rise. I’d be interested to see the effects of warping a standard window like a hann also to move it’s peak. It’s just a case of knowing how it will behave in the spectral domain, which I guess we could just measure with framelib~ by zero-padding the window and taking the FFT to see its spectral shape.