If you're worried by metro's irregularity then you should use Eric Lyons' samm~ (sample accurate multiple metronomes). This is described in http://www.sarc.qub.ac.uk/~elyon/LyonPapers/SampleAccurate-Lyon-ICMC2006.pdf
Hallo, schiemer@uow.edu.au hat gesagt: // schiemer@uow.edu.au wrote:
If you're worried by metro's irregularity then you should use Eric Lyons' samm~ (sample accurate multiple metronomes). This is described in http://www.sarc.qub.ac.uk/~elyon/LyonPapers/SampleAccurate-Lyon-ICMC2006.pdf
Note that samm~ is almost exactly as sample-accurate as Pd's metro. I tried to illustrate this in attached patch. samm~ however is also available for Max, where the metro according to Eric's paper is not as accurate as the one in Pd, so if you need to move patches between both systems, samm~ is a good choice. Of course Eric's system has some convenient extensions to generate polyrhythms etc., but timing alone IMO is no reason to exchange [metro] with [samm~] on Pd.
Frank Barknecht _ ______footils.org_ __goto10.org__
If I understood Erics paper correctly [samm~] allows you to specify which sample of a block is set consistently, if you like a periodic version of [dirac~]. That's surely useful if you want to do somthing in the signal domain like convolution with a impulse train.
What I couldn't figure, is this really equivilent to
[phasor~ 0.0173474647] very accurate signal phasor | [>=~ 0.9999999999999] "floated" 1.0 so we get an equivilence
which should give a single sample set on the top of each phasor cycle. But I seem to remember this not working properly for some reason.
On Sun, 18 Feb 2007 13:53:52 +0100 Frank Barknecht fbar@footils.org wrote:
Hallo, schiemer@uow.edu.au hat gesagt: // schiemer@uow.edu.au wrote:
If you're worried by metro's irregularity then you should use Eric Lyons' samm~ (sample accurate multiple metronomes). This is described in http://www.sarc.qub.ac.uk/~elyon/LyonPapers/SampleAccurate-Lyon-ICMC2006.pdf
Note that samm~ is almost exactly as sample-accurate as Pd's metro. I tried to illustrate this in attached patch. samm~ however is also available for Max, where the metro according to Eric's paper is not as accurate as the one in Pd, so if you need to move patches between both systems, samm~ is a good choice. Of course Eric's system has some convenient extensions to generate polyrhythms etc., but timing alone IMO is no reason to exchange [metro] with [samm~] on Pd.
Ciao
Frank Barknecht _ ______footils.org_ __goto10.org__
On Mon, 2007-02-19 at 04:39 +0000, padawan12 wrote:
If I understood Erics paper correctly [samm~] allows you to specify which sample of a block is set consistently, if you like a periodic version of [dirac~]. That's surely useful if you want to do somthing in the signal domain like convolution with a impulse train.
What I couldn't figure, is this really equivilent to
[phasor~ 0.0173474647] very accurate signal phasor
hehe, i think the number format used in pd doesn't allow such a high accuracy ;-)
| [>=~ 0.9999999999999] "floated" 1.0 so we get an equivilence
which should give a single sample set on the top of each phasor cycle. But I seem to remember this not working properly for some reason.
not each edge of a signal generated by [phasor~] lies exactly on 1 or 0 because of the grid introduced by the samplingrate. see attached patch.
roman
On Sun, 18 Feb 2007 13:53:52 +0100 Frank Barknecht fbar@footils.org wrote:
Hallo, schiemer@uow.edu.au hat gesagt: // schiemer@uow.edu.au wrote:
If you're worried by metro's irregularity then you should use Eric Lyons' samm~ (sample accurate multiple metronomes). This is described in http://www.sarc.qub.ac.uk/~elyon/LyonPapers/SampleAccurate-Lyon-ICMC2006.pdf
Note that samm~ is almost exactly as sample-accurate as Pd's metro. I tried to illustrate this in attached patch. samm~ however is also available for Max, where the metro according to Eric's paper is not as accurate as the one in Pd, so if you need to move patches between both systems, samm~ is a good choice. Of course Eric's system has some convenient extensions to generate polyrhythms etc., but timing alone IMO is no reason to exchange [metro] with [samm~] on Pd.
Ciao
Frank Barknecht _ ______footils.org_ __goto10.org__
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On Sun, 18 Feb 2007 16:46:39 +0100 Roman Haefeli reduzierer@yahoo.de wrote:
not each edge of a signal generated by [phasor~] lies exactly on 1 or 0 because of the grid introduced by the samplingrate. see attached patch.
roman
Ah yes, I remember now. I had to reduce that to[>= 0.95] or something and then you sometimes get 2 or 3 samples set in a row.
So another way I came up with was to use [z~] to delay a squarewave by a sample and then subtract it from itself leaving just an isolated value set to 1.0. Afair that worked okay.
Hallo, padawan12 hat gesagt: // padawan12 wrote:
So another way I came up with was to use [z~] to delay a squarewave by a sample and then subtract it from itself leaving just an isolated value set to 1.0. Afair that worked okay.
Nice idea. You can also just send the square through [rzero~ 1] which does this: y[n] = x[n] - a[n] * x[n-1] usind a=1 then. See att.
Frank Barknecht _ ______footils.org_ __goto10.org__
hi
i don't know to which mail you are answering, but i still believe, that pd's [metro] is absolute accurate (even sub-sample-accurate), when used in combination with [vline~] for example. the only case, where it could be considered inaccurate, is when cpu-load is too high, though when writing to a file, it is still accurate.
roman
On Sun, 2007-02-18 at 11:21 +1100, schiemer@uow.edu.au wrote:
If you're worried by metro's irregularity then you should use Eric Lyons' samm~ (sample accurate multiple metronomes). This is described in http://www.sarc.qub.ac.uk/~elyon/LyonPapers/SampleAccurate-Lyon-ICMC2006.pdf
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Telefonate ohne weitere Kosten vom PC zum PC: http://messenger.yahoo.de