Hi all,
recently there was a discussion about the different ways of avoiding aliasing. One of these is oversampling, but as far as i know there is only one example of it in Pd help patches, and it has a phasor in the subpatch. I m assuming that i could use the Butterworth filter in the subpatch without the phasor, am i right? In this case would it make sense to use it everytime i need i low-pass filter? This is probably a very silly question...
Pierre
The frequency components introduced by aliasing cannot be removed (at least not by low-pass filtering) once introduced, and this happens at the level of signal generation, not at the interface to the dac. Since it is the phasor which is actually aliasing, it needs to be antialiased with a low pass filter before it continues on the a lower sampling rate. So the signal source and the filter need to be in the same upsampled subpatch.
Best! D.
On 4/12/10 5:49 PM, Pierre Massat wrote:
Hi all,
recently there was a discussion about the different ways of avoiding aliasing. One of these is oversampling, but as far as i know there is only one example of it in Pd help patches, and it has a phasor in the subpatch. I m assuming that i could use the Butterworth filter in the subpatch without the phasor, am i right? In this case would it make sense to use it everytime i need i low-pass filter? This is probably a very silly question...
=
dear derek, dear list
is there any possiblity to make an antialaising for the dac output of
a mixed signal?
thanks in advance der.brandt
Zitat von "Derek Holzer" derek@umatic.nl:
The frequency components introduced by aliasing cannot be removed
(at least not by low-pass filtering) once introduced, and this
happens at the level of signal generation, not at the interface to
the dac. Since it is the phasor which is actually aliasing, it needs
to be antialiased with a low pass filter before it continues on the
a lower sampling rate. So the signal source and the filter need to
be in the same upsampled subpatch.Best! D.
On 4/12/10 5:49 PM, Pierre Massat wrote:
Hi all,
recently there was a discussion about the different ways of avoiding aliasing. One of these is oversampling, but as far as i know there is only one example of it in Pd help patches, and it has a phasor in the subpatch. I m assuming that i could use the Butterworth filter in the subpatch without the phasor, am i right? In this case would it make sense to use it everytime i need i low-pass filter? This is probably a very silly question...
=
-- ::: derek holzer ::: http://macumbista.net ::: ---Oblique Strategy # 45: "Disciplined self-indulgence"
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
Sorry, the only solutions I've ever known is to antialias the signal sources by low-pass filtering them before they reach a sampling rate where they could alias, or by using band-limited signal sources in some way or another. Once the aliased frequencies are in, it's pretty near impossible to get them out.
Best! D.
On 4/12/10 11:10 PM, brandt@subnet.at wrote:
dear derek, dear list
is there any possiblity to make an antialaising for the dac output of a mixed signal?
thanks in advance der.brandt
Zitat von "Derek Holzer" derek@umatic.nl:
The frequency components introduced by aliasing cannot be removed (at least not by low-pass filtering) once introduced, and this happens at the level of signal generation, not at the interface to the dac. Since it is the phasor which is actually aliasing, it needs to be antialiased with a low pass filter before it continues on the a lower sampling rate. So the signal source and the filter need to be in the same upsampled subpatch.
Best! D.
On 4/12/10 5:49 PM, Pierre Massat wrote:
Hi all,
recently there was a discussion about the different ways of avoiding aliasing. One of these is oversampling, but as far as i know there is only one example of it in Pd help patches, and it has a phasor in the subpatch. I m assuming that i could use the Butterworth filter in the subpatch without the phasor, am i right? In this case would it make sense to use it everytime i need i low-pass filter? This is probably a very silly question...
=
-- ::: derek holzer ::: http://macumbista.net ::: ---Oblique Strategy # 45: "Disciplined self-indulgence"
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
The simple but expensive answer: just oversample your entire patch!!!!!!! ;-)
D.
On 4/12/10 11:10 PM, brandt@subnet.at wrote:
dear derek, dear list
is there any possiblity to make an antialaising for the dac output of a mixed signal?
thanks in advance der.brandt
Zitat von "Derek Holzer" derek@umatic.nl:
The frequency components introduced by aliasing cannot be removed (at least not by low-pass filtering) once introduced, and this happens at the level of signal generation, not at the interface to the dac. Since it is the phasor which is actually aliasing, it needs to be antialiased with a low pass filter before it continues on the a lower sampling rate. So the signal source and the filter need to be in the same upsampled subpatch.
Best! D.
On 4/12/10 5:49 PM, Pierre Massat wrote:
Hi all,
recently there was a discussion about the different ways of avoiding aliasing. One of these is oversampling, but as far as i know there is only one example of it in Pd help patches, and it has a phasor in the subpatch. I m assuming that i could use the Butterworth filter in the subpatch without the phasor, am i right? In this case would it make sense to use it everytime i need i low-pass filter? This is probably a very silly question...
=
-- ::: derek holzer ::: http://macumbista.net ::: ---Oblique Strategy # 45: "Disciplined self-indulgence"
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-
brandt＠subnet.at -
Derek Holzer -
Pierre Massat