synths, filters and clicks.
6 views
Skip to first unread message
11hzrobot
Apr 26, 2009, 1:40:52 PM4/26/09
to PSPSeq
All synth+SVF combos except the wav+SVF seem to be functioning
improperly. I've uploaded a file "clicktracks" that shows the
problem. The filter works the way I would expect a filter to
work...if you have it set to low pass and turn the frequency setting
all the way to zero then all frequencies are cut and there is no
audible sound. gradually increase the frequency setting and the
lowest frequencies are heard, keep opening the filter until, when it
is totally open (the frequency setting is at maximum) all the
frequencies of the sample/tone that is playing are audible.
So in the "clicktracks" song I have set up synths that play either
kick sounds or low frequency tones, the SVF settings are the same for
all the synths...a low pass filter type, resonance set to 0, the first
four beats the SVF freq is 0, the next for beats the filter is opened
up a bit by increasing the frequency setting, the next for beats it is
increased a little more and the last for beats the filter is open all
the way; frequency setting to maximum. With all but the wav track
there is a loud click that is heard when the filter frequency is 0.
It sounds like a bit of the synth sound gets through the filter before
the filter kicks in and cuts the frequencies. That is just what it
sounds like because I don't know what is actually happening, but it
does not sound right. On some of the synths you can hear the click at
all filter frequency settings. I ran into this because I was trying
different synths to make kick sounds and bass lines and I kept getting
clicking I tried adjusting the attack, but some sound need a fast
attack. So I thought about using a filter to cut all but the low
frequencies and I found that not only did the frequencies that made up
the clicking not get cut by using the filter, but that the clicks
where the only thing heard when the filter is completely closed and
nothing should be getting through the filter. Except the wav
synth...the filter works great with that synth.
any ideas?
improperly. I've uploaded a file "clicktracks" that shows the
problem. The filter works the way I would expect a filter to
work...if you have it set to low pass and turn the frequency setting
all the way to zero then all frequencies are cut and there is no
audible sound. gradually increase the frequency setting and the
lowest frequencies are heard, keep opening the filter until, when it
is totally open (the frequency setting is at maximum) all the
frequencies of the sample/tone that is playing are audible.
So in the "clicktracks" song I have set up synths that play either
kick sounds or low frequency tones, the SVF settings are the same for
all the synths...a low pass filter type, resonance set to 0, the first
four beats the SVF freq is 0, the next for beats the filter is opened
up a bit by increasing the frequency setting, the next for beats it is
increased a little more and the last for beats the filter is open all
the way; frequency setting to maximum. With all but the wav track
there is a loud click that is heard when the filter frequency is 0.
It sounds like a bit of the synth sound gets through the filter before
the filter kicks in and cuts the frequencies. That is just what it
sounds like because I don't know what is actually happening, but it
does not sound right. On some of the synths you can hear the click at
all filter frequency settings. I ran into this because I was trying
different synths to make kick sounds and bass lines and I kept getting
clicking I tried adjusting the attack, but some sound need a fast
attack. So I thought about using a filter to cut all but the low
frequencies and I found that not only did the frequencies that made up
the clicking not get cut by using the filter, but that the clicks
where the only thing heard when the filter is completely closed and
nothing should be getting through the filter. Except the wav
synth...the filter works great with that synth.
any ideas?
H- Seiken
Apr 26, 2009, 1:49:54 PM4/26/09
to psp...@googlegroups.com
No program that runs realtime that I know of can actually cut all frequencies. Load up a really good spectrum analyzer with any common computer based music app and do the same thing and you'll see even with a filter cut, there's still audio making it through that's not within hearing range. Because of the fact there's no ramp on the amplitude to eliminate clicks, you're simply hearing the pop of the synth start cutting off the previous sound.
The only way to truly cut all frequencies is either non-realtime and repeatedly running the same filter on the sound or using an uber expensive hardware filter and even that isn't garenteed to cut everything out.
The only way to truly cut all frequencies is either non-realtime and repeatedly running the same filter on the sound or using an uber expensive hardware filter and even that isn't garenteed to cut everything out.
11hzrobot
Apr 26, 2009, 2:15:02 PM4/26/09
to PSPSeq
If the audio making it through was not within hearing range that would
be great, but the audio making it through the filters is loud. Except
with the wavsvf a filter cut works and sounds good with that combo.
It is all the others that have loud clicks with a filter cut. If the
filter works with the wavsvf why not with the other synths? I'm not
interested in cutting all frequencies, I may have used the wrong
terms, I'm just going for your basic, run of the mill filter cut...you
know...low pass+ freq to zero and you can't hear anything (even if
there are frequencies outside of audible range getting through).
So the click is the sound being cut off just before the amp envelope
triggers it again? That makes sense for adsr settings that have the
sustain or release settings long enough to overlap with the next
triggered sound but what about staccato type settings, like a kick,
that definitely ends (at least audibly) before the next triggered
kick? Is there some inaudible tone still playing that gets cut
(causing the click) before the next sound is triggered. That does
make alot of sense but even if it where the amp envelope causing the
clicks doesn't the filter occur after the amp envelope so the filter
should cut whatever is happening with the amp envelope stage.
On Apr 26, 10:49 am, H- Seiken <haky...@gmail.com> wrote:
> No program that runs realtime that I know of can actually cut all
> frequencies. Load up a really good spectrum analyzer with any common
> computer based music app and do the same thing and you'll see even with a
> filter cut, there's still audio making it through that's not within hearing
> range. Because of the fact there's no ramp on the amplitude to eliminate
> clicks, you're simply hearing the pop of the synth start cutting off the
> previous sound.
>
> The only way to truly cut all frequencies is either non-realtime and
> repeatedly running the same filter on the sound or using an uber expensive
> hardware filter and even that isn't garenteed to cut everything out.
>
be great, but the audio making it through the filters is loud. Except
with the wavsvf a filter cut works and sounds good with that combo.
It is all the others that have loud clicks with a filter cut. If the
filter works with the wavsvf why not with the other synths? I'm not
interested in cutting all frequencies, I may have used the wrong
terms, I'm just going for your basic, run of the mill filter cut...you
know...low pass+ freq to zero and you can't hear anything (even if
there are frequencies outside of audible range getting through).
So the click is the sound being cut off just before the amp envelope
triggers it again? That makes sense for adsr settings that have the
sustain or release settings long enough to overlap with the next
triggered sound but what about staccato type settings, like a kick,
that definitely ends (at least audibly) before the next triggered
kick? Is there some inaudible tone still playing that gets cut
(causing the click) before the next sound is triggered. That does
make alot of sense but even if it where the amp envelope causing the
clicks doesn't the filter occur after the amp envelope so the filter
should cut whatever is happening with the amp envelope stage.
On Apr 26, 10:49 am, H- Seiken <haky...@gmail.com> wrote:
> No program that runs realtime that I know of can actually cut all
> frequencies. Load up a really good spectrum analyzer with any common
> computer based music app and do the same thing and you'll see even with a
> filter cut, there's still audio making it through that's not within hearing
> range. Because of the fact there's no ramp on the amplitude to eliminate
> clicks, you're simply hearing the pop of the synth start cutting off the
> previous sound.
>
> The only way to truly cut all frequencies is either non-realtime and
> repeatedly running the same filter on the sound or using an uber expensive
> hardware filter and even that isn't garenteed to cut everything out.
>
H- Seiken
Apr 26, 2009, 2:21:12 PM4/26/09
to psp...@googlegroups.com
Adjust the attack to make a small ramp up on the kick...does it still happen? Something like ~800 on the DADSR envelope should work.
11hzrobot
Apr 26, 2009, 2:42:18 PM4/26/09
to PSPSeq
yeah it still happens.
On a BAMSVF i have
attack=820
decay=.024090
decaymin= .000100
sustain .007250
release=.008090
it sounds like a little blip ( it doesn't sound like a kick with those
settings because I set the sustain, release and decay low so that it
the tone would not overlap with the next trigger).
with the SVF freq set to zero using a Lowpass setting you can still
hear clicks. Have you had a chance to look at the file I uploaded?
On Apr 26, 11:21 am, H- Seiken <haky...@gmail.com> wrote:
> Adjust the attack to make a small ramp up on the kick...does it still
> happen? Something like ~800 on the DADSR envelope should work.
>
On a BAMSVF i have
attack=820
decay=.024090
decaymin= .000100
sustain .007250
release=.008090
it sounds like a little blip ( it doesn't sound like a kick with those
settings because I set the sustain, release and decay low so that it
the tone would not overlap with the next trigger).
with the SVF freq set to zero using a Lowpass setting you can still
hear clicks. Have you had a chance to look at the file I uploaded?
On Apr 26, 11:21 am, H- Seiken <haky...@gmail.com> wrote:
> Adjust the attack to make a small ramp up on the kick...does it still
> happen? Something like ~800 on the DADSR envelope should work.
>
H- Seiken
Apr 26, 2009, 2:47:34 PM4/26/09
to psp...@googlegroups.com
No, haven't looked at the file. I just know that there's no ramp and when you say 'click', lack of amp came to mind. And you bigger values for the attack? I generally have to play with it on my tunes because it clicks on any setting of the filter and I hate that...sounds cheap, so I make my own ramp...
You may have found a bug afterall...
You may have found a bug afterall...
11hzrobot
Apr 26, 2009, 3:29:38 PM4/26/09
to PSPSeq
ok, so I tried bigger attack settings(.0881000) and small attack
settings(.000100) and the click is still there, but the attack of the
click changes. With the small setting the click has a fast attack and
with lager ones the click attack is not so fast, but either way it is
there and given that the synth is fed into a filter, a filter cut
(lowpass+freq 0,res0) should create a regular old filter cut where you
don't hear anything....synth or click, but it doesn't. What it does
is cuts the synth freqs (you don't hear any synth), but not the click
freqs (you do hear the clicks)
you are right that the amp envelope seems to have something to do with
it because changing the attack settings changes the attack of the
click. But, again, I don't know for sure. Ethan I'm sure will know
what is happening.
On Apr 26, 11:47 am, H- Seiken <haky...@gmail.com> wrote:
> No, haven't looked at the file. I just know that there's no ramp and when
> you say 'click', lack of amp came to mind. And you bigger values for the
> attack? I generally have to play with it on my tunes because it clicks on
> any setting of the filter and I hate that...sounds cheap, so I make my own
> ramp...
>
> You may have found a bug afterall...
>
settings(.000100) and the click is still there, but the attack of the
click changes. With the small setting the click has a fast attack and
with lager ones the click attack is not so fast, but either way it is
there and given that the synth is fed into a filter, a filter cut
(lowpass+freq 0,res0) should create a regular old filter cut where you
don't hear anything....synth or click, but it doesn't. What it does
is cuts the synth freqs (you don't hear any synth), but not the click
freqs (you do hear the clicks)
you are right that the amp envelope seems to have something to do with
it because changing the attack settings changes the attack of the
click. But, again, I don't know for sure. Ethan I'm sure will know
what is happening.
On Apr 26, 11:47 am, H- Seiken <haky...@gmail.com> wrote:
> No, haven't looked at the file. I just know that there's no ramp and when
> you say 'click', lack of amp came to mind. And you bigger values for the
> attack? I generally have to play with it on my tunes because it clicks on
> any setting of the filter and I hate that...sounds cheap, so I make my own
> ramp...
>
> You may have found a bug afterall...
>
H- Seiken
Apr 26, 2009, 3:32:16 PM4/26/09
to psp...@googlegroups.com
Is it there for the other synths?
11hzrobot
Apr 26, 2009, 4:09:56 PM4/26/09
to PSPSeq
It is there for all the synths except the wavsvf and I don't know
about the noise generator, I didn't try that one. But the clicktrack
file has all the synths set up with the same svf, and amp envelope
settings and they all click.
about the noise generator, I didn't try that one. But the clicktrack
file has all the synths set up with the same svf, and amp envelope
settings and they all click.
H- Seiken
Apr 26, 2009, 4:22:13 PM4/26/09
to psp...@googlegroups.com
Tempo doesn't affect it either?
11hzrobot
Apr 26, 2009, 6:00:51 PM4/26/09
to PSPSeq
H- Seiken
Apr 26, 2009, 6:43:31 PM4/26/09
to psp...@googlegroups.com
Sounds like a bug then. :)
Ethan Bordeaux
Apr 27, 2009, 1:21:32 PM4/27/09
to psp...@googlegroups.com
Hmmm, just getting caught up. Been crazy busy recently and will continue to be so for a little while longer. I do know there are some click issues with the filter. They drive me crazy too - generally I try and mask them by playing other tracks at the same time or tweaking synth parameters to make them go away as best as I can. I don't know why they would show up on some generators but not others but I should be able to figure it out once I look at the code and some sample data.
If it's fixable I'll put it in the mythical magical 3.10 release. :)
Thanks for the debug work!
If it's fixable I'll put it in the mythical magical 3.10 release. :)
Thanks for the debug work!
11hzrobot
Apr 27, 2009, 6:49:08 PM4/27/09
to PSPSeq
No problem. Looking forward to 3.10.
On Apr 27, 10:21 am, Ethan Bordeaux <ethan.borde...@gmail.com> wrote:
> Hmmm, just getting caught up. Been crazy busy recently and will continue to
> be so for a little while longer. I do know there are some click issues with
> the filter. They drive me crazy too - generally I try and mask them by
> playing other tracks at the same time or tweaking synth parameters to make
> them go away as best as I can. I don't know why they would show up on some
> generators but not others but I should be able to figure it out once I look
> at the code and some sample data.
>
> If it's fixable I'll put it in the mythical magical 3.10 release. :)
>
> Thanks for the debug work!
>
> ...
>
> read more »
On Apr 27, 10:21 am, Ethan Bordeaux <ethan.borde...@gmail.com> wrote:
> Hmmm, just getting caught up. Been crazy busy recently and will continue to
> be so for a little while longer. I do know there are some click issues with
> the filter. They drive me crazy too - generally I try and mask them by
> playing other tracks at the same time or tweaking synth parameters to make
> them go away as best as I can. I don't know why they would show up on some
> generators but not others but I should be able to figure it out once I look
> at the code and some sample data.
>
> If it's fixable I'll put it in the mythical magical 3.10 release. :)
>
> Thanks for the debug work!
>
> On Sun, Apr 26, 2009 at 6:43 PM, H- Seiken <haky...@gmail.com> wrote:
> > Sounds like a bug then. :)
>
> > Sounds like a bug then. :)
>
>
> read more »
Reply all
Reply to author
Forward
0 new messages