I'm working on re-realizing a piece I wrote for subwoofer that is meant to be felt by touching the speaker cone instead of hearing it. This is done by using sine waves from around 15 to 26 Hz at low volume adn utilizing difference frequencies to get lower frequency pulses. My problem is that the way I have my synth set up, the osc~ objects end up receiving midi messages a unpredictable times in the period which is leading to interference and thus unintended dynamic shifts. Another problem is clicks on note-offs. I can get around those with [line~] or a low pass filter but it would seem more elegant if I could solve both problems by generating sine waves with finite numbers of periods that always started with the same phase. I haven't been able to figure out to this and I've thrown enough hours at it that I think it's time to ask for some help. Any suggestions would be much appreciated.
-Daniel Konar
Could you be more specific, im not sure I entirely understand the problem...
On 8 December 2010 08:32, Daniel K. konarsonarsmonar@gmail.com wrote:
I'm working on re-realizing a piece I wrote for subwoofer that is meant to be felt by touching the speaker cone instead of hearing it. This is done by using sine waves from around 15 to 26 Hz at low volume adn utilizing difference frequencies to get lower frequency pulses. My problem is that the way I have my synth set up, the osc~ objects end up receiving midi messages a unpredictable times in the period which is leading to interference and thus unintended dynamic shifts. Another problem is clicks on note-offs. I can get around those with [line~] or a low pass filter but it would seem more elegant if I could solve both problems by generating sine waves with finite numbers of periods that always started with the same phase. I haven't been able to figure out to this and I've thrown enough hours at it that I think it's time to ask for some help. Any suggestions would be much appreciated.
-Daniel Konar
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/12/10 07:32, Daniel K. wrote:
generating sine waves with finite numbers of periods that always started with the same phase.
10 periods of 20Hz sine wave, subsample accurate timing without needing [block~]:
"0, 10 500" | [vline~] | [- 0.25] | [cos~] | [dac~]
For N periods of F Hz, something like this should work if my mental algebra is up to scratch: "0, N T" where T = 1000 * N / F
Have fun!
Claude
Hi Daniel,
Sorry for the late reply on this subject.
To do exactly as you describe it is possible to take the cosine or sine of a line over period.
Use [vline~] as it is much more flexible.
Take a line running from 0 to 1 in some time, and use the cosine function to obtain exactly one period in the same time.
Note: You don't need the 2 * PI with Pure Data audio objects, they are adjusted for radians (rotation normalised so that 2*PI = 1 )
On Tue, 7 Dec 2010 23:32:48 -0800 "Daniel K." konarsonarsmonar@gmail.com wrote:
but it would seem more elegant if I could solve both problems by generating sine waves with finite numbers of periods that always started with the same phase.
-
Andy Farnell -
Claude Heiland-Allen -
Daniel K. -
Ludwig Maes