How would one (and how hard or easy would it be) write a patch which does the following: Instruct the user to play a sustained note on the violin; then give feedback about how accurate Helmholtz motion is achieved; Could we measure how short the transient takes from nonHelmholtz motion to helmholtz motion? How about changing from a note on one string to one on another, ... Could we write a patch that trains the user to minimize install and end times for helmholtz motion and to somehow show a level of periodicity during the systaining period?
And a different question: Just like how the different non harmonic partials are shifted and forced into a harmonic compromise in a bowed string, does something similar happen when playing simultaneous harmony? I.e. would partials on 2 strings sharing bow and bridge, say fingered at a frequency ratio of close to 2:3 but not exactly 2:3 (i.e. equal temperament), also modelock to the just(=pythagorean/integer ratio) interval 2:3? would the one out of 2 slips on one string be simultaneous with one out of 3 slips on the other?
On 11 December 2010 20:29, Ludwig Maes ludwig.maes@gmail.com wrote:
How would one (and how hard or easy would it be) write a patch which does the following: Instruct the user to play a sustained note on the violin; then give feedback about how accurate Helmholtz motion is achieved; Could we measure how short the transient takes from nonHelmholtz motion to helmholtz motion? How about changing from a note on one string to one on another, ... Could we write a patch that trains the user to minimize install and end times for helmholtz motion and to somehow show a level of periodicity during the systaining period?
A job for spectral tools I would guess. At the point spectral flux hits a first minimum and autocorrelation hits a first maximum, the string will be in stable resonance. You could use something fast like bonk or a schmitt trigger to time those from the onset and get a measure of "how well" the note was bowed.
Actually verifying that the thing is in stick-slip Helmholtz motion is a different thing than looking at the waveform, you need a stroboscope or very high speed camera for that.
BTW if you have never seen the images of Prof Andrew Davidhazy there's plenty of string sequences there. He kindly gave me some for my book cover.
ajf
On Sat, 11 Dec 2010 20:29:11 +0100 Ludwig Maes ludwig.maes@gmail.com wrote:
How would one (and how hard or easy would it be) write a patch which does the following: Instruct the user to play a sustained note on the violin; then give feedback about how accurate Helmholtz motion is achieved; Could we measure how short the transient takes from nonHelmholtz motion to helmholtz motion? How about changing from a note on one string to one on another, ... Could we write a patch that trains the user to minimize install and end times for helmholtz motion and to somehow show a level of periodicity during the systaining period?
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-
Andy Farnell -
Ludwig Maes