Friday, 6 February 2009
In January I traveled to the Open University for a meeting of the Musical Acoustics Network, which has run out of money but is still meeting. In between sessions Jim Woodhouse pointed out that I was wrong to say (as I did in a previous post here, and repeated in a discussion session, a video of which you can see by following the above link) that delay line digital waveguide models can't reproduce dispersion and hence inharmonicity. And he should know, he helped to invent the technique in the 70s. What you do is let the wave packet propagate along the waveguide without dispersion and then pass it through an all-pass filter whose phase response is designed to delay each frequency by the amount it would have done as it travelled the length of the true string, or segment of the string if you're simulating bowing and you need to know the waveform at the bowing point as well as at the bridge. This led on to a discussion with Jim and Murray Campbell about the audibility of dispersion and nonlinearity in plucked string instruments, which we continued by email, and in the course of which it cropped up that nonlinearity is definitely audible in the Kantele (a Finnish folk instrument). The biggest surprise for me was that my Voice Recognition software recognised the word 'kantele' repeatedly and without difficulty, although 'Nyckelharpa' came out as 'New Car Park'.
What a long lay-off, largely due to the problems I've been having writing. I've been trying to get to grips with the aforementioned half keyboard, and with Dragon Naturally Speaking voice recognition software, but neither is anything like as fast as I used to type two-handed on a regular keyboard, so just keeping up with the day job has become quite a challenge, let alone extra-curricular typing like this. But I've been spurred into action by being linked to from the ISVR outreach page. I intend to complete the diary of my parliament week from the notes I made at the time, but they may be interspersed with other things.