Sign up Calendar Latest Topics Chat

  Author   Comment   Page 2 of 2      Prev   1   2

Posts: 215
Reply with quote  #16 
OMG you people are nuts

Senior Member
Posts: 513
Reply with quote  #17 
So a tabla-player, a vocalist, a sitar-player and a santoor-player walk into a bar.
The bartender asks - what do you guys do ?
The tabla-player answers - I create the most beautiful music in this world. And these guys ? One sings along, one strings along and the other one dings along ...

My karma ran over your dogma

Posts: 1
Reply with quote  #18 
Originally Posted by "pbercker"

Not so informative as one might hope ... I suspect you know about "Max" (have you used that for your tabla apps?).

Sorry, I hope this gives a better idea:

The techniques used for manipulating the tabla sounds are mostly: pitch/time shifting, frequency shifting, frequency modulation, resonance filtering, comb filtering, and a few others. As can be easily heard, there is not just one 'filter' (or any other effect) applied throughout, but the processing changes with virtually every stroke.

Why? Because a comb filter for example can ad a nice resonating sound to a loud 'kat' on the baya, but I wouldn't apply it to an open stroke, whereas applying frequency modulation to a sound without definite pitch doesn't make too much sense. A bank of resonance filters on the other hand can color strokes like 'tete' in varying ways, but can't be used effectively with an open stroke like 'ta'. And so on.

How does it work? There is an external function (object) in max/msp called 'bonk~' which, by a spectral analysis of the signal, detects attacks. My program (max patch) includes the score, i.e. all the rhythmical information and also which drum is played when. Starting from the beginning, the program follows me through the whole piece and executes all the details in sound processing, which are fixed for each time point. Of course this is not as simple as it may sound, because the analysis makes mistakes, it may detect an attack when there was just a small noise from a finger moving on the skin etc. Therefore the max patch calculates after each stroke when the next stroke is to be expected and 'opens' a window from shortly before until shortly after that time point. If an attack is detected during that timespan the program reacts accordingly, if not, then that particular stroke will not sound as composed (milliseconds count here because of the percussive sounds), and the program just goes on, calculating the next time point. There are additional mechanisms to avoid that the max patch is loosing track of me, and if I'm not making major mistakes it works quite well. But for security reasons there are around 20 spots in the whole piece where a person sitting behind the computer reassures the coordination of me and the patch.
Previous Topic | Next Topic

Quick Navigation:

Easily create a Forum Website with Website Toolbox.