Diva Meeting Minutes
2009 August 07
Minutes taken by Karl
Summary of last week:
We talked about the things that we want to have ready for when Marguerite returns from vacation (Aug 14?):
- bombproof combination of sensors - event based more reliable so far (Johnty)
- Fixed vowel locations in space according to F1/F2 - needs to be scalable (Allison)
- Better noise source for the Diva synth (Karl)
- tune standard deviations (Sid?)
- tune stops (Sid?)
- repair RBF to avoid zeros in the Diva synth (Allison)
Summary of today's discussion:
Karl:
- I have the bounded buffer working and unit tested in generic threads
- This should stop the drop-outs and make the audio much more reliable
- I now need to implement it with RtAudio and Holmes
- Once that's tested, I'll import it into the main system
Next:
- work on better noise source
- better vowel sounds, especially "OO"
Cam:
- recently completed source calibration. The interface provides a number of different ways to calibrate the orientation of the source including basic presets, simple angular changes, and by sampling three points when moving your hand in a circle.
Next:
- working on the inspector interface.
Johnty:
Troubleshooting hardware problems:
Serial to USB device:
- acquired different serial drivers and they appear to be working better than before
- they still stop but they don't crash the system
- in the process of finding a different Serial to USB device
Bluetooth Glove:
- narrowed it down to the micro-controller on the glove. The Bluetooth still works.
- need to push the button to re-boot the glove
- working on tracking down the particular signals that lead to the crash
Bluetooth Tracker:
- narrowed it down to the Bluetooth connection. The device still streams. After the Bluetooth reconnects the streaming continues.
Allison:
- going on vacation for 3 weeks and is trying to finish things up this Friday before she goes.
- creating the new interface for the vowel space:
- vowels physically located in space according to F1/F2 values (i.e. not individually placed according to taste)
- can move and scale the grid through the Max interface
- doing this leads to all sorts of issues about the orientation of the source. Manually locating vowels with the glove automatically took care of rotation around the vertical axis. Now we have to find a way to manually rotate the horizontal space.
- short-term solution is to locate the tracker in the center of the body
- given the short time-frame, Allison's plan is to get something up and working, but it will be a prototype.
We got into an extensive discussion about a number of different ways that we could locate the source in space.
Given that we are in the process of developing an interface simultaneously in both Max and C++, we should come up with a design document (even if it's a simple one) on the way that we are going to look at the vowel space.
