Forum

First of all, I'd like to thank Sam and the other developers that have been steadily filling up the EigenD/belcanto reference section of the wiki. I remain a little confused on some things:

Each of the many and varied agents within EigenD seem to have a wide variety of input and output ports, which are listed in detail so far. What I am unsure of is how to manipulate them to get certain things to happen, such as connecting things like the strip controllers and breath pipe to some of these ports. I have a vague memory of a belcanto demo that Dave gave at PMT in Birmingham (typed in at light speed from the interpreter keygroup!) which used the verb 'connect' to wire up the strip controller to one of the input ports of an AU (or a rig, or some other intermediate agent). I've experimented with some likely looking strings but the interpreter doesn't seem to like my bad grammar! Is there any chance of some further documentation on the general syntax and semantics of belcanto, and some further info on verbs?

You are quite correct - we haven't yet got to documenting the Plumber, which hooks up Ports for you. It's on the list to do, I'll ask Sam to think about moving it up in priority, although the increasingly immanent arrival of the Workbench does make it less important as connecting Ports just involves the graphical use of a wiring tool in that environent...

John

Does the workbench include any introspection tools at all? I'm thinking of the belcanto strings that pop up in the browser when you trigger certain talker keys. If a graphical action in the workbench could be viewed in terms of belcanto commands as well, it would serve as a great learning tool.
I envisage using the workbench a fair bit in graphical mode, but you always need the ability to write belcanto scripts using a good old text editor :)

Hi

The Pico strip pressure is unfortunately not something that can be controlled - it's a function of the way the ribbon is made. We actually buy this technology in (it's patented) so it's unlikely that this will change in the near future - we don't have much control over how it is designed. It is apparently possible to make the strip more sensitive, but only at the cost of a substantially reduced lifetime due to wear, which we really didn't want. We have investigated a number of other sensor technologies for the ribbon (such as capacitive sensors) but none of them offer the resolution of this one, and when doing large pitchbends it turns out that the current resolution is very nice.

You will find that the Alpha (and to an extent the Tau) strips are easier to press as they're wider and that makes this technology more sensitive.

As for writing Agents in Python, this will be possible when we open source, although most useful Agents are a mixture of Python and C++. Most data manipulation is done in C++ for speed as Python simply isn't quick enough to keep latencies low. The C++ components are assembled and configured in Python.

Hope that helps...

John

Thanks for the info!
After some hours of playing I got more used to the ribbon now (keeping the pressure at the required level and not sliding up to the very end of the ribbon so that the pitch bending doesn't stop).

My plan was to partition the ribbon into two trigger areas that transpose the respective key group up and down by a minor sixth. (The transposing only triggers if the finger leaves the trigger area without having changed the pitch for more than a half tone - so pitch bending would still be possible without triggering the transpose function). That way one could play the Pico with one hand (having the other hand free e.g. for piano accompaniment). And it would allow to split the Tau/Alpha into two keygroups (with a height of 8 keys) that can be played and transposed by one hand each. (Transposing would be done with the thumb for Tau/Alpha)
So one could write pieces that can be played by a Pico duet or one Tau/Alpha player.

As this doesn't have to work at audio rate perhaps Python would be sufficient, will see. Looking forward to the open source version of EigenD, this is really a great piece of software!

The splitting of the strip controller is an interesting idea, though it's not multi-touch capable, so if you wanted to do two shifts simultaneously (one for each keygroup) that wouldn't work. There are some interesting developments in the pipeline (that unfortunately I can't tell you about as yet ) that may make this problem go away though. Keep an eye out later this summer...

John

The 'sticky' behaviour would fix this, since if you run off the end, the strip position stays where you left it. I suppose it makes 'unpitchbending' back to exactly 0 very hard....

I don't have my Tau at hand atm., so I can't verify whether disabling the "Always return to origin" option of the new midi matrix has an effect on the relative strip (there is an absolute and a relative strip source parameter in the matrix).
Back in the days I used a custom Bidule solution that filtered out the final midi message 64 (when it would lead to a big value jump - value 64 could only be "approached" when coming from values close to it).
Don't know whether a comparable effect can be achieved via Belcanto.

Greetings,
NothanUmber