[Fwd: Re: [PD] teabox...any users drivers]

[Stefan Tiedje]

hi stefan,

Am Donnerstag 25 August 2005 10:44 schrieben Sie:

 >> Christian Klippel schrieb:
 >
 >>> > on a general note: whats the sense behind using an audiointerface
 >>> > for controllers, except to block a valuable audio-port? do you
 >>> > really fiddle with knobs and sensors that fast? if so, can i have
 >>> > your hands, please?would make my soldering life much easier ;-D
 >>> > mind you: to use that, you would need an extra audio-interface.
 >>> > add that to the cost of the unit.
 >
 >>
 >> The teabox is sending all of its values thru spdif. It does make a
 >> lot of sense, if it happens that you do have an audiointerface with
 >> spdif already, not using it for audio on stage because all audio is
 >> analog there. Its delivering that as a stream, locked to the audio
 >> rate. Which makes it easy to lock your controls to audio...


hmmm .... on most "normal" cards you can select only one input to 
record. so, if i now want audio input, what then? i need another s-pdif 
interface for that box .... which means another card .... and then comes 
the problem of different cards never having the exact same samplerate, 
which makes it almost impossible to use them inside one dsp chain.

of course, with multi-channel spdif-input cards thats no problem, but 
this is not the usual case.....


 >> With having experienced all kinds of problems with interacting
 >> several USB-Midi, Audio-Midi interface combinations, I think its
 >> quite an appealing idea having only one firewire port delivering all
 >> the data within the same protocol.
 >>


i agree that the usual midi interfaces or way to weak for having big 
amounts of control input. either you have some controllers with a poor 
resolution (7 bits), or less controllers with higher resolution (14 
bits). anyway, midi is somewhat slow for that, so yes, midi isnt the 
best choice for that.

you mean a analogue/spdif audio box which then connects via firewire to 
the host?

oh, and what about system load? doesnt handling that kind of 
control-input via an audio stream have quite a lot of overhead? i mean, 
does it make sense to constantly send a stream of data, even if there 
are no changes on the controllers?

not that i want to make it bad, but i think that there are certain 
constraints for using that device.

i prefer using the usb bus, which is quite fast and available on almost 
all computers and pda's, etc. nowdays. since full-speed usb (which is 
what im going to use in my project soon, its currently low-speed) can be 
also used for audio, i dont expect any latency problems. add to that the 
fact that data is only sent on changes, it should be really fast and 
low-latency.


 >> Of course its pricey...
 >>


sure it is.....


 >> I guess Timothy would be happy to help out creating Pd objects if he
 >> didn't do it already...
 >>
 >> Stefan


greetings,

chris