[Ardour-Dev] ardour3 and tranzport?

[Giso Grimm]

What I can (and will) do is to try to make the current ardour 2.x kernel
based code work also with a3. I would try not to change to much of the
current code, in a second step I would like to add functionality to
configure parameters of the control surface (e.g., time display format).

I am not a programmer, but I am programming a lot, and already had a
look at both ardour Tranzport code and also Mackie control code for
bugfixes.

I can only test with the Tranzport, I don't have an Alpharack and don't
plan to buy one in the near future. I will probably not have time to
work on this before mid of June (there are many gigs to play, Ardour 2.8
recording sessions, and 'real' work before). Until now I used the
patches by Mike from http://tracker.ardour.org/view.php?id=2521 to get
the kernel driver working, and I am using it for about 2 years now
without problems.

If anyone else will work on it before I can do it I don't mind ;-) If by
any chance OSC feedback will be possible until then, I would prefer to
write a Tranzport2OSC standalone driver (similar to
http://vegri.net/tranzport2midi.zip, which is a Tranzport to MIDI
converter with build-in bank switching, written mainly to control ardour
before I got Mike's kernel driver to work).

Thanks a lot, Mike, for your work on the kernel space driver!

Giso


On 04/01/2011 09:57 PM, Mike Taht wrote:
> I'd written the kernel driver for the alphatrack because I'd
> encountered the same problems with fast movement with it, as I had had
> on the tranzport.
> 
> So far as I know  the fast movement issues are totally unresolvable in
> userspace. Still. 4+ years after I wrote these drivers.
> 
> Then the breathtaking scope required to fully utilize the alphatrack
> became apparent...
> 
> Then it took 2 years for the tranzport and alphatrack drivers to
> become more or less standard parts of the kernel...
> 
> during which time I'd moved to the third world, and given up working
> on both the kernel and ardour and was deep into another project AND
> offline 95% of the time.
> 
> I am briefly back in the US and wouldn't mind reducing my guilt
> quotient for this project. Certainly core components of ardour3 have
> moved along enough to make implementing a better interface more
> feasible.
> 
> HOWEVER I'm not the guy to do it - obviously. I did the kernel
> drivers. That's what I know. C++ and I are not friends.
> 
> I am certainly willing to work (briefly) with another ardour
> programmer to try and get the knowledge gaps straightened out, but...
> 
> For those that haven't seen the the old patches to ardour to use the
> kernel drivers, they are here:
> 
> http://www.teklibre.com/~d/frontier.tgz
> 
> I do note that my last couple commits to the kernel tranzport driver
> had issues with the keyboard map. I tried to introduce functionality
> not supported by the transport hardware itself.
> 
> Never ever got the ardour/alphatrack interface to go anywhere.
> 
> the kernel drivers, however, are very stable.
> 
> On Fri, Apr 1, 2011 at 3:39 PM, Harry Van Haaren <harryhaaren at gmail.com> wrote:
>> Apologies for the double post: My reason is that I didn't mention I wrote
>> the script mentioned to use the alphatrack with
>> Ardour. The Alphatrack has some libUSB code available online that make it
>> show up as a standard ALSA MIDI device.
>> This script connected to that, and wrote to/from it using Ardour's OSC
>> interface.
>> Just thought I'd add that to make my mail a bit more coherent :-)
>>
>> On Fri, Apr 1, 2011 at 8:37 PM, Harry Van Haaren <harryhaaren at gmail.com>
>> wrote:
>>>
>>> Hey,
>>> I have no experience coding Ardour internals however I have been a little
>>> active trying to get the alphatrack working with Ardour using its OSC
>>> interface.
>>>
>>> http://harryhaaren.blogspot.com/2009/12/tut-6-ardour-control-using-osc-python.html
>>> Please note that I haven't been all that actively lately in keeping it up
>>> to date, but most basic functionality is there.
>>> AFAIK Ardour has since exposed more of its abilities to OSC, (which can
>>> easily be supported in the Python script) but just aren't atm.
>>> I'll do some testing if thats needed at some point, however due to time
>>> I'm not going to sign up to do a lot of coding..!
>>> Cheers, -Harry
> 
> 
>