RME9652 driver updated

[Karl MacMillan]

Winfried Ritsch wrote:
> 
> [...]
> > using alsa drivers.  Any volunteers to do testing to the code when I am
>  > finished?
>  >
> 
> for sure, I will,
> 
> I saw you used stream mode too, maybe there is an answer.
> 
> The point is normally different devices are not synced by default but
> on hammerfall devices they are, so this could be a trap for
> unexperient user if the alsa-part just uses every alsa-device it
> finds. we need some kind of selection method...
> 
> mfg  winfried

There is definitely a need for some way to select which pcm devices the
user would like to use but I don't have a good solution at the moment -
I don't want to the require the user to have detailed knowledge about
the pcm devices on their card and set lots of command line switches just
to get useful default behavior.  Any ideas?

As for the sync, I plan to move to mmap block mode for the default which
will allow access to the pointer position in the ring buffers on the
card so that sync can be maintained.  As long as only one card is used
there shouldn't be a problem.  Am I thinking about this correctly? 
Another reason to move to the mmap mode and avoid the plugin layer is it
is possible that the ALSA library will insert sample rate conversion,
interleaving, etc. when using the plugin api. This is great for most
applications but probably bad in the case of PD.  Stream mode using the
direct layer will be maintained as an option.

Karl

-- 
_____________________________________________________
| Karl W. MacMillan                                 |
| Peabody Institute of the Johns Hopkins University |
| Network and Telecommunications Services           |
| karlmac at peabody.jhu.edu                           |
| 410/659-8297                                      |
-----------------------------------------------------