On 8/28/06, Winfried Ritsch <ritsch@iem.at> wrote:

Hello,

> > I just came across this object in the cvs that gives access to the hdsp
> > 'madi' mixer in pd, which is great. Will it work with my hdsp interface
> > as well? I can't get it to 'find' and useable soundcards, whereas pd is
> > using it fine.
>
> as far as i know, [hdspmmixer] is really _only_ for the madi. (wini is
> not here for the next 2 weeks, so he cannot be asked).
>
> the main reason for this object was, that there was no mixer (gui) at
> all for the madi, whereas there IS a hdsp-mixer for the multiface and
> the like.

status report:

I started it for our 3D-Mixer to use outputs as Group outputs and with
the "number box Matrix"-Gui so you have a matrixmixer. It should work on the
hdsp cards also, but i havent tested it hard (its cvs now). I hope I will
finish this in next 2 weeks.

The main problem was that on the madi we have (64+64)*64=8192 Fader and
therfore numberboxes as control, which makes pd slow. Even with 4bytes this
exceeds 16k-limit on ioctl for madi to get it in one call, so it has to make
a lot of sytemcalls if also monitoring them. So the goal was to extend the
alsadrivers for memory-mapping of faders and additional power and peakmeters
for in and outs. This would reduce a lot of systemcalls for low latency but I
dont know if it will ever work... any hints.

So its a good idea to make it compatible for als hdsp cards, but I have to
change the arguments so the soundcard can be set, even if there are more than
one in the computer.

mfg winfried

_______________________________________________
PD-list@iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list