USB is probably your biggest issue here, provided the raspberry can handle the synthesis you want to do, you can help by assigning only a minimum of memory to the GPU. No need to do any other rearranging of the system, with raspbian it starts out headless (with X installed if you want to launch it) and capable of quite low latencies. It has a basic minimum of processes running and ready to add what you want from the extensive repository with apt-get. The settings to work headless are mostly the default. These are great advantages in terms of time required to get it running.On 24/02/14 05:09, Milan Lazecky wrote:
hi folks,
i have a yamaha wx5 (midi saxophone). i was thinking to use my midi2usb
cable to plug into raspberry pi which would use some fluidsynth soundfonts
to synthesize music realtime.
do you know about such project existing, or should i work on it from
beginning? (i.e. find some rt kernel, install sw, rearrange system for no
gui etc...)
do you think raspberry can handle it well?
USB implementation is poor and only some audio cards work, plus something else on USB like a midi input or networking will make it trickier. In your case, needing only stereo out, it is certainly easier.
Seriously consider an audio solution that can use HDMI audio out.
Simon
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user