[linux-audio-user] Is there a shortcut to merely test a midi device?

Russell Hanaghan hanaghan at starband.net
Thu May 6 12:16:26 EDT 2004


On Thu, 2004-05-06 at 08:09, Lee Dunbar wrote:
> Howdy all....
> 
> I have been trying to learn a few things at once, possibly too many, and I'd like to cut some unknowns out of the big picture. I'm hoping that someone can help me with a down and dirty test of my midi hardware.
> 
> Details:
> I am trying to learn about setting up and using midi in SuSe 9.0, and also am kinda new to Linux. Also, the keyboard might be defective... the 3 octave keyboard is 'new' in regards that I just bought it last year (was still shrink wrapped), but it is leftover stock - the manufacturer is out of business. To give a clue about the age, the keyboard came with Windows 3.1 software (no mention of 95). I only had Windows ME on a celeron system; when I installed the included software, it could not find the keyboard, but I suspected the age of the software and the differences in OS revisions were more the problem. And I assume that the celeron system was properly configured.... but nothing ever noticed a keyboard.
> 
> I then took the keyboard to my Linux box... it still seems to be uncommunicative, but I'm new to Linux.
> 
> There are just too many variables in the pot right now:
> a] Does the keyboard even work? It powers up and appears to change functions per written instructions. But no communications.
> b] Does my Linux box (SuSe 9.0, AMD Athlon, DFI AK75EC mobo) have a hardware or BIOS defect? BIOS says midi/game is set to midi (only other option for that port is game).... the cable is definitely connected to the correct port.
> c] Does my Linux box have all the proper modules? I see MPU401, virmidi, and all the soundcard playback stuff works fine. (Via 686 based chipset, AC'97 compatible). I can issue aconnect -lio and see a difference when the module is inserted, but do I have all the proper steps completed, etc etc etc.
> 
> Ignore answering all the above, please; I'd rather try to send strings to the midi port and see if the keyboard is receptive, see how it responds, etc, so I can have some of the variables removed from the wide matrix that I presently face.
> 
> Any suggestions on brute force communications w/midi devices? Can I cat some string int /dev/midiwhatever? I once used 'cat ATO > /dev/ttys00' for testing a modem, that was slick, but what commands are sent in MIDI? That kind of test would be, IMO, very useful.
> 
> TIA!
Dont know that your going to get off that easy! :)

If it were me I would do the following...

Use Qjackctl (Assuming your running "X" and guis's) it has a myriad of
useful interfaces for midi and the alsa pcm ports. Before even running
Jack, see if your midi port shows under the midi tab in the "connect"
window. If so, then I would take that as your system seeing your midi
port and subsequent applications. Also cat /proc/asound/devices. I use
alsamixergui for that too...click on "Alsa" logo in top right corner.

If those are there, try running vkeybd (small software midi virtual
keyboard) make sure your sending midi data on the right channel. I think
the Kalsatools package has a little gui app that show what midi data is
being received. Aseqview is another handy little gui app for seeing what
midi data is being sent.

The other thought that comes to mind...does you card or sound chip have
in built wavetable? if not, you might be sending midi stream to it but
it has no sounds to play. I use fluidsynth and the new Qsynth for this
and load soundfonts. Google around some.


Dont know if any of this is useful to you but these things have saved my
butt with the elusive midi connections.

R~

> 
> -- Matthew 28:19,20
> 
> 
> _____________________________________________________________
> Are you a Techie? Get Your Free Tech Email Address Now! Visit http://www.TechEmail.com




More information about the Linux-audio-user mailing list