fbpx
Skip to main content

MIDI Forum

Writing a new part ...
 
Notifications
Clear all

Writing a new part (wavetable)

3 Posts
2 Users
0 Reactions
3,943 Views
Christoph
Posts: 38
Trusted Member
Topic starter
 

I'm thinking of creating an extra voice in the repertoire of the pedal bass voices I'm using. Presently I'm using the GM sounds and some MT-32 extra sounds (which are not noticeably different from the GM sounds stored in the SAM2695 midi synth chip.

From my understanding the chip implements the Roland SC-55 behaviour. I never did such a synthetisation before, so I have no idea to start. All I can deliver is the
waveform. Velocity dependance not necessary.

How many samples would be required? I can send Midi commands to the device and store the values in firmware before sending them to the device during initialization.

Where should I start?
EDIT: I guess I was erroneously assuming that this would be possible but now I believe it ain't after searching through the SC-55 manual.

 
Posted : 14/04/2023 6:38 am
Geoff
Posts: 1045
Noble Member
 

I note your 'Edit' and I suspect that you're correct.

As far as I can tell, your device is is no way either an SC-55, or an MT-32, it's best described as a fixed copy of certain aspects of these two devices. In the case of the SC-55, you've got the sounds, and the midi implementation, in the case of the MT-32 you seem to have the basic sounds, and also one (the GM set) of the many add-on sound sets.

Both real devices had the ability (certainly for the MT, probably for the SC) to copy an existing sound from ROM to RAM, edit it, and then use the new sound in place of one of the existing sounds. This facility may have allowed you to do what you were thinking about. The MT sounds did include samples, but there was no process to add new samples. I suspect that the SC was similar. However, the sample is part only of the final sound, so there was still plenty of scope to create new sounds.

By the way, not sure what you mean by 'Velocity dependance not necessary'? Velocity data will be required. However, I supect that what you really mean is that the V data can be fixed into the system (your software) so it does not need to be provided note-by-note. For example, every note on can be assigned a V of, say, 40 hex. Your software could do that, if that would help your project?

Geoff

 
Posted : 14/04/2023 12:04 pm
Christoph
Posts: 38
Trusted Member
Topic starter
 

By "velocity dependance" (or is it "dependence"? 🙂 ) I meant, that I only need one waveform for different velocity cases.
My understanding of "velocity" is, that one has different waveforms for the sound, depending on wether on pushes the key hard or soft,
like e.g. on a piano.

 
Posted : 14/04/2023 11:29 pm
Share: