[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference napalm::commusic_v1

Title:* * Computer Music, MIDI, and Related Topics * *
Notice:Conference has been write-locked. Use new version.
Moderator:DYPSS1::SCHAFER
Created:Thu Feb 20 1986
Last Modified:Mon Aug 29 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:2852
Total number of notes:33157

1760.0. "Piano Overtone Generator SGU" by SALSA::MOELLER (DEC's hip to the Standards Thing !) Mon Nov 07 1988 12:43

I've been thinking of a MIDI rackmount SGU that does ONLY grand 
piano overtone generation.

Though I have what I consider the premier (affordable) grand piano SGU, the
Kurzweil 1000PX, it still doesn't come alive the way an acoustic does.

We all know that the bugaboo of sample-based (even Roland's 'SA' series) 
piano SGU's is the lack of resonance when notes are struck when the pedal
(MIDI controller #64) is depressed.  Sure, the notes then die away naturally,
but there is none of the harmonic string vibration generated on an acoustic
piano.

Though the harmonics generated by pedal-down playing are very complex, 
they ARE mathematically analyzable, and thus modelable.  The SGU should 
be at least a 16-voice system, for realism.  It would need a very fast 
realtime CPU/FPU, but I believe it is doable, as the unit would NOT need 
to generate the  fundamental frequency, that would be taken care of in
another SGU.

Comments/suggestions from you hardware designers out there ??

karl
T.RTitleUserPersonal
Name
DateLines
1760.1Probably don't need more than a few filtres and oscillatorsMENTOR::REGa little risc averseMon Nov 07 1988 14:5010
    
    	I had been thinking along these lines last friday night.  
    I had the simple idea of a number of tones to represent the instrument's
    (almost any acoustic instrument)  "box"  at its resonant frequency(s).
    These would be triggered, reinforced and continued by sounding harmonics
    of their frequencies.  Presumably one could stimulate buzz and rattle
    emulators in the same sort of way :-^) 

    	R
    
1760.2SALSA::MOELLERDEC's hip to the Standards Thing !Mon Nov 07 1988 15:2722
    < Note 1760.1 by MENTOR::REG "a little risc averse" >
>    I had the simple idea of a number of tones to represent the instrument's
>    (almost any acoustic instrument)  "box"  at its resonant frequency(s).
                                        ^^^          ^^^^^^^^^^^^^^^^^^^^
    Two concepts here.. mine, generating the overtone series specifically
    for a grand piano, in stereo, generated by the same MIDI info, notes
    and pedalling, sent to the actual grand piano SGU.
    
    Reg, yours is more like a 'formant' generator.. it would actually 
    need to be a software-driven stereo EQ unit.. choose SOLOVIOLIN 
    setting, and it would would not need to generate actual overtone
    series, but intelligently impose characteristic violin-body formant 
    frequency EQ on the output of your OTHER SGU (sampler), actually 
    playing the violin sound.  
    
>    These would be triggered, reinforced and continued by sounding harmonics
>    of their frequencies.

    No, I don't believe the 'box' imposes harmonics that would need
    to be generated, but rather, a complex EQ curve.
    
    karl (90�F here yesterday)
1760.3STROKR::DEHAHNMon Nov 07 1988 16:428
    
    This sounds like a great ida. However, it sounds like you want to
    store a lot of harmonics. Unless you do some kind of fancy encoding,
    it's gonna mean a lot of memory. At ~$40/Mbit, that could add up
    pretty fast.
    
    CdH
    
1760.4U try this, k?DYO780::SCHAFERBrad - back in Ohio.Mon Nov 07 1988 16:5412
    Is this a bit more "possible" now that cheap pgmable reverbs have come
    out - especially the multi-FX ones?  The new Alesis thingie allows you
    to stack FX and control them via MIDI controllers in realtime. 

    Perhaps setup as follows: MFX using verb 1 (on) and verb 2 (off) -
    assign KX controller to send MIDI value - map sustain on to verb 2
    level max (which is a bigger room than verb 1), sustain off to verb 2
    level off.

    Maybe?

-b (we hadda incha snow last night 8-)
1760.5yeah well *I* still swim outside at night!SALSA::MOELLERDEC&#039;s hip to the Standards Thing !Mon Nov 07 1988 18:1515
    < Note 1760.4 by DYO780::SCHAFER "Brad - back in Ohio." >
>                              -< U try this, k? >-
>The new Alesis thingie allows you
>    to stack FX and control them via MIDI controllers in realtime. 

>    Perhaps setup as follows: MFX using verb 1 (on) and verb 2 (off) -
>    assign KX controller to send MIDI value - map sustain on to verb 2
>    level max (which is a bigger room than verb 1), sustain off to verb 2
>    level off.

    I DID try this using Performer on the Mac, found a semiautomated
    way to replace the #64 pedal events in a spare track with MIDIverb
    patch changes.  Using ONE MIDIverb.. when it switched from large
    room to small room (#64 off), there was a 'snap' as the 'big' algorithm
    quit.
1760.6So do I (have since I was 7 (oneupmanship rules!))GLORY::SCHAFERBrad - banished to Michigan.Tue Nov 08 1988 10:319
    I remember your problems with snap - have you tried this with a
    MIDIverb II?  Also, with a multi-FX unit, the snap will *probably* not
    be a factor, since you won't be switching patches per se, but rather
    munging a controller value somewhere. 

    Can you tell Performer to map #64 off to "scale controlller x from 100
    - 0 over delta time"?

-b
1760.7CANYON::MOELLERWed Nov 09 1988 12:145
    < Note 1760.6 by GLORY::SCHAFER "Brad - banished to Michigan." >
>    Can you tell Performer to map #64 off to "scale controlller x from 100
>    - 0 over delta time"?

   nope. 
1760.8Oh well.DYO780::SCHAFERBrad - back in Ohio.Wed Nov 09 1988 12:175
    Hmmm - then about all you can hope for is a FX box that does silent
    changes from mute -> big room -> mute.  I might have a go in a few
    minutes. 

-b
1760.9Where there's a will...DYO780::SCHAFERBrad - back in Ohio.Wed Nov 09 1988 12:379
    Just tried doing this with a MIDIverb II ... patch change is quiet as a
    mouse.  Of course, going from a very small room to a very large room
    results in a marked difference in ambience. 

    Couldn't figure out in a short time how to force the KX to map the
    pedal to both sustain on/off and use-patch-22/use-patch-11.  Surely it
    can be done ...

-b
1760.10hey, I have two feet, I'll buy another pedal!SALSA::MOELLERProton Spin Memory supportWed Nov 09 1988 16:3914
    < Note 1760.9 by DYO780::SCHAFER "Brad - back in Ohio." >
>    Just tried doing this with a MIDIverb II ... patch change is quiet as a
>    mouse.  Of course, going from a very small room to a very large room
>    results in a marked difference in ambience. 

>    Couldn't figure out in a short time how to force the KX to map the
>    pedal to both sustain on/off and use-patch-22/use-patch-11.
              ^^^^
              ||||    
    
	that was MY problem.. once you remap the controller to do this
    patch switch, assuming you can, then it's no longer doing #64 on/off!!!
    
    karl, from his limping DECwindows 4MB VS2000
1760.11My friends call me pontoon shoes.DYO780::SCHAFERBrad - back in Ohio.Wed Nov 09 1988 16:518
    This is really going down a rathole ...

    My OB-Xa won't respond to MIDI sustain, so I have its sustain pedal
    taped beside my KX sustain pedal.  Fortunately, I have wide feet... 

    I'll try this tonight and see what I can come up with.

-b