siebenhirter

I would like to create effects known by different keyboards as "trill" or "repeat" in order either: a) from a MIDI command to a series of MIDI commands with falling volume or b) from a SoundFont (one-shot-attack) to produce soundFont with repeat / trill to sound similar to banjo/mandoline-accompainment.
any idea how to do this with kxProject?

__________________
kX 5.10.00.3537 / Audigy SB0090 [c400] / ProFx v 3.08 5.10.00.3534

Maddogg6

The most obvious answer if to use a midi controller keyboard and record the midi data from a 'trill' performance on that keyboard.
This can also be done by; Quickly alternately striking adjacent keys); holding a midi note and moving the 'pitch wheel', and recording the midi events generated from those actions in a midi track in the sequencer.

If you don't own a midi keyboard controller...
Use a 'Edit Controller' view in a MIDI sequencing program such as Cubase or Sonar. This 'view' is integrated into Sonars 'Piano Roll' Editor view.

Theres 2 ways to do a 'TRILL'..
1) 'Pitch wheel Trill'
Add 'Pitch Wheel' to a held MIDI note event to change pitch on a 'held' note.
The amount of pitch change is determined by RPN 0. (Google 'MIDI RPN' for more info on that - look at the SF2 spec found in the kX document / help area.)

'Pitch Wheel' events can be positive & negative in terms of pitch (lower and raise) pitch by a % of X (in semitones - up to 24, 2 by default)
Full pitch wheel Positive = Pitch wheel event: 8191
Full pitch wheel Negative = pitch wheel event: -8190
* The pitch wheel will spring return to '0' in real life, but a sequencer needs the event when you want that change to occur.

2) 'Note Trill' - a Mandolin and banjo, is more of a 'fast' pattern picking... lets look at *that* closer, whats the pattern in notes (midi note and time)? - whats the pattern in how strong/weak each note was picked (velocity)?
These are made just with midi notes, usually in a lower velocity range. Just make sure also, note ranges are 'realistic' - a banjo with a low C string? not.

siebenhirter

@maddogg6,

thank you for your ideas - now i am trying some different methodes, but until now i am not satiesfied - seems to take some more time and trials. tolerable results i had with delay-/repeat-effects of VST-plugins via ASIO.

another question in another thread is how to prepare soundfonts, to get full levels on default-FXBus14 for chorus-application.

__________________
kX 5.10.00.3537 / Audigy SB0090 [c400] / ProFx v 3.08 5.10.00.3534

Maddogg6

No problem - I have spent some time trying to 'emulate' other instruments more realistically - and while some synths will automate some of these things via appregiators (smart ones) or otherwise - its the mechaics behind how the original instrument works that is needing understanding, then apply that to what ever synth you have access to.
Bear in mind also - there *are* other factors - a simple one to grasp is a snare drum - a real snare drum - the harder you hit it, the 'brighter' the sound - so, to improve a snare sound in a synth is to apply 'velocity scaled filter'

To make sure your soundfont can use Chorus (CC#93) - in CL Vienna - make sure the intrument/patch have that send adjusted - some are set to 0 in the soundfont - and thus, when adjusting CC93 - so little comes out of fxbuss for that send - you dont really notice.

Keep in mind there are several gain stages associated with those sends (reverb 91 & Chorus 93)

1) Sound font Instrument - as in a single layer/split
2) sound font Patch or a global - as in applied to all splits and layers
3) MIDI CC 91/93 levels on a midi channel
4) kX Router Sends C&D levels
5) kX DSP - any gain or attenuation in the DSP ??
Not to mention - these sends can be set as 'relative' or absolute - I use relative which maybe all kX synths can handle, 'absolute' may only be supported by higher version of SF - but I havent played with that to know for certain - relative has worked just fine for me, that is, once I understood the above list... keep in mind - some of my sound fonts are from way back when I had an AWE32 - which was only SF1.0 compatable IIRC. And Vienna is buggy, so I havent done a whole lot of editing since (just patch managing, tweaking here and there - no really patch creating like I used to do).

All of the above make 'repeatability' a pain - but gives *a lot* of control - too much maybe for some..??

edit:
All of the above - is beyond sample quality - crappy samples will never sound real.

more edit:
My soundfonts typically will use a very low send ammount in the sound font. Like 2-5% which seems to get amplified in kX Router send C&D as I have them set to max levels in my kX router -

But the best thing I ever did was stop using FXMix and start using mixers in kX DSP with LEVEL meters - this allows me to see what is sent and compare to the 'DRY' levels (kX Router Sends A&B)