Fixed pitch bend range woes

[Vaultnaemsae]

One thing that kinda irks me about the VG-99's G2M function is the fixed pitch bend range (+/-12 semitones) as there are some synths with fixed range of +/-2 semitones e.g. some of the synths from the Arturia collection.

I had an inkling that perhaps it is something that could be adjusted at a deeper level so had a flick through the VG-99's MIDI implementation but didn't find anything particularly relevant except what you can see in the attached image.

I don't know how it works but looking at this entry I'm wondering if is it possible to limit the VG-99's PB output and if that would result in the possibility to match with unmatched synths.

And further to that, does anybody know of a way to convert or scale these differences to make unmatched devices match?

[pasha811]

That is happening to me as well. Thanks for posting. Now I know I am not dumb is as designed at lower level.
I think the only way is to reduce the bending information range in the VST or have a sort of MIDI filter in the DAW. I will try with what I have (Live 11)
For some reason, I have better results when using VG99 and JV1010 via direct MIDI connection, I can play the trumpet patch without the 'drunk player' feeling... but lately, since 2021 I am on Midi Guitar 2 and that offers better control for solos. So I use VG99 for pads / layers and MG2 for solos at the moment.

[Vaultnaemsae]

That is happening to me as well. Thanks for posting. Now I know I am not dumb is as designed at lower level.
I think the only way is to reduce the bending information range in the VST or have a sort of MIDI filter in the DAW. I will try with what I have (Live 11)
For some reason, I have better results when using VG99 and JV1010 via direct MIDI connection, I can play the trumpet patch without the 'drunk player' feeling... but lately, since 2021 I am on Midi Guitar 2 and that offers better control for solos. So I use VG99 for pads / layers and MG2 for solos at the moment.

Yeah, matching pitch bend range is a make or break setting. The thing is, all
my research has led me to "bend range is a function of the receiving device" so it must be adjusted there. This makes sense but for guitar synthesis as clearly the transformation on the guitar end needs to establish independent bend range info. I suppose you can go chromatic mode and kill OB altogether but where's the fun in that?

So, VSTs like MS-20 V, ARP Odyssey, SEM V, Matrix-12 V become very difficult to use. There is a workaround for two of those mentioned but it's a pain to set it up for every patch. You have to use modulation assigns and pin them to the oscillator frequency. In both SEM V and Matrix-12 V it's not a simple setting to make but it does work. The other have no such possibility AFAIK.

The JV1010 synth has adjustable bend range, right?

I couldn't figure out a way to filter PB in Max 8 or Live 11 in any useful way but then, I'm not really sure I understand what needs to be done anyway. Using the VG-99 at +/-12 and a VST at +/-2 results in weird performance from the VST and I image it's the same for hardware synths.

[Brak(E)man]

I'm curious to what happens?
Playing fretless I need at least 12 semitones bendrange preferably 24
But when I'm bending on a fretted I'm seldom bending more than 2 semi. 3-4 at the most.
And setting the bendrange to 12 sending and 2 receiving only makes a difference if I'm bending more than 2.
If you're only bending 2 semi that's what the 99 should be sending out, right or no?? Even set to 12 .... Sliding more then 12 on fretless sends the vg 99 out of range in other words no slide/ bending more then 12 semi can be heard.

(If not can't any plugin divide the bend with 6 if you have 12 sending and 2 receiving?)
Are you using mono mode ?
If not use a fat finger or some wax potting 🙈

( I usually set the bendrange to different settings to get some nice weird effects when playing both guitar + softsynths )

[shawnb]

You can usually set the pitch bend range, PBR, on both the sending device and the receiving device. 

Some gear has a fixed range.  Not a problem... As long as you can adjust the range on the other device to match.

But they must match for accurate performance.

What happens when they are out of sync is that all wheel events - used for bends, legato, slide, whammy, etc. - convey the wrong frequency/note. 

Legato can literally produce the wrong note.

Another common symptom of misaligned PBR is sour chords, since minute pitch differences can be greatly exaggerated.

[shawnb]

( I usually set the bendrange to different settings to get some nice weird effects when playing both guitar + softsynths )

Of course, if you WANT the exaggerated wheel event effect, you can do that, too! 

Wheel events basically go from -8192 to +8191.  PBR settings define what that means.  Since the VG-99 range is 12 semitones, then a 1-octave performance (e.g., an open string followed by a hammeron 12 frets up) will spit out a wheel event with an 8191. 

If your receiving synth thinks 8191 means two semitones, then that 12-fret hammeron will be interpreted as going up 2 semitones by the receiving synth, because a PBR of 2 means the max wheel event (8191) means 2 frets.

Some keyboard-oriented synths allow you to specify two values - a different PBR for up & down.  E.g., a 2-12 means pulling down on the pitch wheel on the keyboard may go down an octave, whereas pushing up on the keyboard's pitch wheel only goes up two semitones.  In that case, a VG-99 12-fret hammeron is translated as 2 semitones up, but a 12-fret pull-off may be reflected accurately as a 12-semitone drop...

[shawnb]

If both sending and receiving devices are fixed, and have different PBRs, that's a problem.

If you MUST use a device with a fixed PBR of 2, with your VG-99 with a fixed PBR of 12, then you will need some sort of external device to do some math for you.  E.g., a MIDI Solutions event processor or a Bome device, to scale the wheel events appropriately. 

Or go chromatic.

[szilard]

midi solutions makes a device called an event processor. I have an event processor + that I use to use with the vg-99 to interface with a synth that didn't have a programmable BPR.

[shawnb]

Fun with PBRs!!!

While we're on the topic... 

First thing: Note that wheel events only modify existing notes...  They do not initiate a note in and of themselves...  So, when playing 100% legato, on the very first hammeron, the transmitting device may sense there is no note & send a Note event, not a Wheel event.  Then communicate via wheel events thereafter.  Also, a clearly plucked or hard hammeron will also retrigger by sending a Note (not a wheel). 

Also...  When you exceed the max PBR with legato/whammy/slide, it may try & match your playing and send a new note as well...

So... 

PBR Detect:  If you don't recall your current PBR setting, e.g., on your FTP, then choose a non-chromatic voice with a hard clear attack on your synth, then run a slide up a string.  If the note rings every 2 frets, then your PBR is 2.  If the note rings every 12 frets, then your PBR is 12.

Theremin:  Transmit PBR of 2 (e.g., using your FTP), synth PBR of 12.  Small pitch bends can send the full range (8191) of wheel events.  If your synth thinks 8191 is an octave, well, it's very easy to get a wild theremin effect.

Microtones:  Transmit PBR of 12, synth PBR of 2.  Hammerons/pulloffs on your first 12 frets are going to send values 0-8191.  But, if the synth thinks 8191 is only 2 frets, then you are playing fractions of those 2 semitones.

Very tricky...  Be careful not to play too hard or you'll trigger a new note...  Lots of tricks we use to get rid of glitches (play feel, dynamics, low vel cut) make this harder to do...

[pasha811]

Yeah, matching pitch bend range is a make or break setting. The thing is, all
my research has led me to "bend range is a function of the receiving device" so it must be adjusted there. This makes sense but for guitar synthesis as clearly the transformation on the guitar end needs to establish independent bend range info. I suppose you can go chromatic mode and kill OB altogether but where's the fun in that?

So, VSTs like MS-20 V, ARP Odyssey, SEM V, Matrix-12 V become very difficult to use. There is a workaround for two of those mentioned but it's a pain to set it up for every patch. You have to use modulation assigns and pin them to the oscillator frequency. In both SEM V and Matrix-12 V it's not a simple setting to make but it does work. The other have no such possibility AFAIK.

The JV1010 synth has adjustable bend range, right?

I couldn't figure out a way to filter PB in Max 8 or Live 11 in any useful way but then, I'm not really sure I understand what needs to be done anyway. Using the VG-99 at +/-12 and a VST at +/-2 results in weird performance from the VST and I image it's the same for hardware synths.

JV1010 has Adj PB range... but the editor software doesn't work anymore...too old...so I cannot edit it and I use it as it is. It turns out that I have too much of legacy stuff. Maybe it's because I am myself legacy ;-)
That's why I invested in MG2 after a long trial. BTW

[admin]

JV1010 has Adj PB range... but the editor software doesn't work anymore...too old...so I cannot edit it and I use it as it is. It turns out that I have too much of legacy stuff. Maybe it's because I am myself legacy ;-)
That's why I invested in MG2 after a long trial. BTW

Seek a Third party JV-1010 Editor for PC/Mac
https://coffeeshopped.com/patch-base/editor/roland/jv-1010

[pasha811]

Seek a Third party JV-1010 Editor for PC/Mac
https://coffeeshopped.com/patch-base/editor/roland/jv-1010

Thanks admin! Just downloaded and testing it.

[Vaultnaemsae]

I'm curious to what happens?

For me the problem is that the G2M is not solid when the PBRs don't match. I don't actually bend very much at all in my playing. But all sorts of mess and misfires creep in when the PBRs don't match in my experience.

I think the only way is to reduce the bending information range in the VST or have a sort of MIDI filter in the DAW. I will try with what I have (Live 11).

I found this device for use in Max4Live. It does exactly what you were talking about. I set Omnisphere to +/-12 and dropped this in front of it. At about 18% it would modify the bend to a +/-2 PBR. Very useful and solves the problem; at last in Ableton Live and Max 8 standalone, I guess:
https://maxforlive.com/library/device/6534/midi-pitch-bend-scaler

If you MUST use a device with a fixed PBR of 2, with your VG-99 with a fixed PBR of 12, then you will need some sort of external device to do some math for you.  E.g., a MIDI Solutions event processor or a Bome device, to scale the wheel events appropriately. 

I'll be testing the Bome software to see how it handles this. Good tip, thanks!

[Vaultnaemsae]

Still not sure what the sysex command in the OP is for though!!!

[shawnb]

That's not a sysex.  That's the layout of the wheel event.

In the old days, manuals had all kinds of fun stuff in them.  Now you're expected to hunt it all down yourself....

[Vaultnaemsae]

Actually, seems I had a couple of things wrong.

1. The Max4Live Midi PBR scaler device does work as I described when put in front of a VST and if you send in PBR data from a controller. I was using the keyboard and wheels in Animoog Z for iOS. It only allowed 2 semitones of pitch bend to be heard from the VST. I assume it's just chopping the upper and lower range of the PB data. But it seems to have no effect on the VG's PB data output. I'm probably doing something wrong -- seems like it should work be fine.

2. Arturia Korg MS-20 V does actually have a per preset PBR setting, now. I'm pretty sure it didn't upon release.
*SEM V and Matrix-12 V still don't have intuitive PBR settings.

I'm now thinking maybe chromatic mode with an external PB wheel-like device might be the way to approach this.

[shawnb]

No need for an external device.  Just go straight from the vg-99 to the Artura, and set the Artura's PBR to 12.

Keep it simple!

When in doubt, find a way to monitor, e.g., using MIDI-OX, to ensure that what is on the midi bus is what you think it is.

[Vaultnaemsae]

No need for an external device.  Just go straight from the vg-99 to the Artura, and set the Artura's PBR to 12.

Keep it simple!

When in doubt, find a way to monitor, e.g., using MIDI-OX, to ensure that what is on the midi bus is what you think it is.

Well, that's the point. Matrix-12 V has no PBR adjustment. You have to mess around with the modulation matrix on a patch by patch basis. Same with Korg Odyssey except it has no option to even mess with PBR via a mod matrix.

There are others too — especially for iOS synths. They have great sounds but unmatchable PBR.

[shawnb]

Sorry, I thought you said the Artura had adjustable PBR...

As noted above, you have 2 main options for the synths that don't:
 - external device to do the math
 - chromatic

Or use an ftp, which has much better g2m. 

[Vaultnaemsae]

Sorry, I thought you said the Artura had adjustable PBR...

One of the synths I originally mentioned has PBR now, don't think V 1.0 had adjustable PBR.

I should probably just forget it and stick to synths that are fit for the purpose.

As noted above, you have 2 main options for the synths that don't:
 - external device to do the math
 - chromatic

Yes, I'm trying to find the right device to do the math. I don't want to use a hardware converter. I experimented with Bome last night but couldn't figure out how to transform the signal properly. The pitch would move but not return to zero. Will explore further. The VG-99's chromatic mode + external bender is an option but I'm yet to test it.

Or use an ftp, which has much better g2m. 

Been there, done that, still doing it a bit with the FTP -- I have two. One is on a guitar with a fixed bridge and an RMC system that suits it quite nicely as there is plenty of space for the FTP and controller and it doesn't interfere with the piezo system. The other is in a closet somewhere. They work pretty well but are bit finicky in some respects.

My main guitar cannot be retrofitted with the FTP without some invasive mods. It always runs through the VG-99 so I'm interested in "bending" signals to my will. Also, I've had some issues with the FTP when used live in the past and I couldn't say it behaves in a predictable manner.

Thanks for your valuable input. I first became aware of PBR years ago after reading some of your posts here

[shawnb]

Yes, I'm trying to find the right device to do the math. I don't want to use a hardware converter. I experimented with Bome last night but couldn't figure out how to transform the signal properly. The pitch would move but not return to zero. Will explore further. The VG-99's chromatic mode + external bender is an option but I'm yet to test it.

A thought...  Internally, the wheel value is actually a number from 0-16383.  It's possible that's how the Bome sees the wheel event.  (I don't have a Bome...)

To get the +/-, you have to subtract 8192 from that...  I.e., 8192 = 0, or no bend at all.

Note also that to go from PBR of 12, to a PBR of 2, I am pretty sure you need to multiply the input x6.  That might not be intuitive, but you need the values bigger, & you need to drop things outside the range.

So if the Bome is seeing wheel events as 0-16383, what I would do is:
- Subtract 8192
- Multiply by 6
- Drop all events < -8192 or > 8191

If it's seeing -8192 thru 8191, you don't need to do the subtraction step.

For tasks like this, MIDI-OX is your friend...  If you don't have MIDI-OX, you can record the values into your DAW & inspect the wheel events there.

Hope this helps,