I'm using the ZTAR with "The Trumpet" and get great results. The Ztar tracks perfectly, I have the AXON MK11 and it's no good. You play one note and get 10 ghost notes on the screen..
I use for recording only for orchestral or combo. The only thing I don't use it for is guitar. Anyway, it has plently of touch controllers which you can program in keyswitching, vibrato. C11 etc... I have one patch set up just for the Trumpet. If you haven't already, check it out. The company is Starr Labs.
Z7-S Mark 2 with TCA-1 trigger cap assempley offers 6 programabe pads (over strings). One thing that I've found is for me it is more natural with 2 hand tap method. It's polyphonic so you can play mutiple notes on one string.
I was about to buy "The Trumpet" and "Mr. Sax T", for use with a MIDI guitar and breath controller. However, I discovered what looks like a major shortcomming - the pitchbend range. For effective midi guitar operation, the range should be set to at least 12 semi tones (to ensure sliding, hammer-ons and pull-offs come accross properly). There are some other issues as well. I will start a seperate thread about that in a couple of days.
My main guitar-to-midi interface is an Axon AX-100 and it works very well. The problem you described, Tony, indcates incorrect setup (either parameter settings, hex pickup installation, or both). Playing style is a factor, but unless you technique is very sloppy, the Axon is quite forgiving. Alan, the performance of the GR-33 (I have one) is no match for the Axon, see below.
I've been using the Axon with Yamaha VL technology (physical modeling for sax and trumpet emulation, etc.) and a lot of other gear (incl. software) successfully for years.
I had a ZTAR in the past and I still have a "poor man's ZTAR", the Yamaha EZ-EG (there's also a EZ-AG). It's very similar to the ZTAR, but not as sophisticated (and not as well built), although when running it through a computer and using some scripting, you can make it do almost anything. I use it very little, because the Axon works very well.
The Roland GI-20 guitar midi interface, although not as fast as Axon, works quite well if you use a RMC piezo hex pickup as found on some Godin and Brian Moore guitars (and in custom installations on many other guitars). Tracking accuracy is noticeably better than with magnetic hex pickups (like the GK-3, etc.). On the Axons there's hardly any difference between the different pickups.
With Roland GR20/30/33 synths, the mag/piezo differences above apply, but the MIDI output is also slower than the GI-20. Response of the built-in sounds is not bad.
I had read several posts on the Axon forum and it sounds like it is a viable option for midi guitar players that do not want to completely change technique (e.g. two-hand tapping). However, the ZTar offers several intriguing options including playing multiple notes simultaneously on a string...I know the Axon will support string bends, something I'm still not sure would work well on the ZTar. I know the GR-33 does not cut it, even with my Brian Moore Custom C60P (w/RMC Piezos). But, after reviewing the Axon videos and forums, I tend to agree there is some setup and technique required to fully utilize the Axon.
I'm not sure I can afford the entry level price of the ZTar ($2000 US) vs. the Axon ($700 US). It would be great if Starr Labs would lower the price bar for their technology.
Still thanks for all the info on the Guitar Controllers.
Good luck with that stuff. I have a Strat with a Roland Pickup and a Godin which comes with the pickup built into it. Maybe your'e right but I don't agree. As far as my teqnique, I've studied with the infamous Dennis Sandole and have done hundreds of "Big Name" sessions in NY so I don't think my technique is suffering too much. Anyway, I said what I had to say and this really is a disscusion for another FORUM.
I didn't mean to question your technique. I was trying to make Alan aware that the Axon is more "forgiving" then other guitar-to-MIDI converters, when playing technique is not precise.
Among some general midi-guitar info for Alan was my main issue: the KEY REQUIREMENT for effective control by a MIDI GUITAR of virtual instruments is a PITCH-BEND RANGE OF AT LEAST 12 SEMITONES (so that hammer-ons, pull-offs, slides, string-bends, etc, are translated properly). Looking at the documentation, this appears to be impossible in Mr sax T. and the Trumpet.
This a very important issue for midi guitarists (and some are not aware how important it is). I was about to start a thread about it, when Alan's post appeared. I'll still make it a separate thread (with some added issues, e.g. portamento, controller assignments, etc.), to get the developers' attention.
The apparrent pitch-bend limitation has stopped me from buying the sax and trumpet.
Although useful, the ZTAR (have owned one) and Yamaha EZ-AG/EG (still own one) are not real guitars, but "switchboards" that lack the (subtle) expression possible with coventional guitar technique. Those techniques are very usefull when playing virtual sax. I used it successfully with Yamaha VL physical modelling.
I agree. I think this is a very useful thread, especially since other guitar players (like us) may want to use the Trumpet (I actually started off playing trumpet and think the sounds I've heard so far are real close) and Sax plugins live (I mainly play cover music).
For me however, the Ztar would not be a good option in this situation since one requirement I have is to switch realtime between the guitar and plugins. It's also very informative to know that the Trumpet and Sax to not support the full-bend range. Again, I would like to thank both of you...for me (and I hope others), this has been a very informative topic and maybe the folks at Sample Modeling will be able to address the bend-range issue. Again thanks to both of you!
No response yet from Giorgio or Stefano. I'd still like to know if, what looks like a pitch bend range limitation in both the trumpet and sax, in fact exists. If it does, can an extended pitch bend range be introduced in the near future (at least 12 semitone range required for MIDI guitar, as explained earlier).
If you can promise that, I'll buy both the Sax and Trumpet shortly!
While I'm here, a couple more MIDI guitar tips for budding MIDI guitarists:
The main advantage Axon has over other converters (early transient detection) only applies when selecting 'plectrum/pick' as playing style. When you select 'finger', only normal pitch to midi conversion is used, which is slower.
For all converters, putting felt or something similar under the strings between the nut and 1st fret helps stop glitches when lifting fingers from the strings, etc. Should only be done when you don't intend to use open strings, of course.
Sorry Udo, but the Pitchbend range is, and will remain, limited to 2 semitones. Please note that portamento is carried out by overlapping notes. There is actually no reason for increasing the range of the pitchbend. The only thing one would get is a chipmunk effect, definitely unrealistic.
To clarify, although the pitchbend range needs to be large for Guitar pitch-to-MIDI conversion, the pitchbend amount applied to the instruments will only be a fraction of the full range.
The large range is required to ensure string bends, etc, are translated accurately into proportionate pitchbend data. The range on the guitar-midi converter and instrument should be the same for proper operation.
If the Konakt 2 Player responds to MIDI RPNs (Registered Parameter Numbers), I may be able to set the required range by sending RPN pitchbend range messages from the guitar-midi converter.
Re portamento requiring overlapping notes, I'm aware of that. However, it can apparently also be achieved using MIDI CCs, because wind controllers don't provide overlapping notes, but appear to work properly with the Trumpet and Mr Sax T.
BTW, where can I read the SyncroWaveTrig patent details?
I didn´t follow this thread, but today I had the first opportunity to read it. You´ve apparently explained the problem in a wrong way. You don´t need the full 2 octave range of the pitch bend - a steeper curve within the existing maximum of 2 semitones would be enough. This means - a certain, low PB value should lead to a much higher pitch change, otherwise the pitch bending you apply playing a MIDI guitar is not sufficiant.
So why didn´t you try the Wind Controller Mode? In this mode a new Pitch Sensitivity knob apperas in the GUI, which let´s you rescale the PB up to 400%. This might completely solve the problem, since we are using some particular curves, leading to a very high pitch change at very low PB values.
So, please try it and report what happens.
Morover, there are other possibilities like using free MIDI utilities allowing CC and PB rescaling...
Please note that The Trumpet was previously developed for keyboard use only and we are very happy we could also optimise it so well for Breath- and Wind-Controllers. At the moment we are not able to make any modiffications exclusively for MIDI guitar controllers - except advising the users to play it in the Wind Controller Mode with the 400% setting I mentioned above, or using some clever rescaling in case 400% is still not enough.
Firstly, I don't have Mr Sax T. and the Trumpet yet, so I can't try anyhing A demo version would be very useful! As soon as I can establish that they will work reasonably well for live playing with midiguitar (with and without a breath controlller) I will buy both.
Working with curves is not accurate enough. Apart from pitchbend, there are other things like hammer-ons and pull-offs that need to be accurately translated.
It's the best practice to use an extended pichbend range. It's so simple and works so well (as mentioned, the actual pitchbend amount applied to the instrument is only a fraction of the full range).
Does the Konakt 2 Player respond to MIDI RPNs (Registered Parameter Numbers)? If so, I should be able to set the required range by sending RPN pitchbend range messages.
I don't understand why the limitations are deliberatly built-in. Having a default configuration with those linitations is ok, of course, but beyond that, there should be total flexibility. Not just the pitchbend range, but also controller assignments and mapping, etc.
Udo wrote:
The large range is required to ensure string bends, etc, are translated accurately into proportionate pitchbend data. The range on the guitar-midi converter and instrument should be the same for proper operation.
Udo wrote:
Apart from pitchbend, there are other things like hammer-ons and pull-offs that need to be accurately translated.
Since you are using a guitar to play a trumpet, not everything you can do with a guitar will have a 1-to-1 correlation with the trumpet/sax (this distance goes for keyboards as well). The task of interpretation of string events into midi events is entirely up to your converter.
Regrading pitchbend, if your converter can be set to a smaller max pitchbend, then set it less and experiment to find which setting feels best (it may be 2 or it may not and external rescaling may still be necessary). Pitchbend capability greater than two semitones for a sax or trumpet is just not to be found in the sample library industry as a whole (more on this below). That being the case, it would be prudent for your converter manufacturer to make a product that possesses settings with this in mind.
Udo wrote:
I don't understand why the limitations are deliberatly built-in.
This is a virtual instrument. Players use a controller (keyboard, wind controller, guitar, etc...) to play, what is in essence, a virtualization of a real trumpet/sax. Therefore, it is constructed to behave as a real trumpet/sax. The virtual instrument intercepts pitchbend messages and responds just as the real instrument would. For instance, with Mr. Sax T, when the player moves the pitchbend wheel all the way down, the tone is bent two semitones down (as well as other effects added to emulate a real player bending). But, if the pitch wheel is moved all the way up, the tone does not bend two semitones higher, but rather only a fraction higher because this is how a real saxophone works (generally). To create pitchbend effects that cannot be produced with the real instrument would not be appropriate for a product going for realism (and could lead a player to embarrassingly believe that such bends are possible with the instrument when they are not :oops:).
Udo wrote:
If the Kontakt 2 Player responds to MIDI RPNs (Registered Parameter Numbers), I may be able to set the required range by sending RPN pitchbend range messages from the guitar-midi converter.
Although RPNs can be sent, this will not change the pitchbend behavior of the virtual instruments nor should it as explained above.
Udo wrote:
Re portamento requiring overlapping notes, I'm aware of that. However, it can apparently also be achieved using MIDI CCs, because wind controllers don't provide overlapping notes, but appear to work properly with the Trumpet and Mr Sax T.
Overlap must occur for portamento to be triggered. Also, the vast majority of wind controllers overlap the notes automatically on slur.
My opinion is that setting max pitchbend or rescaling or a combination of both is definitely the way to go and will yield the closest approximation of string bend to resulting midi pitchbend. I hope this info helps
Two key points I mentioned a number of times appear to be overlooked:
- There would be no chipmunk effect. Only a fraction of the extended pitch bend range is actually applied to the instruments by the guitar pitch-to-midi conversion after the translation process (and if required, playing technique can limit it further, to suit the instrument being emulated).
- I don't own Mr sax T and the Trumpet yet, so i can't experiment. I need to be sure that they can be made to work properly with midi guitar and breath controller before i buy (although keyboards will also be used).
I'm very familiar with adapting playing style and technique to suit the instruments being emulated. I've been using virtual instruments for a long time, including Yamaha VL physical modelling (using custom models), etc, and used mdi guitar (always with extended pitch bend range for translation) and keyboards (both with and without breath controllers). i have also used Yamaha WX wind controllers. webbiedave wrote:
Udo wrote:
I don't understand why the limitations are deliberatly built-in.
This is a virtual instrument. Players use a controller (keyboard, wind controller, guitar, etc...) to play, what is in essence, a virtualization of a real trumpet/sax. Therefore, it is constructed to behave as a real trumpet/sax ................
There are plenty of examples of musicians using conventional instruments in ways that were not intended (some have become common practice). Take the sax: false fingering, tongue slaps, and more
What I meant was that, outside the basic configuration designed to produce "normal" behaviour, the parameter ranges, etc, should be left more open/flexible for users to experiment with. Many of the results may be useless, but others could be interesting/inspiring and usable (although they may sound nothing like conventional instruments). It should be left to the user's discretion.
That flexibility could also have made The Trumpet and Mr. Sax T. suitable, "out of the box", for MIDI guitar. An extended pitch bend range for translation would have made things so simple and easy (and, as I said before, only a fraction would actually be applied to the instruments, so it would sound right).
Re your other suggestions, from past experience, working with curves, etc. is not accurate enough. Anyway, the guitar midi converter and ALL the connected instruments must have matching pitch bend ranges. So, even if it could be made to work with a 2 semitone range, it would be incompatible with the other instruments.
Re RPNs, did you actually try to change the pitch bend range by sending the CC sequence I mentioned in the RPN thread?
Pitch bend range selection has not been included in the Sax and Trumpet User Interface, but that doesn't necessarily mean the Kontakt Player doesn't respond. RPNs are precisely defined in the MIDI specs and most devices respond. I contacted NI about it, but no response yet.
Udo wrote:
Re RPNs, did you actually try to change the pitch bend range by sending the CC sequence I mentioned in the RPN thread?
Yes. They have no effect. This also holds true for many, many sample libraries as they decide how to respond to pitchbend messages depending on the particular instrument.
Rescaling has been suggested multiple times by multiple posters and it shouldn't be overlooked as a viable solution. It is a tool that can also be applied to countless libraries that also limit their pitchbend ranges.
As a wind controller (and keyboard) player, I have to -- many times -- come up with external solutions to better play particular libraries.
Comments
I'm using the ZTAR with "The Trumpet" and get great results. The Ztar tracks perfectly, I have the AXON MK11 and it's no good. You play one note and get 10 ghost notes on the screen..
I use for recording only for orchestral or combo. The only thing I don't use it for is guitar. Anyway, it has plently of touch controllers which you can program in keyswitching, vibrato. C11 etc... I have one patch set up just for the Trumpet. If you haven't already, check it out. The company is Starr Labs.
Tony
BTW, which version of the ZTAR are you using?
Tony
My main guitar-to-midi interface is an Axon AX-100 and it works very well. The problem you described, Tony, indcates incorrect setup (either parameter settings, hex pickup installation, or both). Playing style is a factor, but unless you technique is very sloppy, the Axon is quite forgiving. Alan, the performance of the GR-33 (I have one) is no match for the Axon, see below.
I've been using the Axon with Yamaha VL technology (physical modeling for sax and trumpet emulation, etc.) and a lot of other gear (incl. software) successfully for years.
I had a ZTAR in the past and I still have a "poor man's ZTAR", the Yamaha EZ-EG (there's also a EZ-AG). It's very similar to the ZTAR, but not as sophisticated (and not as well built), although when running it through a computer and using some scripting, you can make it do almost anything. I use it very little, because the Axon works very well.
The Roland GI-20 guitar midi interface, although not as fast as Axon, works quite well if you use a RMC piezo hex pickup as found on some Godin and Brian Moore guitars (and in custom installations on many other guitars). Tracking accuracy is noticeably better than with magnetic hex pickups (like the GK-3, etc.). On the Axons there's hardly any difference between the different pickups.
With Roland GR20/30/33 synths, the mag/piezo differences above apply, but the MIDI output is also slower than the GI-20. Response of the built-in sounds is not bad.
I'm not sure I can afford the entry level price of the ZTar ($2000 US) vs. the Axon ($700 US). It would be great if Starr Labs would lower the price bar for their technology.
Still thanks for all the info on the Guitar Controllers.
Good luck with that stuff. I have a Strat with a Roland Pickup and a Godin which comes with the pickup built into it. Maybe your'e right but I don't agree. As far as my teqnique, I've studied with the infamous Dennis Sandole and have done hundreds of "Big Name" sessions in NY so I don't think my technique is suffering too much. Anyway, I said what I had to say and this really is a disscusion for another FORUM.
I didn't mean to question your technique. I was trying to make Alan aware that the Axon is more "forgiving" then other guitar-to-MIDI converters, when playing technique is not precise.
Among some general midi-guitar info for Alan was my main issue: the KEY REQUIREMENT for effective control by a MIDI GUITAR of virtual instruments is a PITCH-BEND RANGE OF AT LEAST 12 SEMITONES (so that hammer-ons, pull-offs, slides, string-bends, etc, are translated properly). Looking at the documentation, this appears to be impossible in Mr sax T. and the Trumpet.
This a very important issue for midi guitarists (and some are not aware how important it is). I was about to start a thread about it, when Alan's post appeared. I'll still make it a separate thread (with some added issues, e.g. portamento, controller assignments, etc.), to get the developers' attention.
The apparrent pitch-bend limitation has stopped me from buying the sax and trumpet.
Although useful, the ZTAR (have owned one) and Yamaha EZ-AG/EG (still own one) are not real guitars, but "switchboards" that lack the (subtle) expression possible with coventional guitar technique. Those techniques are very usefull when playing virtual sax. I used it successfully with Yamaha VL physical modelling.
For me however, the Ztar would not be a good option in this situation since one requirement I have is to switch realtime between the guitar and plugins. It's also very informative to know that the Trumpet and Sax to not support the full-bend range. Again, I would like to thank both of you...for me (and I hope others), this has been a very informative topic and maybe the folks at Sample Modeling will be able to address the bend-range issue. Again thanks to both of you!
If you can promise that, I'll buy both the Sax and Trumpet shortly!
While I'm here, a couple more MIDI guitar tips for budding MIDI guitarists:
The main advantage Axon has over other converters (early transient detection) only applies when selecting 'plectrum/pick' as playing style. When you select 'finger', only normal pitch to midi conversion is used, which is slower.
For all converters, putting felt or something similar under the strings between the nut and 1st fret helps stop glitches when lifting fingers from the strings, etc. Should only be done when you don't intend to use open strings, of course.
Best,
Giorgio
There would be no chipmunk effect.
To clarify, although the pitchbend range needs to be large for Guitar pitch-to-MIDI conversion, the pitchbend amount applied to the instruments will only be a fraction of the full range.
The large range is required to ensure string bends, etc, are translated accurately into proportionate pitchbend data. The range on the guitar-midi converter and instrument should be the same for proper operation.
If the Konakt 2 Player responds to MIDI RPNs (Registered Parameter Numbers), I may be able to set the required range by sending RPN pitchbend range messages from the guitar-midi converter.
Re portamento requiring overlapping notes, I'm aware of that. However, it can apparently also be achieved using MIDI CCs, because wind controllers don't provide overlapping notes, but appear to work properly with the Trumpet and Mr Sax T.
BTW, where can I read the SyncroWaveTrig patent details?
I didn´t follow this thread, but today I had the first opportunity to read it. You´ve apparently explained the problem in a wrong way. You don´t need the full 2 octave range of the pitch bend - a steeper curve within the existing maximum of 2 semitones would be enough. This means - a certain, low PB value should lead to a much higher pitch change, otherwise the pitch bending you apply playing a MIDI guitar is not sufficiant.
So why didn´t you try the Wind Controller Mode? In this mode a new Pitch Sensitivity knob apperas in the GUI, which let´s you rescale the PB up to 400%. This might completely solve the problem, since we are using some particular curves, leading to a very high pitch change at very low PB values.
So, please try it and report what happens.
Morover, there are other possibilities like using free MIDI utilities allowing CC and PB rescaling...
Please note that The Trumpet was previously developed for keyboard use only and we are very happy we could also optimise it so well for Breath- and Wind-Controllers. At the moment we are not able to make any modiffications exclusively for MIDI guitar controllers - except advising the users to play it in the Wind Controller Mode with the 400% setting I mentioned above, or using some clever rescaling in case 400% is still not enough.
Best
Peter
Firstly, I don't have Mr Sax T. and the Trumpet yet, so I can't try anyhing
Working with curves is not accurate enough. Apart from pitchbend, there are other things like hammer-ons and pull-offs that need to be accurately translated.
It's the best practice to use an extended pichbend range. It's so simple and works so well (as mentioned, the actual pitchbend amount applied to the instrument is only a fraction of the full range).
Does the Konakt 2 Player respond to MIDI RPNs (Registered Parameter Numbers)? If so, I should be able to set the required range by sending RPN pitchbend range messages.
I don't understand why the limitations are deliberatly built-in. Having a default configuration with those linitations is ok, of course, but beyond that, there should be total flexibility. Not just the pitchbend range, but also controller assignments and mapping, etc.
Udo wrote: The large range is required to ensure string bends, etc, are translated accurately into proportionate pitchbend data. The range on the guitar-midi converter and instrument should be the same for proper operation.
Udo wrote: Apart from pitchbend, there are other things like hammer-ons and pull-offs that need to be accurately translated.
Since you are using a guitar to play a trumpet, not everything you can do with a guitar will have a 1-to-1 correlation with the trumpet/sax (this distance goes for keyboards as well). The task of interpretation of string events into midi events is entirely up to your converter.
Regrading pitchbend, if your converter can be set to a smaller max pitchbend, then set it less and experiment to find which setting feels best (it may be 2 or it may not and external rescaling may still be necessary). Pitchbend capability greater than two semitones for a sax or trumpet is just not to be found in the sample library industry as a whole (more on this below). That being the case, it would be prudent for your converter manufacturer to make a product that possesses settings with this in mind.
Udo wrote: I don't understand why the limitations are deliberatly built-in.
This is a virtual instrument. Players use a controller (keyboard, wind controller, guitar, etc...) to play, what is in essence, a virtualization of a real trumpet/sax. Therefore, it is constructed to behave as a real trumpet/sax. The virtual instrument intercepts pitchbend messages and responds just as the real instrument would. For instance, with Mr. Sax T, when the player moves the pitchbend wheel all the way down, the tone is bent two semitones down (as well as other effects added to emulate a real player bending). But, if the pitch wheel is moved all the way up, the tone does not bend two semitones higher, but rather only a fraction higher because this is how a real saxophone works (generally). To create pitchbend effects that cannot be produced with the real instrument would not be appropriate for a product going for realism (and could lead a player to embarrassingly believe that such bends are possible with the instrument when they are not :oops:).
Udo wrote: If the Kontakt 2 Player responds to MIDI RPNs (Registered Parameter Numbers), I may be able to set the required range by sending RPN pitchbend range messages from the guitar-midi converter.
Although RPNs can be sent, this will not change the pitchbend behavior of the virtual instruments nor should it as explained above.
Udo wrote: Re portamento requiring overlapping notes, I'm aware of that. However, it can apparently also be achieved using MIDI CCs, because wind controllers don't provide overlapping notes, but appear to work properly with the Trumpet and Mr Sax T.
Overlap must occur for portamento to be triggered. Also, the vast majority of wind controllers overlap the notes automatically on slur.
My opinion is that setting max pitchbend or rescaling or a combination of both is definitely the way to go and will yield the closest approximation of string bend to resulting midi pitchbend. I hope this info helps
Two key points I mentioned a number of times appear to be overlooked:
- There would be no chipmunk effect. Only a fraction of the extended pitch bend range is actually applied to the instruments by the guitar pitch-to-midi conversion after the translation process (and if required, playing technique can limit it further, to suit the instrument being emulated).
- I don't own Mr sax T and the Trumpet yet, so i can't experiment. I need to be sure that they can be made to work properly with midi guitar and breath controller before i buy (although keyboards will also be used).
I'm very familiar with adapting playing style and technique to suit the instruments being emulated. I've been using virtual instruments for a long time, including Yamaha VL physical modelling (using custom models), etc, and used mdi guitar (always with extended pitch bend range for translation) and keyboards (both with and without breath controllers). i have also used Yamaha WX wind controllers.
webbiedave wrote:
Udo wrote: I don't understand why the limitations are deliberatly built-in.
This is a virtual instrument. Players use a controller (keyboard, wind controller, guitar, etc...) to play, what is in essence, a virtualization of a real trumpet/sax. Therefore, it is constructed to behave as a real trumpet/sax ................
There are plenty of examples of musicians using conventional instruments in ways that were not intended (some have become common practice). Take the sax: false fingering, tongue slaps, and more
What I meant was that, outside the basic configuration designed to produce "normal" behaviour, the parameter ranges, etc, should be left more open/flexible for users to experiment with. Many of the results may be useless, but others could be interesting/inspiring and usable (although they may sound nothing like conventional instruments). It should be left to the user's discretion.
That flexibility could also have made The Trumpet and Mr. Sax T. suitable, "out of the box", for MIDI guitar. An extended pitch bend range for translation would have made things so simple and easy (and, as I said before, only a fraction would actually be applied to the instruments, so it would sound right).
Re your other suggestions, from past experience, working with curves, etc. is not accurate enough. Anyway, the guitar midi converter and ALL the connected instruments must have matching pitch bend ranges. So, even if it could be made to work with a 2 semitone range, it would be incompatible with the other instruments.
Re RPNs, did you actually try to change the pitch bend range by sending the CC sequence I mentioned in the RPN thread?
http://www.samplemodeling.com/forum/viewtopic.php?f=2&t=211
Pitch bend range selection has not been included in the Sax and Trumpet User Interface, but that doesn't necessarily mean the Kontakt Player doesn't respond. RPNs are precisely defined in the MIDI specs and most devices respond. I contacted NI about it, but no response yet.
Yes. They have no effect. This also holds true for many, many sample libraries as they decide how to respond to pitchbend messages depending on the particular instrument.
Rescaling has been suggested multiple times by multiple posters and it shouldn't be overlooked as a viable solution. It is a tool that can also be applied to countless libraries that also limit their pitchbend ranges.
As a wind controller (and keyboard) player, I have to -- many times -- come up with external solutions to better play particular libraries.