Hi,
using pianoteq 3.02 as vst in Max/MSP I encountered this issue: pitchbend commands directed to different mici channels are recognized only on channel 1. Can you check this please?
I need the different channels detuned by pitchbend.
Ernst
I've had the same problem; a midi file with the piano on Channel 1, and a flute voice on Channel 3. The flute track has some pitch bends.
Even though I selected only Channel one in Pianoteq (Options/Midi/Notes Channel), the pitch bends for the flute (channel 3 / track 3) were used by Pianoteq for the piano part.
I doubled checked this on several occasions, and am listening to a piano with pitch bends as I type this.
Glenn
Hi Glenn,
yes, this is exactly the same problem. It did'nt occur in PT2. But the same behaviour in both V2 and V3 show the note-off messages. If you send a note 64 on to channel 1 and then a note-off 64 to channel 2, the note on channel 1 stops. So the note-offs aren't "channel-true" either.
Ernst
Ugh... this should be checked with all the other MIDI messages too, then. Maybe they are all involved somehow
I'm sure as soon as Modartt guys see this topic in a few days there will be 3.03 patch out
ouch, that's quite a nasty bug!
ouch, that's quite a nasty bug!
Yes, you should hear a piano with pitch bends that are on the flute track in a piece by McPartland and Peggy Lee (In The Days Of Our Love - sometimes called Afterglow).
It would be bad enough applied to the melody, but of course the bends are applied to the left hand bass notes too.
In order to get rid of the bends, I had to create a duplicate MIDI file, and remove the flute track from the file to get the piano sound clean.
Ugh is a good word for this one.
Glenn
The good thing about this forum is that it's a BugBuster !
:-)
It is a bit tricky to fix since we expose only one "pitchbend" parameter instead of 16 (one for each channel), but we will try to hack something for the next update.
Hey, I can't replicate this bug. Tested in FL8:
Inserted PTQ3, set in MIDI Notes channel 1, and PITCH BEND to 1 instead of "ANY" (this is important, I think!), set MIDI in/out port to 0.
Inserted NI Colossus, loaded any instrument, set the MIDI channel of that track to 2, set MIDI in/out port to 0 (so they are on the same port!.
Played a piano run while sustain was held.
Switched to Colossus and played with pitch bend.
Pianoteq didn't pitch bend.
So I suggest you to change MIDI channel next to Pitch Bend from ANY to the MIDI channel you're actually using! And then the PB messages from all the other channels WILL NOT interfere!
MODARTT, no bug correction is necessary! "Notes channel" is used only for detecting note on messages from the MIDI stream! For all other parameters, you have a parameter to decide from which MIDI channel they should receive the messages!
Yes, you are right, what you are describing is the correct solution for the problem of Glenn.
However the problem of Ernst is a bit different: with current versions, any pitchbend that gets through the midi filter affects only notes which have been played on channel 1. What Ernst wants is "pitch bend sent on channel x affects all notes which have been triggered on channel x". This is a trick often used for doing some sort of microtuning via pitch-bend messages (with limited polyphony as there are only 16 channels !)
OK, I guess I need some more detailed instructions on how to use the midi mapping, etc.
I have a file as follows:
Track 1 / Channel 1 = piano
Track 2 / Channel 2 = vibes
Track 3 / Channel 3 = flute
Track 4 / Channel 4 = bass
Track 5 / Channel 10 = rhythm
I have PT Options/Midi set as follows:
Current Midi Mapping - Pitch bend Channel 3
Notes Channel = 1 (this plays the piano sound only)
When the file is being played, the screen in Options/Midi shows the midi controls scrolling by, and the Pitch Bends are definitely on Channel 3. But the piano sound has the pitch bends from Channel 3.
How do I set up PT to avoid this?
I have saved this Pitch Bend setting as "Pitch Bend Channel 3".
I've already submitted the file to PT Help, but the info I got back said the file was fine. The problem I'm having is that I don't find the manual covers this area in any detail, so I can't figure it out.
Glenn
Glenn, you are triggering vibes, flute, bass, drums from other VST I guess?
Try removing the setting Pitch Bend Channel 3 to Channel 1 only and you won't get bends from the Flute channel.
When you set Pitch Bend to Channel 3, this tells Pianoteq to apply pitch bend messages FROM that channel (3) to the Notes channel (which is, in your case, 1).
Glenn, you are triggering vibes, flute, bass, drums from other VST I guess?
Try removing the setting Pitch Bend Channel 3 to Channel 1 only and you won't get bends from the Flute channel.
When you set Pitch Bend to Channel 3, this tells Pianoteq to apply pitch bend messages FROM that channel (3) to the Notes channel (which is, in your case, 1).
ED:
I'm not triggering the non-piano channels with anything - I use Pianoteq in stand alone mode. (As an aside I was going to try the True Pianos demo, but it requires a host which I don't have or need or want).
I'm don't want anything from the other channels to sound (2, 3, 4, and 10) - they are used exclusively for vibes, flute, bass and drums respectively.
When PT is set with Pitch Bends as "Any" I get pitch bends on
Channel 1. When PT is set with Pitch Bends on Channel 3, I get pitch bends on channel 1 (piano).
It seems that no matter what setting I use, the pitch bends are applied to the piano (channel 1).
How do I tell PT to NOT use the pitch bends that are on a different channel than the piano.
Other than delete the channel(s) with pitch bends, there must be some way to not have the piano channel use the bends.
Glenn
I can't replicate that. Seems like the bug has something to do with playing the MIDI files with the in-built player, than directly contacting with other synths over other MIDI channels, I don't know.
Try this: set the range of both positive and negative side to 0, then you won't have pitch bend at all
ED:
I tried setting the pitch bend range zero to zero.
Didn't work.
Thanks for the idea.
Glenn
When PT is set with Pitch Bends as "Any" I get pitch bends on
Channel 1. When PT is set with Pitch Bends on Channel 3, I get pitch bends on channel 1 (piano).It seems that no matter what setting I use, the pitch bends are applied to the piano (channel 1).
If you set PT to only listen to pitch-bends coming from channel 1 (that is your piano channel), it should not play any pitch bend from other channels. That is, in the midimappings list, on the row for "Pitch bend" parameter, set the channel filter (next to the "Pitch Bend" label) to 1. Then pianoteq will ignore any pitchbend message that is coming from any other channel.
This is exactly what I wished to say, and I thought I was clear enough.
That is, in the midi mappings list, on the row for "Pitch bend" parameter, set the channel filter (next to the "Pitch Bend" label) to 1. Then Pianoteq will ignore any pitch bend message that is coming from any other channel.
Ah so. Now it's clear, and it works.
I had interpreted it as working the other way around.
The manual doesn't seem to cover much ground on this feature (Section 3.3.3. on page 18 of the English version).
Glenn
What Ernst wants is "pitch bend sent on channel x affects all notes which have been triggered on channel x". This is a trick often used for doing some sort of microtuning via pitch-bend messages (with limited polyphony as there are only 16 channels !)
Yes, Julien, my problem isn't solved by that. I have to return to V2.3 because this works. I am wondering, why V3 is not working the pichbend in a channelspecific way. The same, as I mentioned previously, applies to the note-off-messsages.
Will there be a fix to this?
I mean, it's bad enough that pitchbend doesn't work notespecific so it should at least seperate the channels. Please take this requirements of mine into serious consideration...!
I am wondering, why V3 is not working the pichbend in a channelspecific way.
You can consider that as a bug, next version will behave like version 2.3. But having note-off that take into account the channel is another problem, which does not fit very well into the "physical modelling" approach , there is only 1 virtual piano, not 16 different pianos stacked together, so that is a bit complicated.
I mean, it's bad enough that pitchbend doesn't work notespecific so it should at least seperate the channels. Please take this requirements of mine into serious consideration...!
First - pitch bend was never (in MIDI specification) note-specific, it is monophonic and applies to all notes on the same channel in the same way.
And, as we stated above, it WORKS in channel-specific way, you just have to set it up in the MIDI screen manually.
Well, I believe I made my point clear to Julien. While you seem to be happy with one channel pitchbend and that's nice for you! But it obviously doesn't solve the matter which I brought to attention in this thread. So please don't lecture me on the base of shortsighted misunderstandings. That doesn't bring the issue forward. I admit, my ambition is a special use, but it is not yet too late to challenge industrial dogmas as midi.
And by the way: this pitchbend-issue is a bug even within the dogma. What I tried to say is: I am already compromising with using pitchbend in multichanneling. This is a workaround the missing note-specificness of pitchbend commands. And this workaround is now blocked by the mentioned bug that makes it impossible to run Pianoteq on more than one midichannel with different tunings.
And the missing channel-specificness of note-offs makes it impossible to play a legate quartertone scale on two alternating midichannels because Pianoteq doesn't separate the note-offs by channels. Which is amazing since Pianoteq well separates the note-ons. Maybe, Julien, it's not that difficult to extend this behaviour to the note offs as it seems? It's just imagining a layer of 16 keyboards and about 2800 strings...?
There is only one device that I know which is capable of this and that's my Kawai MP 9500. This is simply a sampler. But there I am stuck with the build-in sound.
You see, that's what drives me to this forum: I'm looking for progress so I send a ticket when I meet regress.
And by the way: this pitchbend-issue is a bug even within the dogma.
It's not a bug, it's a feature
Really, there are not THAT many instruments that exert polyphonic pitch bends of various degrees that would justify a polyphonic pitch bend midi message. Imagine how would a controller for such beast look like! Per-note pitch bend? On 88 keys? That's kinda ridiculous, if you ask me. I think polyphonic aftertouch is a decent substitute for that.
Anyways, that issue aside, there could be a workaround for you, if a little unorthodox.
Load up 16 instances of PTQ3 and set each one to a different midi channel (both Notes channel and Pitch Bend channel to the same channel), then you should have the separation you need, hopefully.
I don't know about the note-offs in that situation, but if they aren't separated by channel, that's also something that should be taken care of, Modartt. Note-on and Note-off make a pair and they should belong to the same channel they're played on.
I'm sorry if you got the feeling of me lecturing you, I was just stating that polyphonic pitch bend didn't have enough justification to be included in MIDI standard when it was created. And even today, I don't see much point in it, really, as it's rarely used.
I'm actually quite surprised that Pianoteq (the primary purpose of which is to model PIANOS), even has pitch bend.
Don't take following too seriously - it's meant as "tongue in cheek".
I was in the shop of my piano restorer/dealer friend yesterday pm, and we couldn't find one piano out of eight grands and four uprights that had pitch bend!!
However, one of the grands he just acquired (a 1926 Hamburg Steinway 170 model O that has a fabulous sound) has key bushings that will need replacing as the keys tend to wobble a bit.
We were wondering if perhaps this was intended to achieve pitch bends as those German piano builders don't seem to do anything without a purpose in mind.
Glenn
I just run Pianoteq through a wah-wah to get my pitch bends.
Hmmmm... An idea (or a vessel) just popped in my head: I don't have PT on this office computer -anybody know if you can send just the sympathetic tones out a separate output, say for processing differently from the actual notes.....? (..probably should start a new string... and no, I didn't search first.....)
To the OP: have you considered getting a Haken Continuum?
Load up 16 instances of PTQ3 and set each one to a different midi channel (both Notes channel and Pitch Bend channel to the same channel), then you should have the separation you need, hopefully.
Hi EvilDragon...:There is of course a CPU liimit to your solution. Mabe you know, there is a saying: it's like shooting sparrows with a cannon.
But seriously, in all this involvement in tricks and workarounds I almost forgot to say that also the use of pitchbend on 16 detuned channels is a compromise for microtonality. And it's also a waste of midi-channels by the way. The best und really simpliest would be, if midi notes were not integers but decimals. If each Midi note would have two decimal digits it would contain enough information for all conceivable tunings. And it would be really simple.
For example a quartertone above 60 would be 60,5 and not 60 on channel 9. I have built an"hammond" organ in Max/msp which is based on those midi-cent note information. It's working on one single midi-channel. And the decimals don't give the CPU any more work than Midi-integers. What causes CPU-sweat are voices - and voices I don't need more than a welltempered piano does because I play mostly with my hands on a normal keyboard.
In fact that's my real wish to pianoteq: to learn midi-cent!!
But the pitchbend-channel-bond would be nice as well (for the start...).
Ernst, with all due respect, I think that what you're wishing for is a liiiiitle bit out of scope for Pianoteq. It's a virtual piano, based on a real piano which doesn't have that smooth pitch division. Also, MIDI doesn't quite support it per-note, as far as I know. Or, if it does support it, it needs several other types of messages PER note (other than note on and off) to detune the sound. You're the only one here requesting such things as midi-cents, so IMHO I can't see this happening with Pianoteq. On the other hand, the mentioned error that a note-off from another channel cuts off note-ons from PTQ Notes channel should be solved (and is even obligatory!).
Your only solution might be a Haken Continuum (as it doesn't have keys, right?) plus a Kyma Paca(rana) connected and doing marvellous stuff.
You don't even need Kyma, Max/MSP can do wonders when controlled by Continuum, too. The only caveat is the price of the darn thing.
Like I said, most of the acoustic pianos I've been able to find don't have pitch bend - and if they have I haven't found the pedal for it.
I just realized - some pianos have four pedals - maybe that' what the fourth pedal is for.
The purpose of Pianoteq (at least as I understand it) is to duplicate the sound of an acoustic grand piano (and uprights too), I'm not sure why pitch bend is included. In the version called "Violteq" it would be essential.
Glenn
Or, if it does support it, it needs several other types of messages PER note (other than note on and off) to detune the sound. You're the only one here requesting such things as midi-cents, so IMHO I can't see this happening with Pianoteq.
Well, what the hell. I don't see it either. But my idea of a virtual piano is that it is not only smaller and cheaper but that it can do other things than a real one. It is a kind of hybrid entity, it's deriving from a physical conception and a sound that we love. But in itself the virtual piano is completely different. It has it's own reality. So it sould emancipate from the limited possibilities of a physical piano. What it already does where customers ask for it Why not follow this track of invention furthermore?
To the "detuning" of the sound or rather pitch: The midi note contains all information about pitch and this stays so even if you give it two digits more. In the state being you can regard the midinote as the pitch information 60,00 which is exactly what comes out of the synthesizer in Hz. Not about 60. Exactly 60,00. So it doesn't need more information to create 60,25. Only a string in the program. Or two...
I was indeed looking this Hakan Continuum up. That looks quite nice. And is full of possibilites. Actually I am a pianist, and keys are my interface (that much to limitations...) and this Hakan Continuum obviously has no keys, rather a long pad. Is it expensive?
I was indeed looking this Hakan Continuum up. That looks quite nice. And is full of possibilites. Actually I am a pianist, and keys are my interface (that much to limitations...) and this Hakan Continuum obviously has no keys, rather a long pad. Is it expensive?
It's unfortunately expensive. 3400$ for the almost 4-octave one, 5400$ for the almost 8-octave one. You can purchase the "pad" which has labeled black keys for reference separately. It's all explained on Haken's page.
It's my dream to have one. Plus, it looks weird and beautiful.
http://www.cerlsoundgroup.org/Continuum...faces.html
http://www.cerlsoundgroup.org/Continuum...rices.html
You have a lot of examples of what can be done with it in the Examples section!
I was indeed looking this Hakan Continuum up. That looks quite nice. And is full of possibilites. Actually I am a pianist, and keys are my interface (that much to limitations...) and this Hakan Continuum obviously has no keys, rather a long pad. Is it expensive?
It's unfortunately expensive. 3400$ for the almost 4-octave one, 5400$ for the almost 8-octave one. You can purchase the "pad" which has labeled black keys for reference separately. It's all explained on Haken's page. It's also limited to 16 note polyphony, because it sends pitch info one note per MIDI channel. But, it's not really intended to play piano. On the other hand, microtonality is its forté.
It's my dream to have one. Plus, it looks weird and beautiful.
http://www.cerlsoundgroup.org/Continuum...faces.html
http://www.cerlsoundgroup.org/Continuum...rices.html
You have a lot of examples of what can be done with it in the Examples section!
Forgive me, who am I to say... but this interface withou real keys would not allow a real pianist to play like in a real piano, cause he would be used with the feal of real keys and the fact of the keys feeling help to locate and be aware about position of other keys around.
True, it's a different type of controller. And I already said that it has the optional strip which has a two-color scheme to discern the pitches. The black strips are also differently textured so you can feel it's a different "key".
Mind you, Continuum isn't meant to do piano. It's a revolutionary new thing that's meant for exploring microtonality. And since ernst's posts made me aware of his wishing into exploring that, I recommended it.
Anyone can play it, it's just a matter of time to getting adjusted to a slightly different controller. Like when you switch from piano to synth action keys.
Great! The update solves the pitchbend issue! Thanks and best wishes to all Pianoteqnicians!!!:-)