Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Pitch bend range
178 views
Skip to first unread message
Jos
Feb 22, 2012, 9:10:06 AM2/22/12
to
Hi,
Say I want to do wild things on my sampled fretless bass.
I set the pitch bend range to -12 - + 12 semitones. That means that Pitch
Bend value +8191 bends my bass tone an octave higher.
Now what do I do if I want to bend the tone a major third upwards? That
is 4 semitones. I would think, things being linear, that I would pick a
pitch bend value of (4/12) * 8191 = 2730. But that sounds way too high.
Something logarithmic must be going on here, but I can't manage to work
it out on my own.
Is there a specification for this? Of course it may depend on the
developer of the instrument to put out whatever frequency for a given
pitch bend value. But I would have thought there would be a standard for
it.
Any pointers?
Thanks,
Jos.
Say I want to do wild things on my sampled fretless bass.
I set the pitch bend range to -12 - + 12 semitones. That means that Pitch
Bend value +8191 bends my bass tone an octave higher.
Now what do I do if I want to bend the tone a major third upwards? That
is 4 semitones. I would think, things being linear, that I would pick a
pitch bend value of (4/12) * 8191 = 2730. But that sounds way too high.
Something logarithmic must be going on here, but I can't manage to work
it out on my own.
Is there a specification for this? Of course it may depend on the
developer of the instrument to put out whatever frequency for a given
pitch bend value. But I would have thought there would be a standard for
it.
Any pointers?
Thanks,
Jos.
Gerd-Ulrich Meyer
Feb 22, 2012, 12:23:28 PM2/22/12
to
Hallo,
On 22.02.2012 15:10, Jos wrote:
> Hi,
>
> Say I want to do wild things on my sampled fretless bass.
> I set the pitch bend range to -12 - + 12 semitones. That means that Pitch
> Bend value +8191 bends my bass tone an octave higher.
> Now what do I do if I want to bend the tone a major third upwards? That
> is 4 semitones. I would think, things being linear, that I would pick a
> pitch bend value of (4/12) * 8191 = 2730. But that sounds way too high.
> Something logarithmic must be going on here, but I can't manage to work
> it out on my own.
>
Well, as MIDI is strictly linear, there couldn't be any logarithm, in my
opinion.
> Is there a specification for this? Of course it may depend on the
> developer of the instrument to put out whatever frequency for a given
> pitch bend value. But I would have thought there would be a standard for
> it.
>
> Any pointers?
After looking at
http://www.ultimatemetal.com/forum/off-topic/680677-midi-pitch-bend-range.html
(maybe you have to copy and paste the link)
i agree, that your pitch-bend value is correct. Have you checked the
other values (1,2,3 semi-tones for example)? Are they correct?
Otherwise i would have a look at my sample respectively at my
sample-player (i don't know, if you have a sampler, a sound-card or
something else).
>
> Thanks,
>
> Jos.
Maybe it's a first small help.
Gerd-Ulrich Meyer
On 22.02.2012 15:10, Jos wrote:
> Hi,
>
> Say I want to do wild things on my sampled fretless bass.
> I set the pitch bend range to -12 - + 12 semitones. That means that Pitch
> Bend value +8191 bends my bass tone an octave higher.
> Now what do I do if I want to bend the tone a major third upwards? That
> is 4 semitones. I would think, things being linear, that I would pick a
> pitch bend value of (4/12) * 8191 = 2730. But that sounds way too high.
> Something logarithmic must be going on here, but I can't manage to work
> it out on my own.
>
opinion.
> Is there a specification for this? Of course it may depend on the
> developer of the instrument to put out whatever frequency for a given
> pitch bend value. But I would have thought there would be a standard for
> it.
>
> Any pointers?
http://www.ultimatemetal.com/forum/off-topic/680677-midi-pitch-bend-range.html
(maybe you have to copy and paste the link)
i agree, that your pitch-bend value is correct. Have you checked the
other values (1,2,3 semi-tones for example)? Are they correct?
Otherwise i would have a look at my sample respectively at my
sample-player (i don't know, if you have a sampler, a sound-card or
something else).
>
> Thanks,
>
> Jos.
Maybe it's a first small help.
Gerd-Ulrich Meyer
Hueyduck
Feb 22, 2012, 1:55:13 PM2/22/12
to
Jos a écrit :
>
> Any pointers?
>
Hi Jos,
like in life, you have to check how the raw data is used and transformed
by your software or hardware.
For instance, in Kontakt, you can set the velocity not to have be
linearly linked to volume. You can chose any curve you need.
Could it be that between cubase and the sound generator you are using,
there is a setting that allows the response of the bend not to be linear?
And anyway, if you have to set things right for a fews note, you'd
better set them by ear ;-)
Huey
>
> Any pointers?
>
Hi Jos,
like in life, you have to check how the raw data is used and transformed
by your software or hardware.
For instance, in Kontakt, you can set the velocity not to have be
linearly linked to volume. You can chose any curve you need.
Could it be that between cubase and the sound generator you are using,
there is a setting that allows the response of the bend not to be linear?
And anyway, if you have to set things right for a fews note, you'd
better set them by ear ;-)
Huey
swanny
Feb 22, 2012, 6:02:19 PM2/22/12
to
are being used.
If this were so, the value 20 would be 4 semitones above the root.
Jos
Feb 24, 2012, 4:14:46 AM2/24/12
to
Thanks Gerd-Ulrich and all the other people.
From that link (and from some other experiments) I conclude that it is
indeed supposed to be linear.
I'm not sure what went wrong, then, but I tried it again yesterday on a
clean project and it seems that it is indeed linear on this given sound
module. That is, at -12 - +12 semitones pitch bend sensitivity, I measure
the +4096 pitch bend as giving a tritone (+6 semitones) above the
original pitch. This is on an old Roland JV-2080 module with a Bass&Drums
expansion card.
And in the original project, I ended up tuning things by ear, of course.
So thanks for helping me sort this out.
Have fun,
Jos.
On Wed, 22 Feb 2012 18:23:28 +0100, Gerd-Ulrich Meyer wrote:
> http://www.ultimatemetal.com/forum/off-topic/680677-midi-pitch-bend-
range.html
From that link (and from some other experiments) I conclude that it is
indeed supposed to be linear.
I'm not sure what went wrong, then, but I tried it again yesterday on a
clean project and it seems that it is indeed linear on this given sound
module. That is, at -12 - +12 semitones pitch bend sensitivity, I measure
the +4096 pitch bend as giving a tritone (+6 semitones) above the
original pitch. This is on an old Roland JV-2080 module with a Bass&Drums
expansion card.
And in the original project, I ended up tuning things by ear, of course.
So thanks for helping me sort this out.
Have fun,
Jos.
On Wed, 22 Feb 2012 18:23:28 +0100, Gerd-Ulrich Meyer wrote:
> http://www.ultimatemetal.com/forum/off-topic/680677-midi-pitch-bend-
range.html
Gerd-Ulrich Meyer
Feb 24, 2012, 5:30:15 PM2/24/12
to
Hello,
On 24.02.2012 10:14, Jos wrote:
> Thanks Gerd-Ulrich and all the other people.
>
No problem, it was me a pleasure to help.
> From that link (and from some other experiments) I conclude that it is
> indeed supposed to be linear.
That's the point where i (and the others) am not too sure. It seems to
depend on the sound module you use.
> I'm not sure what went wrong, then, but I tried it again yesterday on a
> clean project and it seems that it is indeed linear on this given sound
> module. That is, at -12 - +12 semitones pitch bend sensitivity, I measure
> the +4096 pitch bend as giving a tritone (+6 semitones) above the
> original pitch. This is on an old Roland JV-2080 module with a Bass&Drums
> expansion card.
Something i only know by name.
> And in the original project, I ended up tuning things by ear, of course.
>
That's always the best method - we make music for the ears, not for the
mathematical brain :-)
> So thanks for helping me sort this out.
>
> Have fun,
>
> Jos.
Greetings
Gerd-Ulrich Meyer
On 24.02.2012 10:14, Jos wrote:
> Thanks Gerd-Ulrich and all the other people.
>
> From that link (and from some other experiments) I conclude that it is
> indeed supposed to be linear.
depend on the sound module you use.
> I'm not sure what went wrong, then, but I tried it again yesterday on a
> clean project and it seems that it is indeed linear on this given sound
> module. That is, at -12 - +12 semitones pitch bend sensitivity, I measure
> the +4096 pitch bend as giving a tritone (+6 semitones) above the
> original pitch. This is on an old Roland JV-2080 module with a Bass&Drums
> expansion card.
> And in the original project, I ended up tuning things by ear, of course.
>
mathematical brain :-)
> So thanks for helping me sort this out.
>
> Have fun,
>
> Jos.
Gerd-Ulrich Meyer
0 new messages