Pitch bend range

37rog · 2025-07-27T10:59:47+00:00

There was an error rendering this rich post.

My pitch bend wheel broke, or rather the potentiometer itself (the graphite coating wore off), unfortunately I can't send it for repair, so I have to replace it myself. I picked up a potentiometer with the same resistance characteristics, but I don't remember how many semitones it should regulate. According to the standard, the midi pitch bend regulates 2 semitones down and 2 semitones up, but mine is completely different, it regulates a little more than 1 semitone up and also down.

Could anyone with working A-series keyboards tell me how many semitones your pitch bend wheel regulates and what readings the midi monitor gives in the lower and upper positions of the wheel?

I am attaching screenshots of my readings, are they correct?

Find more posts tagged with

Comments

JesterMgee

How many semitones the PB wheel adjusts an instrument is typically up to the instruments PB range setting (if it has one). Most instruments will have a PB range setting to set how many semitones the wheel bends so that isn't really going to offer any insight if your wheel is working nor does the screenshots you posted. The actual physical range you will see in the instruments GUI will always be the full range, just the affected bend range is dependant on the PB range setting of the actual instrument if it has one.

The value of the PB range if its a 8bit CC range will be from min 00 > FF max in hex or 00 > 127 in dec with the PB wheel sitting at approx 40 hex / 64 dec in the rest position.

A quick and simple visual test would be to open any plugin that shows the PB as part of the instrument (Kontakt for instance) then adjust your PB wheel. You should see it adjust to the position you adjust it on your keyboard. Hopefully there is no factory calibration needed for the pot value as this will be something you will be unable to do, maybe it was designed smart enough to test the value on startup otherwise you may find your center point is off.

Without any explanation of what your images are showing, impossible to tell but if this is an example of your bend range being all the way down:

Then this is not right and seems the value of the pot you installed is likely not the correct value.

Did you test the one that came out or does it have any value marking? Was the one you installed the exact same value?

37rog

I use the MidiView program for midi monitoring, and I do not understand why the values in the extreme positions do not correspond to either 1 semitone or 2 semitones as per the midi standard, but are somewhere between them. With the modulation wheel, the extreme positions give values of 0 and 127 as it should be.

I installed the same potentiometer with the marking as the original one - B10K, and also measured the resistance: the original potentiometer had a resistance of 10.8 kOhm, the new one had 10.2 kOhm, and the new one also had a shaft with a chamfer in the opposite direction, but I turned it around and adjusted the central position, as it was on the original potentiometer, I do not think that all this can be the reason. And by the way, the original potentiometer has recently behaved very incorrectly in the extreme positions, well, perhaps this is due to the erased graphite ring.

I would like to know what values a working pitch bend wheel produces without any virtual instruments running in a program, such as MidiView. Maybe that's how it should be for me, or maybe I need to continue looking for the cause and solution.

JesterMgee

I would like to know what values a working pitch bend wheel produces without any virtual instruments running in a program, such as MidiView. Maybe that's how it should be for me, or maybe I need to continue looking for the cause and solution.

Won't matter if an instrument is loaded or not, the values it should read are as mentioned:

Min 0 and Max 127 (FF)

I'm surprised the original one had any issue with the tracks wearing off unless you hammer it constantly, I have gear that is 30+ years old (including an old M-Audio midi controller) that has been well used and have still never needed a pot change. In fact i've only ever replaced one pot in gear I own where the brushes had broken.

You may have to pull it apart again and do some more tests but it could simply be the pot, although is 10k is not the same resolution as the original. Maybe it's a difference between a linear and log pot (unlikely) or just maybe the value is different in the center than the ends, something that may need calibration which I am sure NI can do but that service tool isn't available to us plebs.

nico5

isn’t pitch bend a 14 bit number (not 16), cobbled together from the 2 MIDI databytes, which are 7 bits each?

So the max value of a pitchbend message is 16383 , while minimum is 0.

turning the PB wheel to the max should return 16383 in a midi monitoring program.
Having it in the middle (no bend) should show 8192,
and all the way to the bottom should show 0.

How many notes that represents is entirely up to the receiving hardware or software and often even different for different sounds/programs/patches in the same hardware or software. I’ve regularly seen everything from 2 semitones to 2 octaves up or down

And of course in non western, and/or more microtonal music it may be used even more differently.

In the pitchbend sending hardware, there’s probably a way to calibrate the potentiometer, so it maps its analog voltage to the required digital numbers between 0 and 16383.

Related reading:
https://sites.uci.edu/camp2014/2014/04/30/managing-midi-pitchbend-messages/

37rog

https://community.native-instruments.com/discussion/comment/235504#Comment_235504

yes, I also have m audio for a long time and everything is fine there, it is strange that the potentiometer on this keyboard quickly failed and I did not hit it)

marking B10K means linear, but I think I will try logarithm again. I also have an assumption about the dead zone in the center, because of which the voltage does not quite reach the maximum values.

Thank you for your help!

37rog

https://community.native-instruments.com/discussion/comment/235522#Comment_235522

thanks for the answer, now I know for sure that the maximum value should be 16383 and I will continue to look for a solution.

JesterMgee

isn’t pitch bend a 14 bit number (not 16), cobbled together from the 2 MIDI databytes, which are 7 bits each?
So the max value of a pitchbend message is 16383 , while minimum is 0.

turning the PB wheel to the max should return 16383 in a midi monitoring program.
Having it in the middle (no bend) should show 8192,
and all the way to the bottom should show 0.

Depends. I mistakenly stated 16bit where it should have been 8. Original MIDI standard was a low resolution 8bit value of 0-127 and since he was using the ableton midi monitor showing the 8 byte value I picked up on that:

64 is the mid point so 0 should be the lower value and 127 the higher.

i've not had the need for a long time to check midi data.

nico5

I mistakenly stated 16bit where it should have been 8. Original MIDI standard was a low resolution 8bit value of 0-127 and since he was using the ableton midi monitor showing the 8 byte value I picked up on that:

I'm afraid, the Ableton MIDI monitor and your memory about how MIDI messages are constructed are misleading you.

MIDI values are always effectively 7 bit, since the first bit always has to be 0 - indicating that the byte is a databyte, rather than a statusbyte. 0 to 127 only takes 7 bits - full 8bit values would be 0 to 255.

That's because there are two kinds of MIDI message bytes:

MIDI status byte (always starts with a 1 as the first bit)
- generally contains the message type plus the midi channel
MIDI data byte (always starts with a 0 as the first bit)
- databyte 1 typically contains cc numbers, note numbers etc.
- databyte 2 typically contains cc values, note velocities etc.
- in pitchbend both databytes are combined to form a 14 bit value.
- channel aftertouch and program change messages do not have a databyte 2

There's more, but if you don't believe me, here's the official MIDI 1.0 spec:

https://midi.org/summary-of-midi-1-0-messages

There are also myriads of explanations all over the web. I also mentioned one of the top Google results in my prior post (although the NI forum in it's mysterious ways didn't make it a clickable URL).

JesterMgee

K, will leave this conversation.

37rog

I still can't solve this problem with the potentiometer, but now I know the exact reason. Their original potentiometers are not like the usual ones in stores, although they are identical in appearance. Their Pitch Bend potentiometer has a S-shaped curve of resistance dependence on the rotation angle, so the values change quickly from the central position up and down, and slow down at the edges, and it turns out that its full rotation angle is about 240-260 degrees. On a purchased potentiometer, the full rotation angle is 300 degrees and 40 degrees are superfluous, because of this, the sensitivity drops when turning the handle. It would be possible to install a purchased potentiometer if there was an opportunity to calibrate, but such potentiometers are not available anywhere in stores. In stores there are ones with a smaller degree, even with 260 degrees, but again, they are not suitable in design. In the end, I came to two solutions: the first was to try to insert a potentiometer of the wrong design into the pitch wheel holder, and the second was to grind down the stops on the wheel itself or on the holder so that the potentiometer would turn to the desired voltage.

37rog

I restored the original potentiometer using conductive glue, thanks to everyone for the advice, the topic can be closed.

Quick links

Unanswered questions