I’m curious as to what the internal knob resolution on the front panel is? i understand there is selectable 7-bit/14-bit CC resolution with midi, but i’m hoping the internal resolution (of let’s say the filter cutoff knob) is higher?
The control voltage CV is 100% analog, no steps, but the “recording” of a parameter is digital. So also the recall is a quantized value, until you turn the related knob. That’s my understanding…
that is what i was hoping but couldn’t find any information confirming this in the manual, voyager specs etc.. you’d think they would advertise that if it were the case which kind of got me worried it wouldn’t be so?
14-bit digital, as far as I know. You might be getting confused with the pitch bend resolution which is 7/14-bit switchable.
The LP introduced “RAC” as a feature, so my pick is that the Voyager uses only digital scanning, with 16384 values per pot.
Basic Editing of sounds
Sound editing of continuous parameters (like Filter Cutoff) is done from the front panel by enabling a parameter with its switch (its LED becomes illuminated), and then adjusting its value control. There are 4 value controls, one for each section: Modulation, Oscillators, Filter, and Envelopes. Each is surrounded by a ring of 15 LEDs that show approximately the stored or edited value of the current parameter. The value controls are analog, and when a parameter is activated, the analog control signal is switched to directly control that parameter (RAC or Real Analog Control). For each section, only one continuous parameter can be activated at a time for editing. For parameters that have multiple possibilities (such as Mod Source) pressing that switch advances through the possible Mod sources. For On/Off type switches like Osc. Sync, the LEDs are On when that parameter is on or Off when the parameter is off.
^ if so kind of dissapointing (that the low-end phatty has better internal resolution then the flagship voyager) also according to this thread (http://forum.moogmusic.com/viewtopic.php?f=2&t=13737) the voyager uses 14-bit CC and not 14-bit NRPN so the values are actually only 254 (double 127) and not 16384 ?? i hope i’m wrong but hopefully someone from moog can confirm..
7-bit are 128 values, 8-bit are 256 values.
14-bit are 16384 values.
What is the resolution of a Voyager pot anyway?
270 degrees/16384 values = 0.016479492 degrees per value.
Arc length formula gives 2pi r(theta/360)
= 2pi 0.406(0.016/360)
= 0.000116774 inch of travel per value
= 0.0029660596 mm per value
This is for a small pot (0.812 diameter), I couldn’t find the size of the large ones.
You will have a difficult time differentiating between two values.
The Voyager uses two CCs per parameter to generate MSB and LSB for full 14-bit resolution over MIDI.
The Voyager MIDI output on the ones I have here is only about 256 steps per knob. This is a “bit” confusing because the internal resolution is much higher. An easy way to confirm this is to put a MIDI monitor on the output and observe. Roughly you can only get two different values on the LSB before the next increment on the MSB. So aprox. 256 steps.
The confusing part is if you use a MIDI recorder and turn local off, and loop the MIDI back into the Voyager. Because of the lower resolution of the MIDI output, the Voyager will react differently than under it’s own internal control. I noticed this as I used a poly Voyager setup for a few years, and the poly note assignment required the MIDI to be looped back into the Voyager.
The Voyager is what it is, a digital controlled analog synthesizer. There are all kinds of trade offs that had to be made to give the Voyager patch memory. It is easy to hear the digital artifacts but that’s the trade off. MIDI and digital scanning of the knobs is only so fast. Try turning more than one knob at a time and you begin to hear the stepping as it tries to scan both knobs at once. To keep from getting ghost readings from all the knobs at once, there seems to be some type of digital focus. Once a knob looses focus, the next time you turn it it will jump a little to the new value. But once the knob is recognized, the resolution is very high and much higher that the MIDI it is sending out. So the Voyager will respond to those high resolution MIDI messages. It will save patches at that same high resolution the dials have internally. But can only send out a lower resolution MIDI. At least on the ones I have here? Put a MIDI monitor on it and see what you get.
Correction: MIDI 7-bit CC messages are 128 steps, not 256
Right, this is the confusing part. The Voyager sends out two 7-bit messages per step. For filter cutoff it sends CC19 as the MSB and CC51 as the LSB. Not sure if you would call this 14-bit or just two 7-bit messages. On the MIDI resolution sent though the MIDI out, it is tough to get more than two values per LSB, before the next increment on the MSB. So aprox. 256 steps sending both CC19 and CC51.
Here is the MIDI readout for the filter cutoff. Try it!
B0 13 4E Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 78
B0 33 34 Controller Channel: 1, Controller: 51 (), Val: 52
B0 13 4E Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 78
B0 33 78 Controller Channel: 1, Controller: 51 (), Val: 120
B0 13 4F Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 79
B0 33 3C Controller Channel: 1, Controller: 51 (), Val: 60
B0 13 50 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 80
B0 33 00 Controller Channel: 1, Controller: 51 (), Val: 0
B0 13 50 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 80
B0 33 44 Controller Channel: 1, Controller: 51 (), Val: 68
B0 13 51 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 81
B0 33 08 Controller Channel: 1, Controller: 51 (), Val: 8
B0 13 51 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 81
B0 33 4C Controller Channel: 1, Controller: 51 (), Val: 76
B0 13 52 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 82
B0 33 10 Controller Channel: 1, Controller: 51 (), Val: 16
B0 13 52 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 82
B0 33 54 Controller Channel: 1, Controller: 51 (), Val: 84
B0 13 53 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 83
B0 33 18 Controller Channel: 1, Controller: 51 (), Val: 24
B0 13 53 Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 83
B0 33 60 Controller Channel: 1, Controller: 51 (), Val: 96
In theory the combination of the two CC’s should give me 16384 different values. But in practice I can only get aprox. 256 steps from the MIDI output. If you are getting more steps than I am then they made a change to the MIDI processor somewhere during the Voyager run and I have an early one. The internal resolution is much higher than the MIDI output.
thanks for the info guys (although some of it is flying over my head) so to sum things up, the midi knob resolution is 2 x 7-bit CC’s but what about the internal resolution, still no exact numbers.. is it really higher (certainly feels so)? i mostly sequence notes and do param tweaking live so could benefit from higher internal knob resolution..
Recombining two 7 bits MSB and LSB message into one single decimal value involves some calculation.
For example, in your table CZ Rider, if we take the last values for CC19 MSB = 83 and CC51 LSB = 96, performing the calculation: ((MSB * 256)/2)+LSB we have therefore ((83*256)/2)+96 = 10720 resulting decimal value.
Internally the Voyager works with 14 bits resolution for the front panel knobs (two 7 bits NRPN values), but it averages those values down to 0-255 values, for display purposes on the LCD screen.
And, if I’m not mistaken, for the Mod Wheels you can specify in software what resolution is sent via MIDI out, either normal 7 bit CC values (0-127), or NRPN dual 7 bits messages (0-16383).
Sorry for the poor explination. That table was the actual MIDI output recorded with a MIDI monitor. I was turning the filter cutoff as slowly as possible to record how fine the resolution is on the MIDI output. Perhaps someone could do the same test. I can get no more than 2 increments of LSB to every increment of MSB. I should be able to get 128 (0-127) increments from the LSB in theory. So I’m getting no more than 256 steps end to end with this example using the filter cutoff. This is limited just to the MIDI output the Voyager is sending. The internal resolution is much higher, but no way to measure this. And it can receive and reacts to these fine 14 bit MIDI messages sent from another controller no problem. Just can’t seem to send them and is only an issue if you are recording MIDI CC’s from the Voyager, or sending these same CC’s back to the Voyager.
Another example of the actual MIDI output from the filter cutoff, turned slowly counter-clockwise from the highest setting.(CC19-127MSB:CC51-52LSB):
B0 13 7F Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 127
B0 33 34 Controller Channel: 1, Controller: 51 (), Val: 52
B0 13 7E Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 126
B0 33 74 Controller Channel: 1, Controller: 51 (), Val: 116
B0 13 7E Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 126
B0 33 34 Controller Channel: 1, Controller: 51 (), Val: 52
B0 13 7D Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 125
B0 33 70 Controller Channel: 1, Controller: 51 (), Val: 112
B0 13 7D Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 125
B0 33 30 Controller Channel: 1, Controller: 51 (), Val: 48
B0 13 7C Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 124
B0 33 70 Controller Channel: 1, Controller: 51 (), Val: 112
B0 13 7C Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 124
B0 33 30 Controller Channel: 1, Controller: 51 (), Val: 48
B0 13 7B Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 123
B0 33 70 Controller Channel: 1, Controller: 51 (), Val: 112
B0 13 7B Controller Channel: 1, Controller: 19 (GP Controller 4), Val: 123
B0 33 30 Controller Channel: 1, Controller: 51 (), Val: 48
I’m not seeing 128 steps of CC51 between every CC19. Just 2.
Sorry, I misunderstood…
Then, perhaps the Voyager is set to minimize the amount of MIDI messages sent, in order to avoid overflowing the receiving devices ? (just speculating here)
Perhaps the values are only sent out each time the values for the LCD display are updated (averaged to 0-255) ? (again speculating)
Maybe Rudi Linhard could shed some light on this ?
Hi folks,
the Voyager’s DA converter has 16 bits. The AD converter for the pots has 12 bits.
BTW you can avoid the steps when you’re turning two pots at the same time when you switch off the Parameter Display in the PANEL menu.
Have fun,
Rudi
Thanks for a quick reply Rudi ! 
Hmmm, I’ve always thought that the pots had 14 bits resolution. Interesting.
But is the Voyager supposed to send out every single value change of the LSB via MIDI out when turning only one knob slowly ?
Not at the moment, but I’m working on a good solution… 
Thanks again Rudi. Much appreciated !
There you have It CZ Rider, the answer directly from the man himself.
12 bit resolution should give you 4096 steps.
The use of two CCs spaced 32 apart is how MIDI supports 14 bit resolution (except for pitch bend). Controllers 0-31 are the MSB, 32 to 63 are the LSB. So looking at controllers 19 and 51, these are in fact linked together.
Yes, that is part of the Non Registered Parameter Number (NRPN) scheme of the MIDI implementation, where any of the first 32 controls can be optionally paired with a control offset 32 higher. This is the rare 14-bit Continuous Controller feature of the MIDI specification, and NRPNs simply take advantage of that existing option to offer 16,384 possible values instead of only 128.
But the problem that CZ Rider has observed is that the Voyager doesn’t send ALL the data via MIDI out, even though internally it works at this high a resolution with all the 4096 steps.
And, as Rudi Linhard pointed out, it is a very small bug in the software of the Voyager that he his working on.