Having seen some of the comments they’ve made to Kickstarter backers, I don’t think you’re being sensitive in anyway. I decided not to back the project (for practical reasons), and am now quite glad of that decision. (Of course I still hope it works out okay for you!)
1 Like
omg THANK YALL for this thread!
I’ve often thought similar. I have an old Faderfox LV2 that I absolutely love the form factor/options for, but hate the built in automapping that I’m always fighting against.
Would love a UC3/4 with a joystick - have honestly thought about trying to order a custom one tbh, but it seems insane to have to go that specailized.
thanks for all the resources about/glad to know im not alone in this ha!
3 Likes
It would almost be perfect if it had midi in/out connections rather than just USB midi. (It’s also something I miss on the Softstep btw) How do you find the build quality? At that price range it’s kinda tempting to go for it anyway.
1 Like
It’s held up to some abuse. I used it as a tap tempo controller while VJing at outdoor festivals, so it’s seen some things. I think it used to be more expensive, it’s definitely not built like a $100 device. Metal case, metal pedal, but plastic bottom plate and button caps. Caveat, I don’t know if they’ve changed the product at all in the N years since I bought mine, but the pictures look the same.
1 Like
I have to say I’m quite happy with Sensel Morph and Faderbank 16n. With those I cover pretty much everything I need. Important to note that I use TouchDesigner as MIDI effects hub to define ranges and smart interaction/automation for Ableton and lighting control.
Interesting thread. Please, does anybody know a USB powered MIDI controller with pots or faders that spits out a higher resolution of pot/fader positions than 128 values (via MSB/LSB)?
Fader fox mentioned above does 14 bit midi
1 Like
Oddly, the Behringer BCR2000 does as well. I’m surprised it’s not more common.
2 Likes
even though faderfox supports 14bit, the encoders themselves aren’t that high resolution. the midi fighter twister has a better resolution. then there’s always the arc of course : )
there’s an informative thread on modwiggler about it
2 Likes
@Rodrigo: Thanks, but the BCR2000 used encoders, whereas I am looking for either a fader- or potentiometer-based device.
@Smuff: Thanks, are you refering to the LV2? As far as I understand it, it was optimzied for Ableton Live, not as a generic controller, it is also out of production, and at least the manual does not mention MSB/LSB.
@2197: Thanks, but MIDI fighter twister is unfortunately also encoder-based, arc as well. Will look into the MW thread!
2 Likes
The recently released Intech Grid controllers (see also) are capable of reading up to 12 bits from their pots & faders (0-4095). I haven’t had much time to experiment with them yet since I just got mine on Friday; initial results are promising but noisy. The editor software is basically a semi-graphical interface for editing Lua code that runs on the devices, so it’s probably possible to program in some hysteresis…
1 Like
That’s a helpful thread.
I did see the JOUE thing on there, which looks kind of like a mini Soundplane, which is kind of intriguing but I have to say I really don’t like the idea of putting MIDI/MPE out behind a firmware paywall…
On another note, I got my address confirmation email so my OffGrid should hopefully arrive soon. The most recent update does have some promising comments about being able to configure/use it in a more open ended way:
Our goal is to serve the communities we can do most for, the early adopters, the curious kids, the beatmakers, the script ninjas, and the underserved alike.
3 Likes
Love the thread - and love me some nice, small, flexible MIDI controllers… where are they!!!
Some thoughts about resolution:
Folks often ask for more than the 7-bit, 128 value resolution that a MIDI CC offers. For pots and faders, this directly maps to physical distance, and for most purposes, this is close or as precise as you could hope given the devices and mechanics of hand control:
-
In a typical 100kΩ pot, with 270° of rotation - this means you get a resolution of about 2.1°, (and 0.8kΩ). I suspect in a small or medium knob this is probably a little less precise than you can control - but I bet not by much. (I searched for studies, but all I could find were focused on strength not positional precision). I also suspect the average linear taper pot isn’t all that much more precise.
-
For a medium travel fader - say 60mm, then you get a resolution of about ½mm - which seems more than one could expect for precise linear positioning over that distance. Even if that range represents 80dB of reduction - it is still a precision of 0.6dB… which for volume setting is pretty much at the limit of audible differences in almost all circumstances. (For audio taper mapping, you get more than 0.1dB accuracy in the top end.)
-
Other common controls are similar or worse: Joysticks are just two potentiometers with even less angular range, hence less precisely positionable. An XY touch pad, while larger in both dimension than the typical fader, generally is highly limited by the pitch of the sensing technology and usually far less precise than 7 bits.
So - the mechanics of the device generally preclude any useful precision beyond 7 bits (or close to it).
This isn’t to say that there aren’t circumstances where finer control isn’t desired (like frequency settings) or achievable. Larger knobs and longer faders can usually double the available precision - though that is really only one more bit! Encoders, on the other hand, by leveraging acceleration and multiple turns can, of course, often offer far more precision. The Arc, though really careful knob design clearly is an example.
14-bits, when needed, is more range than almost any musical parameter: It is almost exactly 1 cent of tuning if the control is mapped over the whole audible range. It is 1/100th of a dB when mapped over the range of human hearing - far more resolution than one can perceive. It covers a range from 1ms to 16s for timing operation at 1ms resolution. If exponentially mapped (which is more useful for longer time ranges) - easily covers from sub-ms to epic opera duration.
For most electronic music synthesis parameters, it is rarely useful or viable to have such extraordinary ranges (filter frequency and micro tuning being the two common exceptions) - and the available 7 bit range generally maps the usable range of a parameter with sufficient precision. For the exceptions, 14 bit is almost always enough… and the 24 bits of MIDI 2 buys you nothing in perceptual space - and it is unlikely you have a human input with anywhere near that resolution!
16 Likes
Sorry no I meant the UC4
For me a bit part of it comes down to mapping and resolution. So if you take a 7-bit encoder and have that mapped to a filter cutoff, you’ll often hear awful “stepping” since it’s a low amount of possible positions, which is then worsened by most encoders skipping values anyways.
To mitigate that you can add slew at the software level, which smooths things out, but adds latency to fast movements.
So for my mini fader thing I’ve gone with 14-bit MIDI even though the throw of the fader is pretty small (typical dj crossfader range (50mm?)), but it means the slew in my software is next to nothing.
For my purposes too, I also have audio feedback in the signal path, so having that many more places where the fader can come to rest can offer more dynamic feedback control (e.g. coming to rest at 12000 vs 12001).
There’s also the fact that usable/musical resolution is not necessarily the same as being able to manually and repeatably control a single increment of the parameter. That’s largely what makes modular/analog stuff appealing to people is that “infinite resolution” offered by analog parameters etc… It’d be crazy to think you could dial in the same cutoff (as per the example above) more than once.
7 Likes
Unfortunately, playing multiple oscillator and filter frequencies, including audio-range crossmodulation and feedback levels is at the very heart of what interests me. 
For these use cases, turning a pot on my analog synthesizer gives me an amount of precision I can not achieve with a comparable pot type on my MIDI controller that’s being quantized to a mere 128 steps. Smoothing the transitions between these steps while turning a digitized pot will make the steps inaudible, but it won’t solve the fundamental issue: the values in-between two steps are not accessible.
I do agree though that the full range of 14 bit values is not always needed. But that’s the beauty of the MSB/LSB concept as far as I understand it (please correct me if I am wrong): it can be scaled as needed. As an example (as far as I recall), the Elektron A4’s MIDI implementation uses MSB/LSB to increase the resolution to 256 values on certain parameters.
2 Likes
Of course you are both right - in that when used to control frequency, 7 bits is too coarse, unless the control is limited to a very small range. Many synths do use a 7 bit CC for control of center frequency of a filter - which works as long as all you’re doing is statically shaping a sound a bit.
You both have called out cases where one needs much more precision:
Turning oscillators in modulation requires very very accurate tuning - 7 bits is not enough. Further, I note that on many analog synths (looking at my Serge here…) there is a separate fine frequency knob, because you just can’t position the full-range frequency knob accurately enough - no matter that it is analog. (Tuning the LFOs and oscillators on the Recursive Machine, with those really tiny pots really becomes part of the performance I find - it takes quite some deep concentration to perform!)
Slewing frequency also is going to require very fine resolution if not to be heard. Even here, I’d be skeptical of 14 bit resolution - I’m pretty sure you’d need some slewing if you want it to be “analog smooth”.
3 Likes
I believe it is not uncommon for synths to internally (when scanning knobs) to use resolution higher than 7 bit; e.g 10 or 12 bits (DAC & ADC) when indirectly controlling parameters and in preset storage. I believe they also often slew the output of the DACs meaning that even if the values are stepped, the stepping isn’t as clearly audible. The result is that you can’t very accurately fine tune the frequency to a specific value, but you can adjust them on the fly without too many strange artifacts.
1 Like
The thing I miss most when working with controllers that send discrete, quantized values isn’t precision, but imprecision – the ability to dial something in, depart from it, and then return somewhere close to where I was before, but not identical. To never get stuck repeating myself, but to always have somewhere new to go.
1 Like
Not sure if this relevant to anyone, but I recently got the Noise Machine Mini MIDI Controller. It’s a tiny BLE MIDI device with 4x3 keypad, slider, and knob.
Here’s more in their own words:
NMSVE works by wirelessly connecting to software on your smartphone, laptop, or PC/MAC via bluetooth. Consisting of 12 buttons, a knob, and a slider, NMSVE allows you to choose from 12 different midi channels and offers 3 different modes of use… All in a case that’s smaller than your wallet!
Although, embarrassingly I haven’t had a chance to take it out of the box and use it yet 
TMI Update
Spurred by this thread I was able to give the device a test. Connecting to Ableton over Midi BLE (MacOS) was a breeze. My biggest complaint (that I knew when I bought it) was the lack of USB Midi or DIN Midi. Although the source code is open source it’s a ESP32 based board and hacking USB Midi is non trivial, but DIN Midi is maybe probably possible given the exposed pins.
1 Like