Analogue Interfaces to Digital / Hybrid Devices
February 14, 2016 12:16 PM   Subscribe

The KORG minilogue is a really sweet modular, analogue synthesiser. It must, I think, have a hybrid (digital / analogue) architecture, and that raises a bunch of UI / UX questions...

  • First, am I correct in thinking that the analogue (knobs and switches) interface presented to the user is indirect? Rather than controlling the analogue modules directly, it must send signals to a digital computer, which then adjusts the analogue circuits appropriately. If not, how do things like presets and polyphony work?
  • Second, what happens to the analogue controls when the presets change? Do the knobs and switches move to the new positions? Surely not. So how does the UI design avoid confusing the user?
  • Third, is there a good book or article that discusses issues like this in UI / UX design?
H/T
posted by andrewcooke to Technology (9 answers total) 3 users marked this as a favorite
 
The KORG minilogue is a really sweet modular, analogue synthesiser.

It is not modular. Modular synths are made up of separate modules.

No the knobs do not turn when you change presets. When you move a knob with your hand though, the value you hear for that parameter will jump to be the right one for where the knob is now pointing.
Analog poly synths have existed with this kind of UI ever since the SCI Prophet 5 in 1978, so the behavior is not surprising to musicians.
The novel thing with the Minilogue is that they have managed to hit such a low price point.
posted by w0mbat at 12:45 PM on February 14, 2016 [1 favorite]


First: yeah, there's probably a microcontroller and some flash memory doing the digital lifting and patch/automation storage. Second: in audioland/synth control land this is called "takeover" and different synths/software implement it differently or give the user control of how they want the UI to work -- it can either have the value jump as soon as a knob is touched (which is musically problematic) or have the value begin to change when the knob reaches the value of the new preset. Can't really speak to your third thing, synth UI has a really well developed language.
posted by drapatz at 12:46 PM on February 14, 2016 [1 favorite]


Response by poster: sorry, you're right on the modular part.

i will search for "takeover". looking at videos i get the impression some switches have lights that indicate the "real" setting. also, i guess another approach might be to have the knobs fastened to an encoder rather than a potentiometer and so be able to spin (so without a marker they have no "fixed" position).
posted by andrewcooke at 12:50 PM on February 14, 2016


First, am I correct in thinking that the analogue (knobs and switches) interface presented to the user is indirect? Rather than controlling the analogue modules directly, it must send signals to a digital computer, which then adjusts the analogue circuits appropriately.

This seems to vary from synth to synth. Some use the knobs as direct control to avoid "stepping" but save the position to the computer.(and then do what i describe below when you load the patch) Some do exactly what you describe. Some do what you describe, but use a very high discrete number of steps that is borderline imperceptible.(this was not true of older digital/analog hybrid synths from the 80s)

Second, what happens to the analogue controls when the presets change? Do the knobs and switches move to the new positions? Surely not. So how does the UI design avoid confusing the user?

On every synth i own, even the very old ones from the very beginning of patch libraries, the knobs don't move. They're just "disengaged". Bumping a knob/slider will reengage(takeover) it and abruptly knock that parameter to whatever position that control is in. This can be very annoying if your knobs/sliders are not in perfect condition and are a bit flaky and the sound abruptly changes.

The only synths i've seen with motorized knobs/faders were either extremely high end at their time of introduction(and OLD) or newer non-analog combo units like the MC-909. It's a very unusual feature.

If not, how do things like presets and polyphony work?

It could be anything from internal sampling, truly having 2 VCOs per voice(it's only 4 voice, that's not all that ridiculous especially since these aren't discrete oscillators/filters but just synth-on-chips, or at least it appears to be that from the size of the unit and board shot), or switching in to some kind of divide down mode. Personally i think it's either #2, or a mix of #1 and #2. If 1+2 was done before filtering or in interaction with it, it would be REALLY easy to cheat that without people ever being able to notice unless they set up something to specifically test it. It's a lot better than synths that retrigger the filter in situations like that, judging from the clips of this unit i've heard online.

The more i listen to this, the more confused i am about whether it's really poly or just cheating. "Voice Mode lets you flexibly configure the four voices" is evidence it might really be, but there might also be some clever trickery at work here. It could just be rapidly switching parameters. If it truly is an 8 oscillator/4 voice rig for $500 though that is groundbreaking and korg did something really ridiculous here. That's basically one of these which is the only real competition in this space for half the money. It's a huge market disruption either way when it's only $100 more than say, a minibrute... But if it's just straight up what they're representing it as then... what.



On point #3 i can't help you. There is likely good stuff out there though.

On preview

also, i guess another approach might be to have the knobs fastened to an encoder rather than a potentiometer and so be able to spin (so without a marker they have no "fixed" position).

Essentially all modern synths work this way. Even some relatively old ones. It's unusual for the knobs to have limits now, only sliders do. Very many older synths with patch storage/takeover used just sliders for some reason and this may have been part of it.

As a side note, holy shit. How had i not heard of this thing? Even if it is cheating a little bit it's so exciting! This is the synth i've been wanting a major manufacturer to make for years. And holy crap that's so cheap! Not only that, but it actually sounds good. It has a very DSI prophet 08(which actually used DCOs)/later era analog korg "harsh" sound. I expect to immediately start seeing big names using this live, and hearing this sort of sound on big albums. James Blake was using the prophet 08 on overgrown(example) and played it much more wide open and harsh live... And it sounded like this thing. This seems like it would sound like a MONSTER behind a quality chorus/delay/reverb.
posted by emptythought at 12:54 PM on February 14, 2016 [2 favorites]


I fired that off just too fast, but on further searching korg is stating to retailers/reviewers(example) that this unit is truly polyphonic and has two oscillators per voice. WOW.
posted by emptythought at 12:55 PM on February 14, 2016 [1 favorite]


This is very simplistic version of how analog synths originally became programmable, but it might make things a bit clearer.

A non-programmable analog synth has a circuit board with various potentiometers (or "pots") on it in various places in the circuit. A potentiometer is a variable resistor, and as you turn the attached knob on the front panel the pot's resistance changes, which changes the behavior of the circuit. For example it might change the cutoff frequency of the filter.

To make that parameter programmable, you have to make things a bit more more complicated. The potentiometer now just attenuates a control voltage that goes to an ADC to a little computer. The little computer can in return output a digital signal to a DAC, whose output goes to a transistor that is now sitting in the circuit where the original pot used to be. Now the computer can see the knob position, and can also set the gain of the new transistor to the matching setting.

You do that to all the pots. Now when you change presets, the little computer can tell all its DACs to output the right voltages to set all the parameters at once. It can see when a knob moves and can then adjust the setting for that one thing.
On a polysynth you would have a little computer per voice card, and a central computer that feeds them all the same numbers to set them all to the same settings at once.

Anyway the end result is that you have a fully analog audio path, but the configuration of it is digitally controlled and can be changed in an instant.
posted by w0mbat at 2:39 PM on February 14, 2016 [1 favorite]


Of course the evolution of that is that you end up with a pure "controller" circuit board attached to the front panel, which has nothing but knobs and and switches that just send configuration data to the CPU.
Then you get one or more "voice boards" which receive all their settings in digital form, along with note start and end messages, pitch bend values, etc.
On the Minilogue all the "voice boards" are in fact just different sections of a single PCB. There's a teardown on YouTube.
posted by w0mbat at 3:11 PM on February 14, 2016 [1 favorite]


Some analog synths with digital controls and patch storage feature a "manual mode." My Juno 106 , first manufactured in 1984, does this. When you call up a preset, the sound doesn't necessarily match the positions of the physical controls. When you move a given control, that parameter will change to match the physical position of the control you moved, but all the rest will stay in the position programmed into the preset (and the preset LED will change from e.g. 75 to 7.5. to indicate as much).

But when you press the button to enter manual mode, nothing physically moves, but now what you see is the same as what you hear.

A modern synth like the Bass Station II, however, works similarly in that it's analog and has patch storage, but a manual mode isn't possible because not every control is accessible simultaneously from the front panel -- only oscillator 1 or oscillator 2 can be edited at one time, for instance, and you flip a switch to determine which one. When you call up a preset and then change a parameter, an arrow next to the preset display indicates whether the saved value for that parameter is less than or greater than its current value.
posted by ludwig_van at 10:27 PM on February 14, 2016 [1 favorite]


Just FYI;

Usually when people speak of a digital/analog hybrid they would be referring to something else entirely, like digital oscillators with analog filters. It’s really the audio path that people are talking about when they say "digital" or "analog". You’re talking about digital control, which is not part of the audio path, and has been very common since the 70’s (?). Calling this a hybrid would be confusing to a lot of people.

I’m not an expert, but I think the gist of that is right.
posted by bongo_x at 10:51 PM on February 14, 2016 [3 favorites]


« Older Best route to get cat halfway across the world?   |   Valentine's Day Special! Newer »
This thread is closed to new comments.