@pichenettes has used “Clouds 2” here, so while you’re right, I think it’s a fair name for it
OK, so let’s turn our speculative attention to Stages. Shots of the panel are all over the interwebs after it was momentarily leaked in an introductory video for Marbles.
So what do we have? From top to bottom, row-by-row:
- six rotary pots across, collectively labelled (green LED symbol, orange LED symbol) SHAPE / TIME (red LED symbol)
- six green/orange/red LEDs across
- six black push buttons across, with dotted lines linking each to:
- six vertical slider pots across, with LEDs in the sliders.
- between the third and fourth sliders, a vertical legend. At the top is a black circle with a loop in it. then a gap, then a groups of three LED symbols, the green one being an upwards slope, the orange being two stepped overlapping horizontal lines, and the red one a single horizontal line.
- six input jacks across collectively labelled (green LED symbol) TIME / LEVEL (orange LED symbol(red LED symbol)
- six input jacks across, collectively labelled GATE
- six output jacks across with associated LEDs, labeled 1 to 6 (in Sanskrit/Devanagri script)
Everything strictly lines up in six columns. And it’s called Stages, a segment generator. And we know that it permits 3^6 * 2^6 * 2^6 permutations. So, the 3^6 term clearly relates to the three colour states of the top LED for each segment. My guess is that next 2^6 term relates to the black symbol with the looping arrow in it. The papernoise design language is now that a black circle means “hold down the button” or “long-press the button”. I think the latter in this case - so a long press on the button for each segment toggles whether that segment loops of not. The final 2^6 term probably relates to an implicit switching behaviour depending on whether a jack is inserted or not in… well, we have three rows of jacks to choose from.
So, the green mode for each segment clearly indicates a smooth transition segment between a starting value and and ending value, with the shape of the transition function being adjusted from log through linear to exponential (or v-v) by the mini-pots at the top. The duration of the segment can be set by the slider and/or by the CV input in the row of jacks below the slider.
The red mode clearly indicates a sustain segment, with the top row mini-pots setting the sustain time, which suggests that the inputs labelled GATE may also be trigger inputs (activating on the rising edge of a gate input). The level of the sustain is set by the sliders and/or by voltage on the CV input jacks in the row below the slider.
The orange mode is a conundrum. The overlapping horizontal lines suggests that it is not a step or sequencer mode. The top row pots set shape in this mode, but shape of what? The sliders and the first row of jacks set the level in this mode, but level of what? The ability to set shape but not duration for this mode is what mystifies me. It’s probably bleedingly obvious…
Then we have a GATE input for each segment, and an output for each segment. Are the segment outputs automatically routed in a sequential fashion through to the first output that has a jack inserted in it? Thus you can have six independent single segment outputs, or three two segment outputs, or two three segment outs, and so on. That is not inconsistent with the final 2^6 term, above, because the cardinality of the powerset of 6 items is 2^6, isn’t it?
OK, and now we also have an Instagram clip as an extra clue: https://www.instagram.com/p/Bg9ETBoBu8L/?taken-by=mutableinstruments
Here are some screenshots from it:
This clearly shows one segment performing simultaneous amplitude modulation (multiplication) and (additive) offsetting of another contemporaneous (looping) segment (assuming that no external VCAs and voltage address are being used). Is that what the orange mode does? But how is the duration set in orange mode?
So what is the orange mode?
The panel text could offer some clues. Note SHAPE/TIME has the green/orange before and red after, whereas TIME/LEVEL has green before and orange/red after. So for the orange mode, the pot is a SHAPE control and the slider is a LEVEL control.
At a quick guess, green mode is a smooth transition to the next stage, orange is a stepped one. How does a SHAPE work for a step? Perhaps a number of steps (staircase-like)?
The sliders have LEDs in them, so some visual feedback could be used there.
Yes, I mentioned that, above:
But if the orange mode is stepped, then why does the icon for it on the panel show two overlapping horizontal lines? I don’t think it is stepped. And how can it have a shape but not a time parameter, unless the time parameter for it is fixed (unlikely)? Or is time for orange-mode segments derived from the preceding or succeeding segment? Or do orange-mode segments fire simultaneously with a green-mode segment that precedes it, and thus they act as multiplicative and/or additive modifiers (with a time-dependent shape)? That would permit the scope traces shown in the post above to be obtained directly from the module (if that is what is happening in the Instagram clip)…
Agree that the LEDs in the sliders might provide feedback on current voltage level, not just on whether that segment is active.
The individual gate inputs for each segment are also interesting. Presumably when a gate goes high for a segment, the envelope immediately jumps to that segment. And looping segments keep looping (and sustain segments stay at the sustain level) while the gate for that segment stays high? And while the gate for the first segment in the chain stays high? In other words, are the gates transitive?
I think you’re right that the orange mode is the “interpolated” mode, the source of the stepped or jagged motion.
COMPLETELY unclear to me how the gates and outputs would conspire to produce this scope trace.
Definitely seems like we’re somewhere between an envelope generator and a sequencer though.
Please don’t spread the rumours.
Is it some kind of combined envelope generator and VCA?
Maybe it can take different audio or CV signals, and act as a kind of sequential switch, with some kind of per-segment gain slope…
Or that could be complete nonsense…
I’ve taken part in some leak speculation on Muffs. Do I understand you right that you prefer that one doesn’t? I don’t mind that and would respect your opinion, just good to know.
We’ve been struck by quite a few leaks during my days at Propellerhead, which often sucks and can warp expectations, but it’s also been a lot of fun following what the users are speculating about and it’s driven interest so I’m torn.
I think Olivier was referring to a preview demo and talk about new modules he gave at Synthfest France in Nantes over the Easter weekend.
Yes, my post was relative to a new leak on the elektronaut forum which has now been “patched”.
Possibly, but I think it is more likely that segments can interact with each other, which means they must be able to fire simultaneously. I suspect the orange mode is “overlap” mode, in which that segments runs simultaneously with the preceding segment, and modifies it additivrly or multiplicatively or both. Maybe set via the level control.
Ah, OK, thanks for the info!
Well spotted! That’s definitely it, with the green LEDson tte sliders, somewhat out of focus in the background. But we can see jacks inserted in just a few of the gate inputs and a few of the outputs, and what appears to be it running in sequential steps in stages 1 to 4, while simultaneously running an independent loop on stage 5 (can’t see stage 6), which in slo-no can be seen to be semi-regular in period - maybe it is like an LFO but is being reset by the gate in?
Glimpses of modules aside, I like what he’s saying in that video. It gets at the heart of what’s been bothering me about the criticism of “consumerism” within the modular community lately. I am still finding my voice and figuring out my identity, musically and otherwise. To borrow The Straight Dope’s tagline, “it’s taking longer than we thought.”
All these individualist tendencies…when the Revolution comes, you’ll all be collectivised into State Modular Synth orchestras, and issued with standard sets of modules as determined by a committee comprising suitably reliable Academicians and Party officials.
On another matter, someone by the name of Starthief recently published a track somewhere or other, demoing some aspect of Plaits or Marbles, I don’t recall which. The track notes revealed that it was based on a six note sequence. Pretty sure said person has a prototype Stages, so I take that as confirmation that it can be used as a six-step sequencer, using the red (sustain) mode on each step. Or maybe the orange mode?
Or maybe the sequence came from elsewhere, done the hard way as part of an attempt to get a quantization/chromatic remapping script for Teletype straightened out, and should not be used as confirmation of anything other than a habit of playing patterns of different lengths against each other polymetrically.
It would have been simpler to program the notes into Marbles’ quantizer, since it was a sort of “010203” pattern.
Hmmm, I wonder if orange mode is, in fact, a clocked step mode. Green is sloped segment mode, with shape and duration settable - once triggered, it runs for the specified duration. Red is sustain mode, with level and duration settable (i.e. probably not a traditional sustain mode, which requires a high gate signal). Orange mode seems to have shape and level settings, but no obvious time setting. But if it is a clocked segment mode, then maybe it learns the duration of the step, using a tap-tempo algorithm? I can see how that might work if all segments are in orange mode, or at least all segments in a chain (where a chain is terminated by a jack in the output socket of the last step?). But unclear how it would work if there are mixed green, red and orange steps in a chain of steps.
yeah, solving the WHY of there being 6 gates and 6 outputs seems like the real puzzle of Stages.