“local nadirs” and “anti-melodic hypersurface” is gonna give my poor mathematical heart hypertension
Signifiance et interprétose sont les deux maladies de la terre, le couple du despote et du prêtre
It’s a quote, from a book of exchanges between Gilles Deleuze and the journalist Claire Parnet, thus immune from being capybara’d. Although is Deleuze himself quoting someone else? The phrase seems familiar, somehow.
I’m not sure whether “ signifiance” best translates as significance or signification, though. The point that Olivier is trying to make, I think, is that, as Deleuze himself would say if he were to paraphrase Pratchett: it’s difference all the way down.
Anyway, this is a sort of reverse archeology, in which the nature and purpose of obscure artefacts must be inferred from glimpses and fragments.
From Marbles (emphasis mine):
Second, every technical evolution Plaits and Stages received were also “backported” to Marbles for sake of uniformity.
Did we already know “Stages”, or is @pichenettes teasing us?
We saw various leaks of it but no official info yet AFAIK.
Well, we got a few things right! And yep, it used probability density functions (or probability mass functions).
Lots of other clever things, including the scale note histogram learning from an incoming pitch CV stream, the note masking on a knob, and internal symbolic musicological manipulation of note pitch values.
But the most interesting thing is this implementation detail, revealed by Olivier on MW:
Another thing: the module detects some common self-patches; and in this case does the signal routing internally (in software, bypassing the hardware output and input). This eliminates some latency or inaccuracies.
This is perhaps the first Eurorack example of the patch-cord-jack-routing-internal-digital-signals paradigm, which I am pretty sure was pioneered by the Mungo State Zero polyphonic patch-programmable semi-modular: http://mungo.com.au/zero.html
Preorder sold out everywhere, here in the UK
Scales are programmed by playing a short jam in the target scale: Marbles learns which notes are more prominent than others.
Edit: Ah, that’s what you meant by “including the scale note histogram learning from an incoming pitch CV stream”, didn’t connect the two.
Don’t wanna distract too much from Marbles, but if Stages does all the things I think it might, the interface design must have been quite the puzzle, eh @Papernoise?
It was done very quickly.
My initial proposal:
Oh Hannes, I so love it when you can squeeze more outputs than I had planned! From that point the “face” was set!
The major change was to remove CV on loop length, in order to get 7 outputs instead.
For a while the external CV processing mode was enabled by setting SPREAD to 0 and by feeding your CV to the BIAS input. But this eliminated the ability of transposition up/down (under quantization) an external sequence, so a button was added to enable this mode.
Hannes is a good designer. Impressive how he takes that already very good first version and then improves on it by grouping related things spatially and making them “float” together.
I did mean Stages and not Marbles, but I’ll satisfy myself with waiting till it comes out if I have to
Oh yes many things it does! 3^6 * 2^6 * 2^6 to be precise. It’s superior to Peaks on that matter.
Remind me what the HPs are for the three remaining modules? Was it 14, 18, and 14?
According to Olivier on Facebook:
2018 Modules: 12, 14, 18
14: Clouds 2.0
TBD Modules: 14, 14, 18
I hope the next one will be Stages
I thought I’d seen him list 5 modules due this year on Facebook. I’ve got 46HP left - so I’m hoping I was right and we’re expecting 14, 18 and 14.
So you think there’s another big beasty module in 18HP on its way. Somehow we know 0 about this one. What it does, what it’s called… Where do I preorder?
Based on Facebook and message board posts, my assumption is that the TBD modules included above could appear any time in the next year or so. Those are the “14, 14, 18 HP” sized modules.
The only other “release date” that I have seen is for Clouds 2.0, which is October at the earliest (according to Olivier on Facebook). There are people on this board that are far more knowing than I am though.
Since we do split hairs on this thread, I suspect the new granular module will not exactly be a Clouds 2.0, likely a reimagining/re-implementation which directly addresses the perceived design “flaws” detailed by Oliver in the post about the discontinuation.
As there seem to be some very knowledgeable computer scientist/engineering types on this lovely forum, I suspect there may be clues in the literature about granular synthesis as well…