Polyphonic earthsea for trilogy/ansible and er-301/just friends/txo


Yeah, range setting wouldn’t very useful for anything more than (certain) half steps. I do like idea of selecting several values though!

and i have not hooked up another jf, but might get around to it tonight. my guess is that they would act the same as having two w/'s on the same bus, which currently means they both respond the same way at the same time. i’ve always thought it would be super useful if multiple jf’s could be controlled individually

anyways, thank you for putting so much work into this ecosystem. its pretty wild to see how this is developing


PolyES is really amazing… I’ve been trying it out on my trilogy ES and it’s a real eye-opener even with just a couple simple patterns going through the various outputs.

I use the MIDI/CV functionality of trilogy ES and Ansible quite a bit; would it be hard to put that back in? (When plugging a MIDI keyboard into trilogy ES just to double-check the module locked up and required a restart before I could use a grid again…)


if this is going to be the case i guess you could still take advantage of having two by mixing individual outputs which will give you 2 different sound sets. but yeah it’d be cool being able to address multiple just friends, and i’m hoping @Galapagoose will consider exposing more parameters over i2c so they could also be controlled from polyes - being able to control and modulate telexo parameters directly from earthsea makes it really expressive.

polyes is a complete rewrite of earthsea so it doesn’t have the MIDI implementation. i do want to add MIDI support to multipass at some point though, so if it makes it easy to add it to polyes i’ll consider it. i can certainly see a benefit of being able to use something like nanokontrol with mod buses, for instance.


I’ve not been running scripts on the TT yet, just some one liners on it while the Ansible with Polyes plays a sequence. With the TT untaxed I’ve not had a crash and at the moment because Polyes is such early beta I didn’t feel like testing where the breaking point is yet. So in preliminary soft touch testing the two are tolerating each other.


This is great! Got a Txo+, JF and the Ansible outs working. (W/ and TT on the same busboard)

Similar to @unity2k Im running a few lines on the TT occasionally to alter Txo+ LFO behaviour while JF is being controlled by poly earthsea. No crashes, its been on for around 4 hours.

@scanner_darkly How would I go about recording a pattern with a set of voices (lets say 5 JF voices) and then playing live over the top with a 6th voice sent to ansible? I breifly played with Ansible earthsea and know there was live/pattern voice allocation but thats different now. Trying to get my head around all these features! I can dedicate voices 1-5 to JF and 6 to Ansible, but if i play live while less than 5 recorded notes are playing, the live note will be routed to JF


unlike ansible earthsea in this version live always shares voices with the current pattern. what you play is what will be recorded and routed to assigned outputs.

so what you need to do is use another pattern for playing live. leave voices 1-5 enabled in pattern 1, disable 6-8, record, switch to pattern 2 and enable voice 6. now you can play live with this pattern or record and switch to pattern 3 (at this point you have both 1&2 playing).


Apologies for noob questions but do I understand correctly that the Teletype is needed to be present (i.e master )

I have er-301/JF/ansible and grids and would love to access polyphony if at all possible



Not needed. But i2c powered bus board/expander


Maybe you remember, you adjusted some slight gate timing for me a few months back so i could play my modules without a lag/trigger issue i had.
I just wanted to quickly inform myself if this adjustment made it into this new release or will i face the issues again that i had prior your adjusted firmware?


Curious to hear more about what is working for you in this regard? I initially balked at the idea of providing i2c control over the synthesis parameters as they seemed ideally suited to CV control, while triggers & pitch (per channel) are not possible with patch cables alone. At some point it just becomes a softsynth which has never really been a goal for a hardware device.

I’m interested in any propositions you have though, and would be happy to consider whether they’re interesting / possible to me, and to consider if it’s something useful to more than a handful of people (hence worth the time investment).


Let me know if you determine the problem is with the board. I tested per instructions (~2.8V across the pins IIRC) but since I was never able to trigger the situation that would require pull-up, I couldn’t test that the boards rectified that problem.

Please contact me if you determine the problem is with the board, and we’ll figure out how to make things right. Same goes for anyone who bought TT bus boards from me.


there shouldn’t be a lag.


i doubt it would be the board. i’m also testing another theory (i sent @n-So a test version to try).

if anybody is able to use @xenus_dad’s board successfully could you post here so we can rule it out for sure?


here is what i had in mind:

controlling parameters that are also controllable via CV/knobs: less of a need here as they can be controlled directly on the module (and this would work better from a performance point of view since you can play polyes with one hand and control jf parameters with another).

one use case that is not covered by this is using the new built in modulation in polyes. one example would be using key tracking to modulate a parameter (say, making higher notes have shorter decay). this is already possible by assigning a voice key value to a CV output and using that CV for modulation, but being able to do this over i2c would provide this ability to something like meadowphysics which doesn’t have CV outputs. this would likely require a separate mode (or some other way to deal with having a parameter that can be controlled both over i2c and with a knob).

same as above but individually per voice. not sure if this is possible at all but it could provide for some amazing expressiveness if you could set time/intone/ramp/curve per voice!

controlling parameters that are not controllable via CV/knobs:

are there any other internal parameters that could be exposed via i2c? this would be similar to providing individual voice control which is not available with the module itself.

right now you can modulate octave/transposition/volume per each jf voice, but such changes only get applied to the next note. would it be possible to have a way to change pitch/volume without triggering a new note? so basically like JF.VOX but without retriggering the envelope.

the above would allow for an approach where you could control multiple voice parameters (each with its own defined range of changes) from a handful of sources, which could be live gestures, for instance. or recording that and playing prerecorded modulations over your sequence, or using the sequence itself as a modulation source.

i think many of these would be useful with teletype as well!


I would say the problem would be per-board, if anything. The PCBs were @tehn’s , sourced from Oshpark, and the components were per the shared BOM. If anything, it would be builder error rather than the whole batch being off.

I also think it’s unlikely to be the board, if the boards work as designed…it’s a pretty simple build. Still, I am not a professional builder, and don’t want anyone to be stuck with something not working that I’m responsible for.


So I’ve been reading up on i2c while I wait for parts to build a busboard but I think I may have hit a hurdle. It says that the i2c cables should be short. I have a 104 12u case and I’d like to connect my ansible and 301 which are in the bottom two rows of my case and a just friends which is at the top row of my case. Will this be ok? Does anyone think otherwise?


I have it setup this way and it is ok! /w top right, ansible top left teletype middle bottom. 7u 104hp.


Great news! Thanks for that. Guess I’ll just give it a go when I have it all up and running.


Just wanted to comment on the suggestion of a Teletype bus board to provide pull up resistance:

AFAIK the boards shown at that link are Teletype boards, designed for the spacing of the Teletype power and i2c connections. I am not sure they would function as needed in a non-Teletype scenario. You could probably connect power and i2c cables to it and let it float in your case, but that leaves a lot of exposed pins untethered and may be a poor plan.

Just commenting here because people have been reaching out to me about buying Teletype bus boards to support Polyearthsea in setups that do not include Teletype.


i did test with the teletype i2c backpack not attached to a teletype (so, just the power and i2c cables connected), and it does work. exposed pins would be a problem for the other board as well - good to call it out though, i added a warning to the page!