ER-301 Sound Computer


I’m drooling over this thing after watching the videos. Can’t wait to explore it and watch it grow!!


Code base? I thought it was closed source? Please tell me it’s closed source, that’s why I haven’t bought one.


This module really excites me - but I don’t think it’s there yet for the price. (£850 to get it to me.) I look forward to seeing how it develops.


its built on

By itself, Luasynth doesn’t do very much. It only comes with a basic demo app for generating or processing audio offline, via the command line. To use that, install LuaJIT. You’ll probably also want SoX to convert into and out of the raw, 32-bit, floating-point audio format Luasynth uses.

Here’s an example:

sox input.wav -tf32 -c2 -r44100 -q -
| ./luasynth delay -len 83 -feedback 90
| ./luasynth amp -gain -6
| sox -tf32 -c2 -r44100 -q - -b16 delayed.wav

This adds a delay to input.wav, reduces the volume by 6 dB, and writes the output to delayed.wav. The -tf32 -c2 -r44100 options to SoX tell it to handle the format Luasynth uses, while -b16 converts the output file back to 16-bit, the resolution of CD-quality audio.

In this example, we’re using luasynth twice. The first usage

./luasynth delay -len 83 -feedback 90

creates a delay unit with the len knob (length of time to delay) set to 83 milliseconds and feedback (amount of feedback) set to 90%. The second usage

./luasynth amp -gain -6

creates an amp unit and sets the gain knob, which is in decibels, to -6. For a full list of available units, run ./luasynth help units, and for help on an individual one, run ./luasynth help .

The sample rate is assumed to be 44100 Hz by default, but can be changed by setting the environment variable RATE. For example, to work at 48KHz, type export RATE=48000 in the shell followed by similar commands to those above.


Ah, but the ER-301 is closed source though, right?


yeah i think it still is but i also vaguely remembering him saying hed be opening it up for community contributions at some point…but im certainly not 100% on that


I have a ciat-lonbarde instruments and a shnth (which is what makes me consider this purchase in the first place) no audio euro rack stuff
this is a tempting bit of gear

I’m considering making a tiny skiff to incorporate this and a power supply to get used with the CL stuff. Does anyone have recommendations for a module or two that would go well with the 301 in a small system?

this looks to scratch a very similar nesting programming itch the shnth does while doing it with a different interface that seems more approachable. I can get ideas together quickly at this point on the shnth but I think this interface would allow for fast programming.

has anyone who owns it come close to using all the CPU in a single patch if so does anything change as you approach high CPU usage or does it just kind of stop working once you hit 100%?

with the shnth it starts to sound like it has taken some cough syrup (I believe this is how peter put it) and get kind of slow and weird


ok I’m going back through this thread and it seems like you don’t want to catch the CPU limit

ansible and an arc may be something that I look into in the future to pair with this.

I currently have a standalone meng qi Voltage memory that will work great for now to do basic sequencing & key playing

(I had an idea for the voltage memory where the top 6 pads worked as normal and the bottom 6 worked to update/trigger a sample and hold for the 6 individual outputs above
1–2--3–4—5----6 regular keys
7–8--9-10–11–12 update/sample&hold trigger
allowing for a different play style than is available now)

karplus-strong tends to work better with filters did I miss that the 301 has filters?

the period o meter and envelope followers look pretty interesting I hope to see some more use from them


Ansible/arc pair fantastically with ER-301 as does teletype.

There is a ladder filter in the ER-301.

Voltage memory sounds sweet. I’m curious about that.


if you have a looper and you set the buffer to say 15 seconds what happens if you record long than 15 seconds?


never mind I saw how it works in a video
it makes sense


never having had eurorack audio stuff does the 301 need some kind of output module to listen to the 4 outputs? or are they ok to go into a mixer or line level stuff?
I’m taking the plunge…


They should be ok going into a mixer, but the levels will be hot. If your mixer has lots of headroom or gain reduction/pads (some Mackie’s have that I believe) it will be totally fine.


hey does anyone here that has the 301 module have the IFM bar module form peter blasser?
I was wondering if they have tried the two together

with the new slice shift param in the sampler player module I bet it could be a lot of fun!


I’m eagerly waiting for Arches, from Soundmachines. The tactile interface I’ve wanted forever (better than a Buchla 223e).


I am thinking about getting a ER-301. I would use it mainly with samples - one shots, granular synthesis and longer samples for background sound with or without time stretch and effects. Since it is not a cheap device but looks great I’d like to learn something about the current state of development…how stable it is meanwhile?

Also I understood that the manual grains unit has a polyphony up to 16 grains but did not find any demos of it - does it really work with 16 grains in parallel? And how many units/tracks that play samples can be used at a time without overstraining the CPU?

Any experiences would be appreciated.


The stable releases are rock solid. They seem to come around every 2-3 months. The dot releases come fast and furious, but there’s no need to take them up if stability is your primary concern.


This is a fairly complex question to answer. I recommend asking the question in so you can get a variety of perspectives.

My perspective is: “I have yet to encounter any CPU constraints in my use of ER-301”.


Same here.
I have done crazy multilayered chains, and never encountered issues.
Never really needed to check/monitor CPU usage (yet)!


I will echo @jasonw22’s responses. My experience is the same.

For example, I had a grains track going full tilt, and was doing live looping on another channel. And had triggered sample slices on another. Didn’t even make a dent in the cpu.


But there are some fairly advanced ER-301 users who do things I scarcely comprehend, and these folks do occasionally seem to have to think a little bit about efficiency. The good news is, the system is so flexible, they seem to be able to find ways to achieve that efficiency without too much effort.