Polyend releases the Tracker firmware source code
polyend.com
polyend.com
https://arstechnica.com/gadgets/2026/10/from-impulse-purchas...
https://github.com/aroum/fm1-custom-fw
see also:
> Reverse Engineering of the M-VAVE FM-1 Pocket Synthesizer Firmware
https://news.ycombinator.com/item?id=49994065 (75 points - 50 comments)
It's also worth mentioning that the release is not just a source code release, they added a lot of great features like sample streaming, and dedicated velocity and timing lanes. And they released APIs so that people can easily create new synth engines and effects.
The combinations of LLMs and agentic reverse-engineering skills have sent things into overdrive in recent months. As well as allowing people to repurpose cheap gear like the M-VAVE it has also revealed that a lot of larger vendors have been outright lying to their users about hardware limitations to more effectively segment their target market. In my the future is high quality controllers/converters with a generic SOC inside. Behringer is already going down this route and has the manufacturing heft to do it quickly and with a decent level of quality. Akai also seems to have jumped in on the trend.
Still got mine. That project produced some very nice effects and synthesizers ..
>M-VAVE
Lest we not forget about Monome NORNS and Zynthians too ..
That's the one! glad to hear yours is still going strong! Come to think of it, it's probably ready for a new lease of life if you let your coding agents go wild.
1. I tried running ghidra from claude about a year ago. I was not impressed, it was not really that good. It felt confused and not knowledgeable enough. I picked it up again a couple of weeks ago, and must say - it is nothing less than spectacular. I am having a small existential crisis here, Terry Tao light. 100x my skill, 1000x my speed it feels like.
2. Like all consumer products, design and brand is essential to the buyer/user experience of audio tools. Removing options in UX is what design entails a lot of the time. You are in reality paying a premium for options to be deleted from menus, gently nudging you into a curated brand culture. Acting like it is inherently bad is missing the point. It is like insisting everybody should roll their own Arch Linux on their laptops.
3. Audio tools/synths/fx/sequencers/controllers are instruments. The value of an instrument is to a great extent dependent on the time invested in learning it, as opposed to inherent things like "audio quality". Music is very in-the-moment, so any cognitive friction that takes you out of the flow is gonna detract from the experience. Having a "fixed" UX is in this sense a way of focusing, honing muscle memory and releasing flow. Watch Jeff Mills hunched over a 909, that is some Paganini shit. Standardization drives adoption, adoption fuels cultural expression, expression creates genres.
From an individual tech-loving buyers' point of view, the AI revolution in CFW is fantastic, from the vendors point of view, not that easy to handle.
The best strategy for a HW-vendor at this point is to get ahead of it and open-source the firmware under your own License Agreement, so you can possibly also integrate community work back into your official firmware.
I'm happy that some move in this direction.
The much more terrible strategy is litigation and moving as much logic as possible away from the product into some cloud-service (requiring periodic online-connection, etc.).
We already see both trajectories, so I hope the customers will decide wisely which strategy to fund...
Looking at how e.g. Roland is repeatedly building new hardware around the same digital "analogue emulation" engine, this shift in the market could make it more viable again to build actual analogue hardware.
Or..."let's add Wi-Fi and a subscription model so the device has to check-in to our servers every week"
(this is already happening, especially many established brands seem to be leaning more and more in this direction)
Korg Volcas are analogue and have sold orders of magnitude more devices than any of these boutique brands.
This new paradigm makes me happy.
Out of all the agentic dev things... vibecoding synth firmware might be the best outcome, so far.
I made an apparently common mistake:
My mistake: MWAVE or M-WAVE
Real name: M-VAVE FM-1
The latter will get you more stock listings on Ali and elsewhere. There cannot be a better stocking stuffer for anyone who is synth-curious.
Or to put it another way, FM is half of RTFM.
However, Tally Ho into the future of open source firmware. Can you imagine what we could have done back in the day if we had these software tools?
I mean, 30 years ago that Nord Micro was all about subtract synthesis in the UI, but it was digital all underneath.
Good news everyone! The FM-1 now has subtractive synth firmwares.
1990’s microcontrollers frequently use mask rom which is baked in at manufacturing and often aren’t installed with a means to dump that code on production instruments.
Also, 1990’s instruments often use gate arrays for the design logic and these are also burned in at manufacture.
So reverse engineering is more likely to require decapping and gate level implementation and probably won’t be some dissasembly and generic emulation.
"Out-of-Warranty Repairs Available: In the unlikely event that something ever goes wrong with your hardware while using custom firmware, it is not covered under standard warranty. However, Polyend is always happy to help get your instrument back up and running through its official out-of-warranty repair service (standard service, parts, and shipping fees apply."Hmm, out-of-warranty repairs for music hardware is pretty standard, at least here in Europe. I've done this with Elektron, Novation, Korg, Adam Audio, Yamaha and a bunch of other brands before as well, with everything from speakers to synths. Is it different in the US perhaps?
I've even sent gear running custom firmware in for out-of-warranty repairs, without any questions asked about it. AFAIK, even if you're within warranty with custom firmware, they would need to prove it's specifically the fault of the custom firmware in order to deny warranty claims, they couldn't refuse it outright just because of the firmware on the device.
No, I just add the brand to my List of Companies That Will Live in Infamy. Sony, Samsung, HP, Dell… the list is extremely long at this point. They have all produced inexcusably bad products that I had the misfortune of purchasing, and they will never see another dime from me or from anyone else wise enough to heed my advice to avoid them.
So for you is just better that you have a product that doesn't work properly, because shipping costs more than the value you lose from that thing not working? Do I understand this correctly?
If I buy a 2K synthesizer, that thing better work 100%, I'm not sure in what world someone would have to live in to not have that fixed...
> No, I just add the brand to my List of Companies That Will Live in Infamy
Cool... What does this have to do with the out-of-warranty repairs almost any instrument-maker offers already?
Just like us, but with graphics or landing pages ;)
(On a sidenote: and even Suno can be used in "correct" ways, as long as money and rights flow the right ways, like Arthur Baker said in his interview with ?uestlove, where he was describing how he would hire musicians and tell them what to play and how, and is saying "what's the difference with Suno?"
Creative / intelligent / motivated individuals instead use the instrument creatively (whatever that means in that person's field) because it's so easy to not make slop if you care just a tiny bit about the output the slop machine is producing.
Still, seeing the slop being churned out for fame still hurt me (i'm a software developer, hardware designer, used to be a musician and back then i had a couple of good ideas but i missed the eurorack train), but i am doing my best not to be an hypocrite about it
Amidst the understandable joy, one thing I've seen almost no discussion about: the change in the incentive to research and develop new algorithms.
Most of the new firmwares are re-implementing familiar synth algs into new places. Almost as quickly, aything new that comes out will instantly get cloned to every other device.
Pending strategy, I've frozen release on my newest instrument.
Like open source software, people feel entitled to everything for free. And plenty are feeling vengeful of designers/companies for withholding alg development. This results in rationalization to defend in some cases piracy.
It's nice that "music wants to be free" and that people can dream up new sounds, but this is challenging for new designs. Say Roland comes up with a new technique "WaveZorb." A few audio examples, an afternoon playing it into an LLM and poof, it's like Doom, available on any instrument. Who needs to buy a WaveZorb anymore? Then, Korg invents "FiltFountain," sells three of them, and the alg is on everyone else's device faster than it can physically ship.
It's not worth fighting, but thinking about: The path forward is to gray goo. With little to differentiate any instrument from another. And little support or incentive to make any new synthesis algorithms.
Its going to be very hard to compete, and therefore will be less attractive for innovators
Half a day of setup and i had my first curtis-style filter on the synth. 2 days of low intensity prompting and I have
- 3 Filters: Curtis, 303, Steiner Parker
- An additional noise source
- The oscillators from braids
- Prophet08-ish oscillators in a stable and instable fashion
- a granular delay, tape saturation and a klon centaur like distortion
- karplus filter and a body resonance fx
- polystep arp (different pattern per note) with scales or gated euclidean patterns
I have interest but no skill in dsp programming (never found the time to pursue it). I love the filters and oscillators, it went from a tame, boring synth to gritty synth I like to play. At the same time I have to stress that this is a result, not an achievement. Anyone with a 20$ claude subscription can do that and people do. So I think it’s smart to open source(and not novel, see modular synths diy modules). Though when gigging I need devices that are stable and working in any conditions and I’m willing to pay for that.
https://drive.google.com/file/d/1kG1jWOXU6-N7aGWstWGiOY-fP5k...
The stock synth can’t do that. But making a svf and oscillators isn’t something musical or groundbreaking, integrating it in an assembly blob is, somehow.
I paid 20€ for improving my synth and confirming my existential crisis as a developer. …but, truth is: I spent quality time with my daughter while claude worked and thats a good thing.
(I/AI had fun porting it to wasm/web! I don't want to seem too self-link'y, so if anyone asks, happy to share the link)
I also have a Polyend Tracker for sale if anyone's in the market..! It's a great device, I'm just trying to focus on having less right now.
I love to see this new trend in the electrical musical implement industrial complex.
Well done, Polyend! Godspeed, and looking forward to seeing how this develops.
The important thing about the alt firmwares is adding alternative and/or add new features.
But if you want to do FM synthesis, the other interfaces are just an opinion that is less integrated with the actual hardware side of the interface.
Other manufacturers could’ve done this. There are amazing F/OSS synth engines out there.
See also, the Zynthian hardware and the ZynAddSubFX synth.
I can't find it now, but somewhere off https://shorepine.github.io/amy/piano.html it talked about the Android work you did.
Edit: There is https://github.com/shorepine/alles#thank-you-to but I remember a more detailed thing too.
I always notice your name because I enjoyed Advogato back in the day and keep using bits of your work elsewhere.
I understand why someone can love Zynthian but it has little appeal to me…I’d rather have an MPC (they also run Linux) because of all the decades of design experience they embody and the volume of wisdom that I can access regarding them.
But that’s me, not you or anyone else.
They have effects, plugins, sampling, recording, etc.
Plus a pretty sophisticated DAW, robust hardware, and a lot of different flavors depending on what you want… And the ecosystem includes desktop tools.
The Xynthan looks like a playground for developers…and that’s cool. But I don’t want to deal with that.
When there’s an open source ecosystem for MPC like the one for Zynthian, it’ll be an interesting time indeed …
No commits for 2+ weeks
Few parts of it build as-is. Missing parts.
Readmes are out of date.
Maintainers not reviewing/taking PRs.
I'm on the discord and questions about building this are going unanswered.
Future remains open, but def not a serious committment yet.
We aren't talking about the same thing.
I am talking about trends and movements, you are talking about the details of a specific repo.
Look at everything springing up around the fm1, https://www.youtube.com/watch?v=uENkYmfIx9o
If only Ableton and Max would follow. Maybe that’s a weekend project.
At least this way there are still cause for eyeing in other directions than Ableton.
Why would anyone ever open source anything? Well, I don't know, it happens :) Parent raised the point, maybe ask them :)
> If you want to extend Ableton you can already do so with Max/MSP. Or write a VST plugin for it.
That's a really tight and constrained definition of what "extending" applications means. VSTs/Max is more like plugins/addons, then actually extending what Ableton can do, but I'm sure if you even know what Max is, you're well aware of this already. Today it isn't really possible to "extend" Ableton in any officially sanctioned way.
Ableton is amazing. I keep coming back to it. I've used it as my daily driver since version 3 (pirated) and I've paid for it since version 6.
It would be awesome to have a Linux version (aside from the headless one in the Push 3 standalone) and it would be awesome to have a native SDK for device development, and it would be extra awesome if they open sourced it (never going to happen, but I can dream).
It's cool if you don't like it, but don't tell other people what they should or shouldn't use. You can offer alternatives, but just telling people they're wrong is arrogant and unhelpful.
Don't shove things I didn't say in my mouth and call me arrogant.
I love the idea of the polyend stuff but they're too pricy when there are perfectly usable trackers with more possibilities like renoise.
The novelty of a handheld tracker wears off quick once you start plugging things in.
Hopefully someone fixes the arpeggiator too. They made some implementation decisions that really messed it up for chiptune style stuff. (Retriggering the envelope on every arp step, not allowing for arbitrary chords(!))