There are lots of people with good ideas (and of course lots of people with shit ideas, that they think are good) who don’t have coding chops, but might have better ideas about how features can be implemented and so on, that may not have been thought of by the original designer/developers. So using AI is a way for non coders, who understand concepts, to create firmwares or mods to firmwares.
If people think that AI is doing everything with these alternative firmwares, that isn’t the case at all, the AI is just writing the code (often a bit wonky) and then the person is steering the idea and its implementation and so on. It still takes time, but just not as much time.
A load of manufacturers will spring up competing to make hardware based around standard chipsets with various form factors and relying on the community to make polished firmware to go on it.
The possibility is that there is going to be products in market soon (maybe 1-2 years) that are pretty close in functionaliry to £800 grooveboxes that you can install on a £100 Chinese box and easily switch to something else if you don’t like it.
That’s pretty scary if you are someone like 1010, Polyend, Elektron, Novation etc.
Sure understand these points. I just don’t get why Roland, korg, Yamaha, elektron , TE etc etc couldn’t have come up with something this capable before if it was possible. I’m sure market demand would have meant that if it was possible it would have been done before. The only reason I can think of is that the programming would have been too costly to conceive and implement?
Open source gives you an infinite pool of enthusiasts all working together.
That’s why Linux can compete with Windows even though it seems crazy you could outcompete Microsoft (I appreciate LInux is largely corportate funded now, but I mean in the old days).
Or similarly Godot slowly but surely eating up the market share of Unity in the gaming world.
The only reason I can think of is that the programming would have been too costly to conceive and implement?
Of course. All development up to a certain quality level takes hours that have to be paid. But even with ‘free’ software I’m not fully convinced all those AI generated firmwares will be seriously supported for a long time. You have to provide updates for a minimum of 5 years to sell consumer electronics in the EU, so you can’t really sell a device as fit for purpose for anything when relying on user-supplied firmwares. Marketing for such a device is impossible.
The there is user support. How do you even answer a question when there can be one of 200 different firmwares on the device? How do you determine if it’s the device or the firmware that’s broken?
Worth remembering that a lot of these alternate firmwares - not necessarily talking about just the FM-1 here, other gear too - are modular, with front end loaders that let users choose what features to add/omit, and many of them are a community effort too.
This makes sense for a number of reasons for individuals, but probably not for a company, whose primary and essential concern is profitability.
As someone developing custom firmwares versions, I reached out asking if they wanted to connect, they told me they would be launching an “open source community” in the next few days. So, we’ll see.
They have just added details regarding their official open source project here:
With how quick development is going, I can see M-VAVE capitilising on this with another device with more advanced CPU and RAM for more complex projects, open source devices with custom firmwares are possibly going to be the thing in 27
If so the topic with M-Vave firmware modification is open.
From there we need to think about how we organize our discussions. This thread has quite understandably gotten a little weedy. There is already an SMK-37 thread too. I don’t know, do we need a separate topic specifically on firmware development ? think-think-think.
A completely separate convo for firmware mods is ideal, it’ll be impossible to have conversations about the default device and the myriad custom firmwares in one thread with how frequently and how quickly development of these things happens
You are welcome to share and discuss both official and third-party firmware, including usage experiences, sounds, projects, bug reports, feature suggestions, and development notes.
So Mvave are fully on board with 3rd party firmware. I bet they have made a load of sales of the back of this.
I suggested in the ai thread companies will start to utilise others making firmwares and release machines happy for alternative firmwares to be released as it will increase sales.
I was quickly shot down
But I think m vave will gladly up production and see these fly out
Really wild that they are officially providing a greatly expanded baseline for developing custom firmware. But it’s a smart business move on their part to quickly embrace all this madness, especially considering that, as you suggested, they are presumably selling a LOT of units over this.
I accidentally violated the site guidelines with a previous post and a link to a ported version of sloop. I wonder if I can get that post retroactively unhidden now that it’s REALLY official.
While this is awesome, they were distributing some GPL work (including my own) compiled with closed source binaries, a violation of the GPL. I raised it and for now they’ve taken it down while the work out licensing. But they’re definitely embracing the space (and imo this shows even more so!)
Glad to see that M-VAVE fixed the FM operator detune issue, which is what got me dabbling in firmware modification in the first place, way before all the custom FW became available.
You can see that (tested again with @Indica 's STR-1 patch) the detune, and therefore the oscillators beating against each other, now match Dexed, unlike in their closed-source FW V15.
