Reflections on the Axoloti and Akso boards and Bela

#1

Isn't is strange that both the creator of the Axoloti board and the Akso board have gone AWOL at the same time.

Are they up to something? :slight_smile:

The Akso board is such an elegant refinement of the Axoloti board, its a shame they are both made of unobtanium.

I bought a Bela and Beaglebone to experiment with and the cost of the Bela proves that the Axoloti boards are almost 'free'. Maybe that doesn't help.

The cost of the full Bela is approaching 2x the price of an Axoloti/Akso. However the academic team that have produced Bela have a chunk of funding coming from research grants as they all work at Queen Mary College, University of London, and also are officers of a UK limited company that has growing assets.

It's hard to compete when a) Bela has its basic "beer and pizza" cost of living carried by taxpayers as the people are salaried by academic positions and b) they can spin off a business on the side that doesn't need to pay any salaries.

I'm not knocking them by the way I think Bela is a great project.

#2

Axoloti is very hard to get these days. Have been for a while.

I was considering getting the Akso board at some point but, tbh I am glad that I didn't as it seems a bit hard to get any updates on whats actually going on. I think intention is good, but it seems like a big project for a single person, like the original Axoloti.

I have Bela too, just got one a month back. Didn't have time to do much with it yet, but I did play around with it a bit and it seems very promising. I love Pure Data, so this is for me, hehe :slight_smile: .

I also have a Pisound based RPI3 and an Organelle and Bela seems to be best for me for now...... All though a bit expensive, I think I like it a lot :slight_smile:

#3

Bela looks interesting and I should certainly look into it further.
There is also Electrosmith's Daisy which looks very promising and has been featured on this forum before. There is an integration of PD vanilla that is still in development. However, I really like Axoloti for many reasons: the simplicity of its programming interface, the start-up time, and most of all the community that I have been following for a few years now. The user library is huge and contains a lot of high quality objects that makes it very easy to create some very nice-sounding synth/fx.

The Akso is really a missed opportunity, and I find it quite sad that the development of the Akso and Axoloti seems to have stalled.
I was wondering a few days ago if it would be possible to create a fully open source Axoloti (hardware+software). I don't know how difficult it would be, but it seems to me that the akso was developed in a relatively short time.
This way it would be possible for any user or group of users to make their own Axoloti in a DIY way in the future.
Unfortunately I don't have the knowledge to take part in such a development, and I obviously don't know if this is even possible. But I would be so sad to see this great community disappear in a few years.
Because I already have 3 axos, this proposition is more focused on keeping this community alive and attract new users in what I consider to be the best option for standalone virtual modular synths today.

#4

Axoloti is open source, hardware and software. That’s how Nic got Akso working quickly.

IMHO, the Axoloti patcher is the thing that makes Axoloti and Akso special. It’s a higher level abstraction than Pd, far less esoteric. But Pd is way ahead on portability.

#6

Thanks, I thought only the software was open source, don't know why!

#7

I think they're probably hold captive by the CIA or something similiar, as if every synth enthusiast looking up some expensive Sherman Filterbank or the likes would go and try to patch something similar before spending the money and in the process learn to program stuff and proactively create instead of consuming it would essentially break capitalism... :boom:

Now seriously, it would be so great if their absence hinted at something new coming up, but together with the boards being unavailable all the time it kinda gives me a "sinking ship" vibe. At least the demand seems to be there.

#8

Keeping the Axoloti going is a matter of finding the right person/people to make a commitment to doing the herculean feat of updating the firmware and the hardware. I mean, if someone commits to recreating the Axoloti from the original schematic, the firmware needs very little adjustment. Having an M7 cortex is nice, while I'm also very satisfied with what the original board is capable of.

I'm currently learning embedded programming (slowly) because I want to do exactly that, by the way. Cheesy as it sounds, Axoloti pointed me to exactly what I want to spend my time doing. Mentioning as I do every time the idea of updating the Axoloti comes up here, that if anyone has interest in working on an update as a long term project, count me in. However it may still be a while before I'm at all helpful.

It would be great if there were an Axoloti core in a DIY format, similar to a Eurorack kit.

#9

I do have a bela too (salt and salt+), but haven't done much with it yet as I'm in between moving to a new house (again...after my current house just got renovated last months haha) and haven't hooked up my modular in the meantime.

But the big difference is indeed the patcher of the axoloti. It pretty much resembles the nord modular environment and synthedit, both of which I had used before. The biggest plus to me is the sheer speed of which you can change code inside the editor environment and check whether it's working, how the output signal has changed or what is giving errors. This increases the rate of learning how to code immensely! I learned how to abuse the fft-code of the spectral-fft module to do ifft and quickly generate large, interesting wavetables simply by trial and error. If I had to do this with bela or synthedit, I'ld probably still be nowhere..

So, I really hope it's not the end of the axoloti, it would be a real loss if no more boards would come out. I'm still not finished writing new modules! And it would be a pain in the ass to have to do it all over again for another board..

#10

From what I hear at work, all electronic circuit boards are made of unobtanium these days. All kinds of electronic components have crazy delivery times (70 weeks or so, meaning if you order your parts now, you'll get them in September next year) or are "on allocation" (meaning you should be happy to get part of your order filled, some time in the future). The only way to get things manufactured at reasonable prices is to have long-term contracts already in place.
This situation is not really new: I remember our hardware developer complaining in 2019 about delivery times for some "chicken-feed" (commonplace passive SMD components like 10kOhm resistors or 100nF ceramic capacitors) suddenly rising to 13 weeks. But instead of "the market" solving the problem it seems to have gotten worse.
So, maybe another production run of 500 Axolotis just turned out to be impossible this year.

#11

So I have heard, but I think in such small quantities it might not be a problem. I've been buying components on RSOnline, Mouser and Farnell this week for building the Pepper Eurorack module for Bela and poked around looking at the STM stuff and in quantities of 100's it seems there is stock for STM32 MCU's and the main thing I find problems regarding stock is always potentiometers and trimmers because they are big and heavy.

I see that two of the main producers of chips that supply the auto industry where shortages are being reported were based in Texas and hit by the power outage there.

#12

I found a blog post from a while back where it said that the patcher for Nord Modular was available and could be used to generate CSound output from existing patches.

Closed source, but anyway I couldn't find it on the manufacturers.

It would be interesting to give it a try, but I think it may have been a transitional thing for annoyed owners of the Nord Modular synths.

#13

I skyped with Johannes last week.
He is definitely alive and kicking!

He told me the main problem now for the production of Axoloti Core boards is the availability of the processor.
I don't know how much I can disclose here now (we didn't discuss this...) but I think I can say I've seen two different prototypes with an M7 processor on board!

So I'm pretty sure this is not the end of Axoloti.
It's just a very difficult project for one man to manage everything. This together with the past and present events it hasn't been easier for him too...

Anyway kudos to this kind community and may the Force be with Axoloti-development :slight_smile:

Johannes Taelman
Where to buy a Axoloti Core / Out of stock
#14

Wow awesome news on all counts

#15

I think the patcher app is really awesome too, loving the Axoloti to death.

#16

I skyped with Johannes last week.
He is definitely alive and kicking!

He told me the main problem now for the production of Axoloti Core boards is the availability of the processor.
I don't know how much I can disclose here now (we didn't discuss this...) but I think I can say I've seen two different prototypes with an M7 processor on board!

ok! this is really good news! I hope Johannes is doing well himself too.
Great to hear that there might be a faster board coming. That would really help with some of my latest modules build around ifft. They take quite some cpu (spikes) which makes it a bit hard to really make a good, usable synth based on them.

#17

I think this must mean the H7 processors as they are all M series Cortex, so I'm guessing something like applying some of the changes in Akso to the Axoloti.

I've been bowled over by the size and convenience of the Akso board and the brilliant module documentation. In fact I was reading a load of your axo-object code last night on there as I was trying to understand the difference between S-rate and K-rate and on his docs he makes this very clear by relabelling them Control-rate and Audio-rate.

I'm planning a hobby project trying to build my own firmware on an H7 as a learning exercise, I've just ordered a Daisy, but it would be great if there were 3 H7 maker boards to choose from as the firmware would be substantially the same on Axoloti, Akso and Daisy.

#18

If I understood it right,
the Daisy cannot run the Axoloti firmware for now, because it doesn't have enough memory in cache.
The form factor and power of the Akso are nice, but not everything is compatible yet from Axoloti, some user modules wont work and there is not much info regarding GPIO and such.
As far as I know, Akso seems on hold and there haven't been fixes for bugs.
The trusty Axoloti might become outdated, but it's at least rock solid...
Let's hope we get some more info from Johannes or Nicolas soon :slight_smile:

#19

Ah, no I didn't mean in my example that I would attempt to run patches on Daisy. I'm thinking of creating something that has no O.S. but is a single hard coded instrument, it's just a way for me to learn things more hands on. I'm really exploring the space of 'maker instruments' out of curiosity.

For working with the Axoloti patcher and 'virtual machine' you need Axoloti and I agree it's the main stable way to do that.

#20

such good news! i was realy getting worried....

#21

I haven't seen your post.

Great to know that Johannes is fine !

At the moment, nothing compares to the Axoloti.
I have a bunch of "Daisy seeds" and - put aside the H7 horsepower - it is really laborious to use compared to the Axoloti.