FPGA-Based DSP Options


#42

Micropython has been extended to several other development boards like some arduino boards etc.. and is continuing to do so for others. You don't need to limit yourself to the micropython board in your image.:wink:


#43

Just out of curiosity... why do you choose a dev board programmed in MicroPython which is relatively new, over a dev board programmed in C like the arduino which has many years of available free resources which has many concepts that transfer over to how you program the modules in the Axo patcher... ?? :thinking:


#44

@Gavin I was reading about that, being able to use it on Arduino etc, but I don't see the point when the board shown above is the official one that was designed from the ground-up to run it. The hardware is Open Source hardware, just like the software is open source, so buying the official board to learn on just makes perfect sense to me. If I were to start making successfull use of it in various products, I'd be able to take whatever components I need and just incorporate them into my own designs, so it's not as if the price of the official board will have any bearing on that.

Basically, I see the purchase of the official board as a sensible one-off move so I have the comfort of knowing that any problems I come-up against while learning it, are down to me, and not some hardware-based problem. Not only that, I like to support the people who brought the project to life, so even if I only ever buy that one official board, I've supported their time and effort in bringing the platform to us.

Regards the C/Python thing, I hear you, but I've been looking into various languages and trying a little of each, reading/watching opinions etc, and I really like Python cause it feels more human-readable and like there is less work to do in order to do the same thing, so less chance of errors and easier to learn. I'm not saying it is easier to learn or that it is less prone to errors, but it looks/feels that way to me.

What's cool about MicroPython is that you're programming a language that's easy to digest, and is getting run "bare-metal" on a piece of hardware that is so simple even a beginner could roll their own hardware with just a few components and a hot-air solder station!

So it's easy to build, easy to program, and very fast (cause unlike standard Python, it runs bare-metal). Also, unlike on a Raspberry Pi, there's no OS to get in the way and/or slow things down or complicate things. You just switch it on, and you have bare-metal Python :stuck_out_tongue_winking_eye:

There's nothing other than a persons own programming skills stopping them from developing their very own standalone computer with custom designed video and sound drivers etc. Just switch it on and off you go, bare-metal performance would come of your MicroPython programming skills!

Getting carried away there, but the point I'm making is it's true, you can do that. All of that said, for all I know, I could get into it and suddenly decide, nah, I'd rather learn C or whatever :grin:


#45

I guess the downside of a Python-based solution would be when it comes to trying to integrate with external hardware (which is likely to happen pretty quickly), you’ll likely have to write your own drivers etc. from scratch.

That’s could be viewed as a ‘plus’ for some, but is definitely a ‘minus’ for most, I imagine.


#46

I know, but that's the kinda stuff I want to learn it for, drivers etc (that's why the FPGA thing grabbed my attention). I'm wanting to get as close to bare-metal as possible so that I can write my own LCD drivers etc. I want to be able to design these things and reuse them across multiple products. Learning this stuff also cuts product costs considerably, cause driving for example an LCD and everything else from a single board or chip, removes the need for having to purchase dedicated driver boards which ultimately, would not perform as good as a custom bare-metal driver would.

If I'm going to learn this stuff, it might as well be stuff that is geared towards what I need in the long run.


#47

It could be a nice way to train and get some practical skills with the benefits of a great dev/experiment community...

FPGA – Intel Cyclone FPGA with 16K Logic Elements, 504Kbit of embedded RAM and 56 18×18 bit HW multipliers for high-speed DSP


#48

Realy need to wait and see what software they support it with...:thinking:


#49

looks promising, some comments seem to indicate that the IDE could be closed source...


#50

There are some other boards too.
I think that in a year or so we will have a practical idea of what can be done with these.


#51

Woaaaa, this is VERY exciting.

Waldorf seems to have teamed up with Exodus, the Valkyrie creator.

So its not Valykrie anymore, now its Kyra :slight_smile:

I love Waldorf synths they are damn cool. look forward for this one :slight_smile:

But 1900€.... is kind of a lot of money.


#52

There's no way I'd pay that for one, mainly because I'm not interested in such a synth, but it's definitely cool and I'm pleased the guy managed to work with Waldorf to bring it to market. The only hardware synths I have my eyes on at the moment are those awesome clones Behringer are putting out.

It means I can get a Model D for three hundred quid, so can't be bad :grin:


#53

there is also this guy:

http://scott.joviansynth.com/FPGA_synth/


#54

And actually also the Organ that I posted a YT video of above, have actually been turned into a commercial product:

http://shop.keyboardpartner.de/HX3-Sound-Engine/HX3-Modules


#55

Arduino just announced their new board that combines a fast ARM microcontroller, WiFi, Bluetooth, and an FPGA. All this is wrapped in a package that provides Mini HDMI out and pins for a PCIe-Express slot. They’re calling it the Arduino MKR Vidor 4000.


#56

now someone should pair this with the multiface/digiface hdsp express card line from rme...


#57

There's a project working towards an Open Source CPU using FPGA!

What great news to hear, and there's even development of a completely Open Source OS to run on it, which is equally great to hear (Completely Open Source OS running on a completely Open Source CPU). In addition to that, because the Open Source OS is designed to run on an Open Source CPU, the performance is not going to be limited to that of the current architectures put out by AMD and Intel.

Can't wait for this to really take off, and the good news is that it will due to it's Open Source nature:
https://openrisc.io/


#58

Mod note: Ive just re-titled this... its an interesting topic, but not really about an "FPGA axoloti" (its kind of veered of into more general fpga options - which is cool)


yeah there are quite a lot more FPGA options becoming available to the 'hobbyist', some full FPGA and also some interesting hybrids.

what I don't get though, is seems like a lot of announcements but not seen many (DIY/hobby/hacking) projects using them... anyone got any references for projects using them.
(current ones, Ive again seen announcements of projects that plan to use them, but then little after that)


#59

Yeah but you forgot to capitalise the "O" in Options so I just fixed it :grin:

I feel the same pretty much, I think the KYRA synth posted earlier is about the biggest example so far (synth-wise I mean). To be honest I didn't really expect Axoloti to become FPGA-based, but still I see it as a very appropriate tech for creating boards that could interface with it etc.

I think if anything will boost the uptake of FPGA, it is that FPGA-based Arduino board posted earlier. With Arduino having the following it has, incorporating an FPGA into it like that has got to have a fairly large impact on uptake I reckon. Only thing that bugs me about it is that I saw a video from the official Arduino dude and he seems to suggest that the Open Source tools for it will be cloud-based.

I hope I'm misinterpreting because I'm not developing anything whatsoever in any "cloud". I suppose that if this happens to be the case, people will port it to be offline anyway, still does my head in every time I hear the term "Cloud" though :roll_eyes:


#60

Exactly. i think those people(like me) who are just hobby programmers, wont get anything out of FPGA's unless other people build it...

FPGAs are damn complicated to program and I guess the point with them is they are COMPLETELY 100% customisable, you can program it to do what you want and only do that and do it really fast, cause it doesnt have any burdons from other tasks..........

But than comes at a price > a lot of hard work, you have to build/code everything from ground. I think 95% of Axoloti user simply won't be able to do that, so FPGA Axoloti would be aimed a very small target group.. Sorry I mean no offense to anyone who might have the skills to do so, hehe, you know who you are :wink:

I tried looking up some stuff cause it seems to be the new "thing" all the companies that has exceptional stuff, uses FPGA. Rme, Antelope, Universal Audio, to name a few, Waldorf(Kyra), but could not make any sense. Anyway, you re better programmer than me, so probably makes more sense to you, hehe :wink:

There are some tutorials around, but without a board to play with, i dunno how much fun it is. But yeah maybe you can get an overview what it takes to program such a thing.


#61

Even Behringer wants to cash in on the FPGA trend, looking for programmers with FPGA experience:

I never liked Behringer, but that Neutron thing, which is an original design, have gotten my eyes up. And this also sounds pretty cool, if they plan on doing some more original stuff.