Hey!
Just saw this. This one is very similar to Axoloti, all though "for guitar". Cant really figure out what the programming environment is going to be, other than open source. It has got 2 dsps. Not produced yet, no demo and no price.
Any thoughts?
http://rezzonics.blogspot.dk/2015/07/dmfx-1-open-source-digital-multi.html
hmm, last year, so perhaps just a concept...
I thought DSP processors tended to use non-free development environments, this one for example the code-studio seems to cost about $800. I dont know if there is a free alternative, if not, then only developers who can monetise it are going to be interested. (imho)
Also I wonder if we should have another category for discussing non-axoloti things,
I think discussing 'related' technologies is interesting, and potentially spark ideas that could be used within axoloti,
but perhaps they should be kept separate from things directly related to axoloti.
note: I think there have only been a couple of topics like this, so probably a bit much for a dedicated category, more likely is we should use a off-topic category, there is a already a lounge category, but it requires trust level 3, which is only about 8 of us...I could lower this to 2 (64 users) or even 1 (300+ users) . Its quite nice to use trust levels, as would stop potential 'advertising' of products by new users.
@johannes thoughts?
btw: there is another way to consider this... if its an open category, then google will see it... which means if people search for similar products, then the discussions on the axoloti forum will show up, which might mean these people might become 'aware' of axoloti, by searching for similar products. just a thought 
Oh yeah. Ups. Forgot to check the date. Will see if it ever happens 
Not much shows up when googling. Saw one clip one youtube that only showed a display working, not really any sound examples.
Yes a NON axoloti category would be nice. I did request this a looooong time ago. But didnt hear back from any of the developers here and the only user comment was " there is enough places for this on the web all ready........ " haha So I dropped it again.
I posted this in the same category as the Organelle was posted in. Hope it was not wrong place.
Anyway, yeah it is good to check what else is coming and take ideas from them. Seems like the DSP board solution is becoming the future for experimental music devices(Microgranny, Organelle, Axoloti, DMFX-1). I think A LOT of these music devices will show up over the next years in many variations.
Still happy with my Axolotis, though
nah, here was fine, just was wondering... I felt a little uneasy when posting the organelle stuff too...
DSP board solutions.
hmm, think we have to be careful with that term, Id usually associate this with DSP chips, which axoloti, organelle, bela etc dont use. they use more generic technology... hence Ive tended to call these things 'digital modular' (not ideal, but I know what I mean )
the cheap/low powered ARM chips have become a lot more powerful, so like desktops, its becoming possible to do DSP on general purpose chips, rather than DSP chips. look at recent iPads, to see how much computing power is now available.
its interesting we are seeing this... though Im not sure how much it will really grow, PD/Max/Reaktor have been around for years, and not taken over. the biggest growth was VSTs, and now arguably iPad music apps.
So I wouldn't be surprised to see more 'out of the box solutions' or re-progammable external 'synths' e.g. imagine Axoloti like a VST host, when you buy plugins. In some ways, this is how OWL is being sold.
But this may be just a reflection, of what I want... a rack of dedicated hardware boxes, driven by my controllers.
(I kind of get the appeal of things like Volcas & TE PO, but I dont need the control aspects, I prefer this to come from dedicated controllers . so I can build things like lego )
DMFX-1 looks complex and expensive to me, and the path from a single prototype circuit board to production is substantial! A 16bit fixed point DSP is no fun to program, a proprietary compiler does not help. Using a 2nd DSP for control, is a bit strange. No external memory means delay will be limited to a few seconds, but that's ok for a guitar delay fx.
Off-topic discussions, so far only a few an still related to Axoloti in a way. I'm OK with opening up a lounge, and lowering the required trust level. So far there hasn't been any spam on this forum, I'd like to keep it that way.
Yes sure. I am not sure I use the right vocabulary. I am just thinking of DSP as digital signal processing of audio signals...... or something like that
But like Microgranny, it is dedicated to one thing. And I think there will be more and more of these devices too, along with the open source projects like Axoloti. Probably more projects like Microgranny than Axoloti. But I think it is becoming the future, cause you can keep cost fairly low and still make something pretty advanced and pretty decent quality.
Yeah I think ease of use is the one of the reasons. Most musicians, with no background in programming etc. are very quickly lost in PD and MAX.
DMFX-1videos, what I could find:
Hello,
I am Rezzonics, the creator of DMFX-1, thank you for the interest you showed on DMFX-1 Digital Multi FX guitar pedal.
I think Axoloti is a great piece of equipment keeping a good balance of performance and simplicity. I think it's great that there is an effervescence of open source HW/SW initiatives like this and I hope there is plenty of room for other players because it can only enrich the DSP and Audio DIY community. I hope we don't see each other as competitors and I would like that some collaboration could be reached in the future.
I understand this is a community forum devoted to Axoloti but since there are several questions regarding DMFX-1 I would like to provide some answer to them, and I will be glad to answer any additional questions you may have.
DMFX-1 is based on TI TMS320C55xx 16-bit fixed point DSP. By playing with the ezDSPC5535 kit, which actually is quite compatible with DMFX-1 and allows doing many things, I realized that LCD and user interface limited a lot the audio capabilities of the DSP, that's why I decided to use two of them, one dedicated exclusively to audio tasks and the other dedicated to user interface, tuner, noise suppression and other no real time tasks. I could have chosen a cheap MCU instead of a DSP for doing this but that would require an additional development environment, so why not use an identical DSP, which cost less than 5$ that is able to deal with LCD, GPIOs, interrupts, I2C bus, digital potentiometers and everything required to manage the user interface?
Development environment is TI Code Composer Studio. License is free for XDS100v2 emulator environment that is what is actually used.
Code is developed in C language (or assembler if you are brave enough). By using good practice techniques, C optimizer can get good code efficiency results.
TI offers DSPlib with plenty of assembler routines for IIR, FIR filters and mathematical functions like sine, exp, log...
TI also offers a CSL library with lots of examples to implement peripheral interface control functions: SPI, I2C, LCD, GPIO, SAR ADC, interrupts...
DMFX-1 also uses DSP/BIOS provided by TI to deal with real time task management.
DMFX-1 will be certainly more expensive than Axoloti, I think it could be sold at 100€ to 125€, it consists of a 6-layers PCB main board and a 2-layer PCB daughter board.
It includes a preamp to allow direct connection of a guitar, output buffers and filters.
I also found that non-linear processing like distortion, overdrive and fuzz are difficult to implement in the digital domain, and not very efficient, so analog distortion circuitry is added.
Data can be routed via USB. The kit could include up to two SD cards.
I conceived this kit for DSP educational purposes or just have fun creating your own digital effects. Programming in fixed point may be more challenging, but I think that's actually part of the fun.
Latency on DMFX-1 will most probably be very similar to Axoloti: ADC/DAC codec latency plus the internal ping-pong buffer latency.
It would be interesting to see performance benchmarks: STM32F427 has 168MHz frequency while C55 has 100MHz, I guess performance will be mostly related with code efficiency more than processor speed. In terms of audio quality I think that 16-bits is more than enough, specially if final use is for guitar effects.
DMFX-1 is using 48kHz sampling, but I think that sampling frequency could be reduced without affecting a lot audio quality (for guitar effects at least) which could greatly improve efficiency.
Sorry for the spam and the long speech, I really enjoy this subject.
Thanks for the info, really interesting to hear about other projects.
are you close to a release?
sorry for misrepresenting code studios cost, I search on the TI site, so didn't find it as part of the eval board offering, only bought standalone - i still find TI's offerings pretty confusing
is the C5535 eZdsp USB Stick Development, similar to STM discovery board i.e. a complete evaluation kit?
Im personally fascinated in all these things, its an interesting time to be a developer in this area, would be interesting to look at the CSL and DSPLib.
in a similar vane, Ive been chatting to someone about the Beagleboard X-15 (http://beagleboard.org/x15), which mixes both conventional ARM A15 and DSPs TI C66 ... but they are as rare as hens teeth at the moment
Thank you.
Yes, I think I could have a first version of the code ready quite soon.
I am looking for help in manufacturing and distributing the HW, any help is more than welcome. And then I would need to build a SW community with people interested in developing code for the HW. Not easy.
Yes, the ezDSP5535 development board is a complete evaluation kit that includes an audio codec, mic and headphones, small graphic LCD and SD card. A Code Composer Studio license is also included with TI ezDSP5535 kit. It costs around 100$.
The Beagleboard X-15 is a beast, it's playing in a different league, the Sitara AM5728 includes 2 ARM Cortex-A15, 2 C66 floating point DSPs, 2 ARM Cortex-M4, graphics acceleration. The processor itself cost 100$.
cool, sounds like your at an interesting stage with some challenges ahead.
the X15, its approx 250 euros (but mouser are saying 26 weeks delivery lead time!) ... which is not bad considering specs, and the combined A15 + DSP could be pretty interesting for a musical appliance. however, there seems no info on the codec, and using small jacks for stereo in/out, I do wonder what the quality would be like....
not much point in having lots of processing power, only to find the noise on the output makes it unusable.
Thank you for the insider knowledge
I think the "analog distortion" sounds interesting. Also the FIR and IIR options. Will follow your project What page should we follow to get the latest updates?
And also I dont think I understand how you would program new patches on the device. Does it come with a patcher environment similar to Axoloti or is is PD based, or etc.?
Thanks again.