Nord User Forum

[Benis67]

Nord keyboards and synth use multiple dsp of the 56k family, first Motorola, then Freescale and lastly NXP. The family of these dsp no longer exists and in the future the available stocks will be enough only to repair broken keyboards. The "insiders" know the problem well and it is also known that it is very complicated, to the extent of the impossible, to convert the machine code used for these DSPs into another platform. There is a lot of discussion about the next piano or synth model that will be presented at NAMM 2020, but it would be very interesting to know how Clavia thinks about tackling the problem. In practice it will be necessary to create a new hardware synthesis platform, starting practically from scratch. A similar situation occurred for Access Virus synths that use the same dsp. Do you have any news about it? https://synthmorph.com/blogs/news/access-virus-future-kemper-amp-virus-ti3

[Mr_-G-]

There are already some synths that do not use DSPs: Novation Peak and Waldorf Kyra use FPGAs. The problem of code compatibility means just that: the code is not compatible, but the underlying algorithms should be implementable in different compilers, perhaps even more efficiently.
So manufacturers will need to hire some additional programmers to implement existing algorithms. I find it a bit hard to believe that there will be a shortage of DSPs for repairs any time soon (but I have been wrong before! ).

[Benis67]

There are already some synths that do not use DSPs: Novation Peak and Waldorf Kyra use FPGAs. The problem of code compatibility means just that: the code is not compatible, but the underlying algorithms should be implementable in different compilers, perhaps even more efficiently.

Yes, Novation Peak and Waldorf Kyra use FPGAs (Intel Cyclone), Roland BMC SoC, Yamaha SWP70 SoC... the biggest problem is that Motorola DSPs are normally programmed, for faster routines, in native assembler. To move to a new platform it would be necessary to go back to sources and algorithms in more standard languages, typically C ++
Purely transferring the Nord algorithms to a C++ is certainly possible but it requires to document all algorithms with all its qualitative / quantitative attributes which could then be transformed into any high level language and prototyping on a newer CPU developer board using the specific compiler. It is a very complicated operation that can take years of work. And it may be impossible if, for example, some developers responsible for parts of code no longer work for Clavia it might be more convenient to start from scratch using only the sample libraries of the old models. it would be interesting to understand if this transition, certainly very expensive, has already started or if on the contrary they want to monetize as much as possible with the remaining dsp, thus adding only minimal upgrades. Obviously I hope this is not the case

[maxpiano]

The "insiders" know the problem well and it is also known that it is very complicated, to the extent of the impossible, to convert the machine code used for these DSPs into another platform.

Complicated yes, most likely, but not impossible, particularly if the code was not written directly in Assembler but maybe using a higher level language such as C, in that case at least the algorithmic part of it (the "core") should be rather portable. Anyway, is there any alternative?

PS: welcome to NUF...

[westharp]

Didn't Nord move to a new platform architecture with the Stage 3? This would explain the following:
- the many bugs in the Stage 3's OS at the beginning
- both the Electro (6) and the Piano (4) have been updated in 2018, shortly after the Stage 3, although the Piano 3 was only two years on the market at that time

I always assumed that they moved to a completely new platform with the OLED-models. Can anyone confirm this?

[Schorsch]

Purely transferring the Nord algorithms to a C++ is certainly possible but it requires to document all algorithms with all its qualitative / quantitative attributes which could then be transformed into any high level language and prototyping on a newer CPU developer board using the specific compiler

True, but there are tools for automatic / semi-automatic code generation, like Matlab's Embedded Coder and DSP Concept's Audio Weaver, both are expensive Matlab-based code generators. It’s possible but this would be still more likely a huge and confidential manual task, rewriting those from the scratch in C/C++ to provide maximum portability and the ability to scale with newer generation of hardware platforms.

[CountFosco]

https://www.youtube.com/watch?v=cidW8PKgQS4

[Benis67]

...............Nord should offer us the option of board upgrades for current models if they were to release new boards using either different DSP chips or JFETs, etc. I'm not talking free, but at least a fair price for these boards.
If I'm correct, which I'm likely not, if it is the case that the DSP chips used on Nord hardware, that would mean that production of these instruments would at some point to be potentially discontinued and re-released using newer board specifications. As I say, I maybe very well wrong here.
lew

It is not really possible to upgrade an Hardware architecture with another completely different . There will always have to be a zero year when synths and pianos are being created with the new architecture

[cgrafx]

There is no compatibility between old boards and new boards now. This is no different than any other Nord model upgrade. There is zero compatibility between Nord models other than stuff that is physically system independent (in this case samples). You can't take parts out of a Nord stage 2 and use them in a Nord stage 3, and programs and synth settings are not transferrable. This will be no different if the underlying physical architecture changes.

To the outside world it will simply be another Nord model.

[FZiegler]

I wanted to buy my first Nord Stage. And wanted it to be played (and repaired, if necessary) for the next 15 years roundabout - considering the invest. Meaning it should be usable as long as the oldest Nord Stage Classics out in the wild. Is this DSP thing a real issue to that point? Or is the computerized setup of the instrument a threat for the life period anyway (I never would keep my laptop for such a period)?

I'm playing my Yamaha CP33 for 12 years now, and it fits like on the first day. Do I have to think differently about a hightech instrument like a Stage 3? What's your opinion?

Nord staff won't give any statement about the DSP thing before issueing a new product line, will they? Will not help to wait a few months...?

I wouldn't care much if I already had one (as things are rarely totally controllable), but now I'm not sure any more.

Thanks @Benis67 for mentioning!