Nord User Forum

maxpiano wrote:

cgrafx wrote:

Benis67 wrote:As is known, the DSPs of the Motorola/Freescale/NXP 56K family have been out of production for years and sooner or later even the Nord Keyboards will have to use an alternative generation of DSP/CPUs for synthesis (presumably ARM-based).
Will NORD STAGE 4 be the first of this new family? And if so, will the differences in sound generation be perceptible?

Nord already switched away from the NXP 56k family when they launched the NS3.

And anyway a DSP executes algorithms which are deterministic (for a given input you have a given output) so, regardless the DSP used (performance aside) same algorithm = same result (identical audio data stream), different algorithm = different resulta. DAC and analog stages can instead introduce differences in the sound output, even when audio data stream is generated by the same algorithms.

The problem is that the code for the 56k family is not obtained by compiling languages such as C++ but by exploiting native, super-optimized code libraries.
To switch to a new architecture you have to write almost all of the code as there is no tool that converts 56K machine language into code for other architectures.
Certain operations that are simple to perform with a DSP can be much more complex on Arm CPUs and vice versa.
In many cases it becomes easier to rewrite the code from scratch, keeping the characteristics of the old models as much as possible: which can be a good thing.

Benis67 wrote:

cgrafx wrote:

Benis67 wrote:As is known, the DSPs of the Motorola/Freescale/NXP 56K family have been out of production for years and sooner or later even the Nord Keyboards will have to use an alternative generation of DSP/CPUs for synthesis (presumably ARM-based).
Will NORD STAGE 4 be the first of this new family? And if so, will the differences in sound generation be perceptible?

Nord already switched away from the NXP 56k family when they launched the NS3.

what is the source of this information? From the videos it is clearly seen that there are 5 DSPs of the Motorola 56K family

Info came from this post
general-nord-forum-f29/namm-is-just-aro ... 89#p138291

Additionally the NXP chips previously used are "No Longer Manufactured"

M1tsos wrote:

mozart999uk wrote:Another forum user (m1tsos)? had spoken to a nord tech and had found out that the NS4 (88) might have a different action (possibly closer to the piano). Be interesting to see if that's true and how it will affect the playability of the other sounds (organ and synth)

It seems like what i said before 6 months was truth

It is going to have triple sensor keybed at the 88 version.. it is possible for kawai keybed also..

I think it would be more likely that it would be the same action as the Piano 5.

Info came from this post
general-nord-forum-f29/namm-is-just-aro ... 89#p138291

Additionally the NXP chips previously used are "No Longer Manufactured"

thanks for the info

Benis67 wrote:

maxpiano wrote: (...) a DSP executes algorithms which are deterministic (for a given input you have a given output) so, regardless the DSP used (performance aside) same algorithm = same result (identical audio data stream), different algorithm = different resulta. DAC and analog stages can instead introduce differences in the sound output, even when audio data stream is generated by the same algorithms.

The problem is that the code for the 56k family is not obtained by compiling languages such as C++ but by exploiting native, super-optimized code libraries.
To switch to a new architecture you have to write almost all of the code as there is no tool that converts 56K machine language into code for other architectures.
Certain operations that are simple to perform with a DSP can be much more complex on Arm CPUs and vice versa.
In many cases it becomes easier to rewrite the code from scratch, keeping the characteristics of the old models as much as possible: which can be a good thing.

Regardless which language you use and if you do it manually or not, an "algorithm" is the (logical) transfer function between (all possible combinations of) inputs and outputs, if you change that then of course you will not obtain the same results (but same would apply if you rewrite code on the same DSP i.e. you implement a different algorithm)

PS: I see we have some notions of DSP/Microcontrollers world and embedded programming in common

I hope it's got a sensible amount of memory. An amount that reflects the fact it's 2023 ...

Can't wait to hear how it sounds, especially on the Hammond/Leslie/Drive side.
Nord generations all have their soundwise particularities.
Hope it will be available worldwide.
And not too pricey.

This is exciting news! I had a feeling that Nord was working on something like this, though I figured that April NAMM would be the announcement.

It's interesting they are announcing it in February. I wonder if it will be available before NAMM or if they will keep working on finalizing features/presets/production between now and then.

As we used to jokingly say about NAMM prior to Covid, it stands for "Not Available, Maybe May" and that was for a January NAMM...now this one might actually be available by May?!? If so, that's cool!

I will definitely investigate its features and decide if it's a worthwhile upgrade from my NS3. I still have one of my NS2 units that suits me quite well after about 12 years.

Frantz wrote:Can't wait to hear how it sounds, especially on the Hammond/Leslie/Drive side.
Nord generations all have their soundwise particularities.
Hope it will be available worldwide.
And not too pricey.

I think something often overlooked by non-Nord owners are how good their DSP effects are.