2 Aug 2012

Update on Nord, a year later

Kinda forgot about this blog again, basically now it’s little more than a scratch-pad.

So, it’s been a year since I bought my Nord Stage 2, and I have both praises and criticisms for Clavia.

The good

Latency

Honestly this is where having real hardware is an advantage (although we’ll get to some criticisms on the hardware shortly). When I was on software, I got the latency down to about 20ms which is playable, but not really what you want for serious work. With the Stage 2, the latency through my whole audio system (all analog) is <1ms.

Tweakability/flexibility on stage

If you’re not yet aware, typically wherever I live, I end up playing for my church. This means I’m on stage a lot. A keyboard that on stage is both 100% reliable and easily tweakable is hard to build, and Clavia’s done it. I take advantage of it too… very often we get into the middle of a song and I find myself adjusting envelopes, filters, oscillator shape, drawbars, etc. It’s extremely nice to have.

MIDI

I need two physical keyboards even if there’s only one set of sonic possibilities. The Stage 2 does a good enough job of dealing with my Axiom Pro 61 that it’s quite usable on stage. I still wish MIDI routing settings could be saved with the program though, especially as they’ve added similar settings in OS releases since v1.26. It can be tedious to remember which zones to enable on the Axiom. I wish I could just set MIDI rules on the Stage 2, save them with the program, and leave the AP61 on “Zone 1” all the time.

The bad

The main reason I’m writing this post is to codify some of the criticisms I have of Clavia’s design. I understand their design decisions, but disagree with many of them.

Architecture

All of Nord’s products are based off of FreeScale’s CPUs and DSPs — specifically, the CPU is the ColdFire. It’s an old, cheap processor, and the instruction set is a reduced version of Motorola’s 68k. There is support within Linux and gcc for it. I have yet to definitely determine what kernel and C library are in use, but the OS X packages for their firmware contain a blob which is definitely unencrypted, and one interesting string in there is “__gnu_cxx20recursive_m”.

The issue is, I’m really starting to see some limitations hit with this architecture. Like 18 polyphony combined between the two synth slots. I hit that ALL the time. There’s only 16MB of main RAM, and from what I can tell, about 16MB of flash space for the OS. These parts probably total <$50 not including the DSPs. Being that this is keyboard cost me $4,450, I would think Clavia would consider building their boards based off of a more modern architecture with processors that scale well while still generating minimal heat, such as ARM's Cortex A9 or so. Meanwhile, the ColdFire has remained largely unchanged through the last 15 or so years. I do wonder how much work it would be for Clavia to port their OS codebase to ARM; I would guess it depends mostly on integer/float sizes and how much of the OS is currently written in m68k assembly. Judging by what their OS looks like in a hex editor, I'd say most of it is probably C/C++. Yamaha, Korg and Kurzweil have all re-architected their platforms in recent years around Linux and either x86 or ARM. I realize Clavia does not want to be like the big guys, and that's why their keyboards are so awesome, but there is some legitimacy to this hardware architecture approach.

Outputs

The Stage 2 uses I²S internally. Converting I²S to S/PDIF is doable with very cheap, off-the-shelf parts, and yet Clavia has yet to offer any boards with a digital output. Perhaps they think their customers are all old geezers that just want an instrument that sounds like something from the 60s or 70s, and don’t care for this newfangled digital crap. Um.

Product maintenance

I have to give Clavia credit for maintaining their products well in terms of library support, but for some reason or another, flash space is so tight on the Stage 2 that they can’t swap out the drawbar and Leslie engines for the new ones included in the C2D. I dunno about you, but I would readily give up the pedal noise feature (which I never use, since I don’t have their $200 triple pedal) for a better sounding B3 and Leslie.

In general, Clavia seems to be against adding significant new features, particularly in the area of sonic possibilities, after release; once it’s released, save for bugfixes, it sounds how it’s gonna sound forever.

Summary

The Stage 2 is great, though I really think it would benefit from some hardware and architecture changes to bring it into this century. Clavia’s done a great job researching how the sound of a B3 or Steinway grand piano works, and I have no regret about paying out the ass for the results of that research. Now what they need is a good hardware engineer who can make their instruments a hundred times more powerful in terms of processing abilities. Once they do that, they truly will have built the keyboard to end all keyboards.

Posted in Uncategorized

One Response to “Update on Nord, a year later” (post new)

1 Mar 2013
William R. Tschanz
 

I have had the stage 2HA88 for almost a year. Mostly at my home studio studying my sheets and charts, modes and theory. Had the Nord out on stage very little. The few times I had she would seem to detune (go out of tune). I thought it was possibly a sag in voltage from the stage circuits. I was running Vrs.1.40 during the first issues last fall. had every major brand of digital keyboards and the interface is quite easy to navigate. Being proficient at the Modes and my theory I have no need for transpose and via the interface programmed all global, voice and pitch stick parameters to the OFF position. (no lights on when switch is activated). I uploaded Vrs. 1.70 and thought I would find things on a steady A440. On stage it happened agian last week. How humiliating to have your Nord keyboard detune in the middle of a performance. I was not playing by ear and had no tonal reference at home to hear if the Nord was holding A440, only when the few times out when in a performance with others did I find this issue. Ghost in the machine? Bad code? Defective flash chips? An untidy floating intiger? Hand wired that particular day maybe the tech had a bad day or a good night the day before? Just pick one and keep guessing. Two weeks before the one year warrenty it is in the repair shop with as of yet no answers. As an artist I am completely overwhelmed that I would have a very expensive instrument(relative to the Keyboards)so defective as to be unplayable. Beware if your Nord, there might be a ghost in your machine.

March 1st, 2013 at 10:09 am

Leave a Reply