randy's Recent Posts

We're putting together the first production Soundplanes this week. Here are some shots of number 0001 moving through the various stages of production.




This is me using fine steel wool to rub down the paste wax that protects the finish. If I ever have an intern he or she will definitely have to do some “wax on wax off”. If you don't understand that reference, put The Karate Kid on your Netflix queue. It was pretty important for the MTV generation.




Gluing the back plate in. Our friends at Gurian Instruments gave us a great tip on using Teflon tubes to control the application of super glue.




Here's the back plate glued into the case. Yup, this is serial number 0001, headed to Portland soon.

It's cool. I always appreciate the input---it's up to me to make any sense of it.

Chris and I are on the home stretch, assembling production machines. I calibrated the first Soundplane and it is performing beautifully. I look forward very much to sending these out next week, thereby reaching a milestone three years in the making.




Me, calibrating Soundplane #008. Out of all the 30 instruments, this one had the worst imperfection in the case, a small tearout in the edge of the walnut power light surround. So I used it to calibrate and refine our assembly process. It's been put together and taken apart about sixteen times. It's going to be a demo unit for the time being.



Here Chris is applying Kapton stiffeners to our sensor flex circuits and inspecting his work with a loupe.


Tomorrow from 9am-5pm or so we'll be at the MMTA (Mostly Modular Trade Association) Summer Synthfest 2012 in Seattle, showing off our first production units. The event is at the Experience Music Project in the shadow of the Space Needle. Come down and see us if you'd like to get your hands on a Soundplane. Details here. [mostlymodular.com]

Something like you describe sounds more possible. Thanks for the input.

I hear your fair request but I feel like I have too many formats to support already.

The way Rack Extensions do UI, I don't think it's even possible to make anything looking remotely like Aalto in that environment.

OK, thanks for the update. hard at work on Soundplane at the moment but I'm definitely going to look at this.

Thanks for the report, will investigate.

Please add your Mainstage version number (2.?) and OS version.

Super! Thanks for sharing.

The AU should be in the right place after install. I have not tested Mainstage 2 specifically so maybe someone else can help. I might try reinstalling Aalto after Mainstage.

THere aren't really any schematics. The version that's most competely documented is the 8x8 one in my these on the DIY page. This requires tying up an 8x8 audio interface.

Never used an Arduino myself, just a recommendation for low data-rate options.

Seriously, make the 1x2 first... even with your electronics person helper. It's a 4 hour project that could save you days of time on the full version!

What I'm saying is, this is mostly uncharted territory and you will have to do some of the experiments yourself. I tried to be informative with details in the thesis, but it's still a hard project.

Thanks for contributing-- I'm listening to all these ideas and scheming!

I'll probably do something really simple for 1.3. Longer term I think a more general scriptability will be the best way to address people's diverse desires.

On epiece of advice at the outset: Instead of jumping into an 8x8 or large, I would make a 2x2 or even 1x2 version. That way you can get a feel for some of the work involved and make mistakes that you will inevitably make, on a small scale instead of a big one. And, you may even wind up with a useful controller.

For the small sizes (1x2 or similar) you can probably use an Arduino to make the USB data and do all the work. For bigger sizes like the Soundplane, this will be too much data, so I don't really have anything simple to suggest.

Hi Says Chris,

I moved yr threads to DIY.

You can check out the G-Tar project from Medialogy lab in Denmark -- this is the only other DIY Soundplane project I know of. For whatever reason it's not a thing a lot of people have done. It has not taken off like the monome. One reason is that it's quite a bit harder-- there's a lot more data processing and a high-powered processor required.

The USB thing is no easy task either. There are so many ways you might want to do it, that I can't give an easy answer without spending a lot of my own time developing such.

I set up this part of the forums to encourage DIY folks to make Soundplanes, and I was hoping that activity by adventurous people would result in more information getting out there, and an easier starting path... but that hasn't happened. I can do some more of this myself, after we ship.

Meanwhile I'm still here willing to answer your questions.

So, the Soundplane A has a blank area on the left side, where the DSP board lives. There is some room inside the case here for a bit more sensing hardware, and we are leaving extra ADC channels available on the board for it.


So, my question to you is: should we put a few sliders up there? Or a big knob? Or something else? Or leave it blank?


The "blank" option has the advantages of keeping the costs of the Soundplane A down as much as possible, and getting us to a finished product faster. If we do add something, it will have to be done with the level of quality the rest of the product has, so it might add a significant cost in materials and assembly time. And you can get a bunch of USB sliders or knobs pretty cheaply, as you like, and set them there.


On the other hand it's nice to only bring one controller to a gig, and it's very useful to have a knob or slider that you can set to a value and leave there.


Thoughts?

That PolyKB demo is pretty cool. I have always thought their oscillators sounded good in the AKS clone etc.

Aalto avoids aliasing in oscillator FM by calculating what aliasing harmonics would appear, then backing off on the index to reduce them to a certain level. You can hear this pretty easily in that high notes have a sort of threshold of mod index after which they don't get any more gnarly. It occurs to me that the same calculations could be used to treat hard sync as a kind of instantaneous FM. I also have some papers around on hard synching sines. Stuff to try for Aalto v.2 or future products!

The sounds of Aalto's oscillator are Buchla-inspired but the algorithm design is not. Having a sine that's waveshaped into square / saw probably leads to some commonalities, but I'm not trying to emulate the circuit.

I don't know of other soft synths that take the same approach of waveshaping a sine, but there have got to be some out there. The magic is really in the details and fine tuning.

I suppose this is just one of many considerations for any prospective modification, but will the advantage of more steps be outweighed by their addition overcrowding the User Interface?

Maybe for now, working at slower tempos, automation could be used to adjust step values while the sequencer plays? This might effectively offer 64 or more steps.

Yes,thanks for the thoughts. Aalto's main goal is to keep things relatively simple, to do as much with as few UI elements as possible. I'll make future products and modules with more options.

The push to finish the first 30 Soundplanes continues, with software development at home and hardware production at the shop. Here's a cool thing: the alder instrument bodies in line for hand sanding. I'm really happy with the way the variety of grain presentations gives each one a unique personality.




Each Soundplane needs a significant amount of hand shaping work after it's milled, to get rid of marks from the planer and to smooth down the sharp edges that the milling tools leave. We'll be at this stage for a few days---time to get some new choons into the iPod. You'd think that Madrona Labs would have better speakers... well yeah, something to consider after we ship.




After the initial sanding will come three coats of finish, the applying of which means a little careful work but mostly just setting up a reasonably temperature-controlled environment and then, that most universal of chores, waiting. Everything should be dry around the time the surfaces and FFC interconnects arrive.

Hi Garf,

Mostly 1.3 will be a stability / compatibility upgrade, so I won't add this new feature, but I'll put it on the list of "features wanted by people who know what's up."

I hear you. I'll do this in the next update unless I run into some problem.

OK, you mean when the pitch of the voice (or any parameter with a patcher input really) has a different value for each voice.

This doesn't strictly depend on the sequencer though that's really the only useful way to get a lot of control.

You can send the "voice" output of the KEY module right to any patcher input, then scale it for some control. So you can play chords for example, without using the sequencer, but the intervals will be the same.

I understand - will think about it.

hmmm.

Hi,

I don't quite understand. What about the voices do you want to offset?

Hi peeps, I'm happy Aalto attracts people of all skill levels with the MIDIs, let's keep it friendly please.

Hi Fraser,

Well I'm pretty sure it's possible but I don't know GB at all so I'd just be looking through the anual same as you. One thing I can point out is that GB will be loading the Audio Units version of the plugin. So if you Google "Audio Unit Instruments Garage Band tutorial" or something I bet you can find help.

In fact I just did that and here's a helpful link.
http://www.daveclews.com/dave/Portfolio_files/GB%20Plugins.pdf

Hi Fraser,

You want "Export song to Disk" in Garage Band to make an audio file. There are a lot of options in GarageBand. I've tested that the Aalto plugin works OK but I have to leave the details of making your track up to you. I don't really know GarageBand that well so I'm sure you'll know it better than I do after a couple of hours with the manual.

I've been working in the shop with Christopher this week making the blanks for the Soundplane surfaces. Here's Chris hard at work gluing a sheet of thick latex to a fiber underlayment. (thanx Ableton for the sweatband, it's coming in very handy)




Cutting the latex sheets with the utility knife was fine for making a single prototype, but for 30 surfaces was way too slow. A gigantic shears might be the best cutting tool for this project, but getting one would be overkill for now. Meanwhile we figured out how to use the scroll saw by making a jig to hold the latex in place. It's a good thing neither of us is allergic --- latex dust goes everywhere.




After the fiber and rubber are glued together, a somewhat messy process, we get to select the area of veneer that will form the playing surface. Of course, each one can be quite different in appearance. I try to select a fairly symmetrical pattern that still has some excitement in it. Here's the first one that rolled off the assembly line.




These blanks will go to Gurian Instruments for laser-cutting and inlay, and then come back to our Ballard shop for sanding and finishing.


I'm also at work on the Soundplane client software and the Aalto update. Since there is so much to do right now, it's good that I have the software work on the one hand and the wood shop work on the other, because I can usually convince myself that the wood shop is a nice break from writing software. I'll spare you the pictures of software development but it sounds like TAKATAKATAKATAKATAK.

OK, I'll see what I can do about this for the update.

Good news, we have found good homes for the first run of 30 Soundplanes. Thus, funding is all in place and production has begun. Look for finished Soundplanes in 8-10 weeks, and more updates now that there will be cool stuff to take pictures of.


It's been a long time getting here, and I appreciate the support from everyone who's along for the ride. Whether you preordered a Soundplane or bought a copy of Aalto, you've helped make this new instrument possible.


Here's a closeup of the final prototype case during finishing. I had tried rounding off all the edges previously, and it ended up looking too blobby. By rounding off the front edge horizontally after the vertical cutout is made, a compound curve is produced that I think is a very nice intrinsic detail. The sharp edges get hand-sanded just enough so that they don't chip easily or hurt your fingers.




The road to production has generated lots of prototypes and castoff parts. You can see some of them on the shop floor here.