randy's Recent Posts
Our first product, Aalto, will be released this coming Monday the 13th of September. Exciting news! But I know Aalto's introduction has left a few people wondering what happened to the innovative hardware we announced. Let me explain.
I started Madrona Labs to make the Soundplane, developed as part of my Masters work in Computer Music at the University of Victoria, into a reliable, portable instrument. The prototype required a bulky and expensive external audio interface, and had low resolution. But it felt great to play and generated enough excitement that I took a leap of faith and started working on it full time. I luckily found a great hardware engineer who wanted to help me make a self-contained version, and borrowed money to fund six months of development.
Six months later we had a great case and mechanical design and a mostly working main board built around a fast DSP from Texas Instruments. But it was clear we wouldn't have a working prototype for at least a few more months.
I had already started working on Aalto, which I had planned to release along with the Soundplane in order for folks who aren't Max/MSP wizards to have something fun to play out of the box. So I decided to release Aalto as a standalone product in order to get some income. Supporting it very well, as I intend to do, will necessarily slow down Soundplane development, but you gotta do what you gotta do! It has been a long road, but we get ever closer to working Soundplanes. Soundplane is still the center of our business model and a driving passion of mine. I badly want to play one. Expect more news soon.
@ignatius: partly CPU efficiency, partly a few tenacious bugs. There was a stuck note happening every once in a blue moon. Hard to fix if you can only reproduce every 15 minutes! I think it's gone now.
@asquare: it's a great idea to share patches here somehow! I will look into how we can embed files here... after Monday.
Eventually, I think it would be great to share/browse patches on a server from inside the plugin.
Yes, Monday the 13th. I really should update the front page.
The work I'm finishing up, mainly optimization, took longer than I'd hoped. I have now set the 13th as the definite release date. Sorry to make you wait another week.
I do plan to do Mac VST quite soon. The decision to do AU only at first came about because of patch management. Hosts like Live, for example, have very usable management for AU patches. But once I started putting the VST version together I realized that I needed to do my own patch mgmt. to make it something I would want to use. And this is a project that could take a little while. So, AU version now, AU/VST update with integrated patch management in ... well, a month would be my guess.
And I'll add your vote to the list of people who want RTAS. I will do this eventually if enough customers want it.
It won't be too far from $100.
Thanks for the support! Currently I am planning for an AU release on Sept 7. It could always slip a bit if something goes wrong but right now it's looking good.
If you're a synth head, it's probably clear that Aalto is inspired by Don Buchla's work. I think that Buchla's 100 and 200 series designs are high water marks of what can be achieved by focusing on the experience of playing rather than on the technology. Design not from a feature set out, but from the knobs in...
Above you see the 259 Complex Oscillator, which is the closest cousin to Aalto's algorithm from among the Buchla modules. If you want 259 sounds, of course you need to get a 259. It is more complex than my oscillator, both in sound palette and interface. What I tried to do was capture the qualities of its sound that made my ears particularly happy, where that was technically possible in a real-time algorithm. I also took cues from how Buchla's designs do more with less, and distilled the interaction down to make a few knobs that did a lot.
For example, having only one scale control per signal input on Aalto (except for the oscillator pitch) was a conscious decision I made with simplicity in mind. It seems to cover around 90% of the sound making possibilities (designers might call them "use cases") with 50% of the UI.
Who else besides musicians needs precise and intuitive UI? Airplane pilots, that's who. These gauges were another inspiration. Note that when you look at a number, your gaze is also taking in the filled-in area of the dial, not darting back and forth to some number box across the screen. I'm assuming these designers knew what they were doing.
Another musical device that works really well is the Weiss DS1. Note all the breathing room around the knobs. Simple, uncluttered and a joy to use. That it sounds phenomenal doesn't hurt.
Check out the Tone Board for the Yamaha GX1. If you spend your year's development budget making a $60,000 synth, you can put some time into UI. See the clear layout of signal flow and how the knobs connect to it and are grouped by the panel graphics. Also, precise markings in seconds on the sliders as opposed to an arbitrary range of 1-10, or no range at all. If only those drawings of envelope shapes were able to move... you'd have something a lot like Aalto's envelopes. It's no coincidence.
In my view the best way to respect the innovators is to innovate. So instead of trying to reproduce some elusive and hyped "analog magic" I am combining all of the influences above and more into the best design I can make for today's technology. I look forward to sharing this work soon.
coding coding coding
Nope, just as an AU and VST plugin. If you just want to play it live, you can find a free host program that will host the plugin.
The beta is closed now. We'll do one for Windows in fall so stay tuned if that's your platform.
:-(
@soto,
I have thought of it. I'm going to wait and see. I'm using the JUCE framework which supposedly allows compiling on Linux, though I'm not sure anyone has tried making a plugin with it on Linux.
The thing I'm most concerned about is support. When I sell a copy of the software to someone I'm making a personal commitment to help make it work for them on any system I support. I'm not sure that I could do this well enough on Linux because of the many versions out there and my lack of experience with them.
So it's not a high priority but I'm interested and appreciate your offer. If you have any ideas about the support thing let me know.
@asquare: You actually hit it right on the head. Aalto is good at making complicated sounds easy to program. It's also quite heavy on the CPU because I'm making sound quality absolutely top priority, and because it's so patchable. And finally, it's more oriented towards helping people make the next Silver Apples than the next Howard Shore soundtrack. So for all these reasons I think topping out at four voices would be fine.
Hi kickprod, thanks for the good words. This summer for OS X, hopefully fall for Windows.
Hi there!
Here's how I think it will work. One purchase gives you a copy for every platform we release it for when available. Your license will let you use one copy at a time.
There won't be copy protection because I think that if you buy software, you should be able to make copies of it. At runtime, however, the software may check to see that you are holding up your end of the license agreement.
Thanks so much!
Aalto will only have the one combination of modules. I wanted it to be something that you get your head around quickly and start making sounds with. I wouldn't rule out doing a configurable modular ... later.
Here's a little movie that shows off some of Aalto's sounds, UI and charming personality. Enjoy!
We are shipping for the Mac this summer, Windows to follow shortly.
Price will be comparable to other high-quality softsynths out there. It will also depend on whether this version is released as a poly synth, which I haven't decided on yet. I'm working to get the CPU usage down to the point where that is reasonable... wish me good speed!
Microtuning most definitely. There will be some kind of aftertouch but I'm not sure about the details yet.
cbm: thanks and well put. Aalto is definitely inspired by Buchla's work. The Music Easel was a model for me in how good design could give a small collection of modules a huge variety of sonic possibilities.
I also like how the Easel lends itself to programming distributions of events, not just single notes. I tried to make Aalto's sequencer flexible enough to offer a lot of possibilities here.
And then there are the woodblock and 259-like sounds that are new sonic territory for plugins. My goal is not to emulate hardware but to make new sounds that satisfy me in the ways the original ones do. In return for all the inspiration I take pains to point out that, if you want to make Buchla sounds, you gotta get a Buchla!
I'm not sure. I am planning to do a closed beta, but may invite a few more people. Rabid followers of the list for example... :-)
Aalto is a new synthesizer coming from Madrona Labs this summer. It looks like this:
Aalto can make many different kinds of sounds, including some that have been difficult or impossible to do with softsynths until now.
Here is a bouncing-ball-style patch made with FM:
Aalto FM Bounce by Randy AKA Caro
Here's a clip demonstrating the timbre and waveshape knobs of the complex oscillator.
Aalto Timbre+Shape by Randy AKA Caro
MP3s do not compress single-oscillator sounds very well---to hear the full-quality sound, use the download links on the Soundcloud players.
Aalto came about because I wanted Soundplane owners to have a rich and expressive sound making tool right out of the box. I designed a small synth that would be easy to learn, easy to use, rich in possibilities, and tightly connected to the controller. Though we still have some work to do on the controller, the synth is nearly done. So we are releasing a MIDI-controlled version of Aalto as an AU and VST plugin this summer.
There's a lot more to say about Aalto, so please stay tuned for more info including a demo movie very soon. If you have questions or suggestions, please visit our forums. I'll post some more info there to get things started.
The Soundplane remains the reason we are doing all of this, and we continue to make progress on it. Brian has recently shifted gears from hardware design to firmware programming. Please check back or subscribe via RSS for an update on the Soundplane A soon!
We are aiming for a price of under $2000 here in the USA. It is possible it will be significantly under, but it's too early to say for sure.
@walker: thanks for sharing your excitement-- we are thinking very much the same, and will have some new hardware parts to show off soon, please stay tuned.
hi David, the Soundplane A will have only a USB connection that transmits the raw sensor data. So you'll need to use a computer with it, running our translator software to make OSC and MIDI messages. We are also working on software synths that will listen directly to the Soundplane data.
This software is being written with the help of a cross-platform framework, JUCE. So while we are developing on Mac and may release there first, we want to support Mac, Windows and Linux ASAP.
good question, I added it to the FAQ, if you have any more, please ask.
Spring is here, and as we get closer to having finished prototypes to show off I am going to be posting updates more frequently. I'm working on two main tasks right now, one of which is finishing the Soundplane enclosure.
We have two prototype enclosures milled by O. B. Williams, a local company that mainly does architectural millwork. They have been around since 1890, which makes them one of Seattle's oldest companies. As you can imagine, they know a lot about wood there, about what species to use for different projects, and how to make things that will look great and last. When I visit them I always learn something new.
CNC machines like they are using have made a lot of new designs possible in wood. Look at the beautiful new bikes from Renovo for example. I think we will see lots more innovative products based on this intersection of natural materials and computer-aided manufacturing, and I'm very happy to be making one.
As you can see above, the Soundplane body is milled out of a solid block of wood. A difference from the first prototype is that pegs are left in to hold the circuit board, which means no more messing with plastic standoffs, less assembly time, and lower costs in the long run. There is a back panel (not shown, boring) with a rabbet cut into it that fits snugly into the piece shown. Then the two are screwed together to make a very solid monocoque structure.
As you can see on the front, the unmilled block is actually two blocks glued together down the middle. Why? Two main reasons. The first is mechanical: the composite is less prone to warping than a solid piece of wood. The glue joint is supposed to be stronger than the wood itself, something I look forward to verifying with destructive testing when we are done with our earlier prototype.
The second reason is sustainability: using 4" instead of 8" boards means that a much younger tree can be used. And young alder trees grow very fast. Alder itself has other benefits, including being beneficial for other trees in mixed stands. Find out more courtesy the US Forest Service. Time was, alder was considered a weed tree in our Pacific Northwest, and used for fuel. But for the above and other reasons its popularity as a craft wood is growing. I love its soft glow with a clear oil finish, as shown here, and I hope you'll agree.
Coming soon: more construction details, and our first software synthesizer.
dondo:
A computer will be needed to use the Soundplane A when it is released. One of our future products may be a Soundplane->CV converter module.
Thanks for the Inkscape link! I didn't know there was something like that.
Check email for driver.
This is a post about some extreme DIY shenanigans for general interest and later reference.
We are getting close to complete prototypes now and finishing up some of the smaller details. One of the parts I'm finishing is the back panel that holds the USB jack. This panel will also serve as the nameplate.
Frontpanelexpress.com has a very inexpensive service for milling aluminum plates. One catch is that you have to make your design using their software which is clunky and Windows-only.
We are, to put it mildly, detail-oriented, and the nameplate logo has to look good, right? So we need our own typography on the plate. The only way to get a custom graphic into their software is to import an HPGL file with the engraver paths. I do my design work on the Mac in Adobe Illustrator, a program I've been using for about 20 years, and am very comfortable with.
So the problem is: get an Illustrator outline from my Mac over to an HPGL file on Windows XP. Piece of cake.
The first way I tried was exporting a PDF, then opening it in the Windows shareware ViewCompanion Premium, which can export HPGL. I tried this, and the resulting HPGL did not import into Front Panel Designer. It showed up as an empty drawing with a size of 0.0394" x 0.0394".
The next way I tried was by installing a driver for an HP pen plotter on Windows, and printing the PDF to a file using this driver and Adobe Reader. Tracking down the right driver was somewhat hard. Then more disappointment: the same empty drawing resulted.
At this point I was just about ready to give up and pay Frontpanelexpress to do the design work as a custom job. I couldn't think of another route to making the HPGL file that made any sense. As a shot in the dark I searched for "print to HPGL." That turned up a DIY page from 1997 with a reference to a Roland plotter driver working on XP. The link to the driver itself was gone, but the Wayback Machine had a link to the old Roland page with, miraculously, the XP driver. And so I installed it and it worked just like that.