September 21, 2013

The Next Episode...

by Chris Randall

The first production run of Audio Damage modules is officially sold out, at least as far as we're concerned. We've sold the entire production run to Analogue Haven, Alex4 (which is a European distro company started by SchneidersBuero), and EquionoxOz. We've pulled the trigger on the second production run, and will have more backplanes in early November. When this second run shows up, we're going to introduce at least one more new module. Possibly two, but not likely.

Simultaneously, we've begun development on a more powerful CPU/RAM backpack that will enable us to make much more sophisticated modules (think a pretty much full-featured Replicant or Dr. Device in hardware). This will take some months, as these things go, and we'll discuss the feature set here and get some input. We made every attempt, and were largely successful, at keeping the initial Euro release secret. But now that the cat's out of the bag, we'll be pretty open about what's upcoming, unless it's a case where spilling the beans would mean opening ourselves up to competition. However, something like Replicant is an advanced case study in DSP in a Euro context, and is difficult to pull of without starting with an existing DSP library tailored to the task, so I'm not too worried about that. If someone else could have done it, they already would have, since it's such an obvious effect for the Euro environment.

Now, this is not to say that we'll only be releasing DSP-based audio effects. We have plans for several different things, now that we have a WORKFLOW, and finally I get to my point, such as it is. Adam is on the way to Japan, and I'm in charge, and when that happens, my imagination runs wild. (Every time he gets back from Japan, I spend the first three days interrupting his tales of travel to tell him about The Next Big Thing I thought of while he was gone.) One of the things I'd like to think about is a desktop synth.

Now, there is no shortage of desktop synths. One could make the argument that market was all full up. But one could (and in fact did) make the argument that all DAWs come with a delay, so why make one? We built a pretty nice business on that concept. So...

Now, this wouldn't have to be a digital synth (like, say, the MeeBlip -> Blofeld range of devices.) We have strong analog chops up in here. My question is this: if you were presented with a new desktop, a little guy, what would you like to see that you haven't seen before?


Page 9 of 11

Sep.27.2013 @ 11:27 AM
"...all the "high-resolution, optical encoders" in the world are just so much artisinal pickles otherwise."

Fucking gold.

Sep.27.2013 @ 11:39 AM
The Aleph looks cool, but the DSP they are using (Analog Devices Blackfin) seems like a weird choice. I programmed it for a few years while at Analog Devices, but there were also PhDs around to do the heavy lifting - and there was a LOT of "heavy lifting" involved in getting stuff running on that DSP. Creating an "open platform" is a noble goal, but I doubt that many open source folks will want to tackle getting stuff running on a 16-bit integer DSP with a very obscure syntax.

Sep.27.2013 @ 1:32 PM
If the bees "modular control environment" makes Aleph into anything like a Nord Modular, I can see it going places. If not, then I'm not sure what the point is. I've never worked with the Blackfin, but I've worked in enough different development environments that I'm sure I could if sufficiently motivated. But I make music to get away from software engineering, not to do more of it. Expecting there to be a large population of musicians who are also embedded developers seems like a tall order.

Back on topic: I hope Chris and Adam continue to build stuff that sells really well. And then sneak in some firmware that can take the hardware far into left field.

Sep.27.2013 @ 3:21 PM
Chris Randall
That's the exact same Blackfin DSP chip that is in the OP-1. Interestingly, the Teenage Engineering guys do the UI stuff on the same chip, which I don't believe is the case with the aleph.

But that should give you an idea of what it's capable of, in a highly optimized setting. I agree with Sean, though. It is no joke programming that thing. Nothing like booting up your little Eclipse derivative and pasting some code you found somewhere in to your Arduino. It's an interesting choice, I'll give them that. One reason we chose the STM32 series for our foray in to hardware DSP is that you can program it in C/C++ on a freely available IDE. I don't know how they've worked it out, because to the best of my knowledge, a Blackfin requires VDSP++, which clocks in at $4200 a seat. If there's some other programming solution for Blackfin or SHARC, I would love to hear about it.

Also, no decimal points, as Sean pointed out. Which is, it must be said, something of a bummer.

Well, monome products are the very definition of the Vleben good; for those that like that sort of thing, this is the sort of thing they'll like. It is a very powerful and well-designed box, there's no denying it.


Sep.27.2013 @ 3:54 PM
Chris Randall
Actually, a quick Google search shows that there are about a dozen Blackfin toolchains, including some freebies. So never mind that $4200/seat stuff.


Sep.27.2013 @ 6:25 PM
I was doing some research into the performance of open source Blackfin compilers until I got sick of it. Which didn't take very long.

Even with free tools, programming the Blackfin would be non-trivial. Straight C code would compile into assembly code with hilariously bad performance. You can use C intrinsics to call optimized assembly code (at least in VisualDSP++ - not sure about the free compilers), but this is going to be fixed point, and probably double precision for decent sound quality.

To get the best performance out of the Blackfin, you have to "think like the Blackfin." I was able to do this at one point in my life, but not because it was enjoyable or Zen or anything like that. I was doing that work to feed my family. Take away the survival motivation, and replace it with "hey, here's a cool open source platform I can develop audio code on without getting paid"...honestly, it is a tough sell.

Sep.27.2013 @ 6:30 PM
"just a fetish object for the cult"

Alas, my fate in life. Wearying, so wearying...

Sep.27.2013 @ 6:34 PM
Chris Randall
Sean, don't take this the wrong way, but you still think like a Blackfin.


Sep.27.2013 @ 6:37 PM

Sep.28.2013 @ 4:05 AM
I'm glad boobs mentioned the Aleph, and fwiw all the Scope stuff and the Solaris also run on AD Sharcs...but I would imagine keeping your platform compatible with your existing code (ie, more 32FLT support than SHARCS tends to offer) is probably ideal when it comes to implementing this since you can leverage your existing work.

On that note, let's first consider your simpler effects: I would think vapor, fluid & liquid can be done within the same basic code. And it stands to reason you can come up with something that leverages the modulation&sequencing elements & sound shaping potential of Ronin/Bigseq/bitcom/Filterstation/replicant... With a reverb unit based on parts of EOS/Mangleverb/ratshack/Adverb and a delay section drawing on....,

At that point I think any sound source you put before that will still sell. And again going with your existing palette on the plugin front...

There's no reason you can't invest in analog sections too, but I really appreciate how you've gone from a few highly capable effects with a vet WIDE range and numerous knobs to very well thought out UIs and exposed control sets over the years. Again taking this and applying it to your desktop hardware, and you should in my opinion wind up with a very focused device that can go from fast transient blips & bursts to sequenced mangled madness...

In A Way That Makes Immediate Musical Sense when programming & performing, basically.

Page 9 of 11



Sorry, commenting is closed for this blog entry.