[sdiy] Axoloti

rsdio at audiobanshee.com rsdio at audiobanshee.com
Tue Dec 15 07:14:51 CET 2015

On Dec 14, 2015, at 5:04 PM, Ian Fritz <ijfritz at comcast.net> wrote:
> Johannes sounded pretty discouraging as far as putting more sound sources and waveshapers into Axoloti.  Not sure why -- I had the impression that all the anti-aliasing issues had been solved ages ago. :-)
Solved, yes, but not without a cost. The processing is not free. You'll always have more voices if you allow aliasing, and fewer voices if you don't allow it (or minimize it). I think Johannes was merely pointing out that it's not something you can ignore. Besides, not all algorithms are perfect at eliminating the aliasing, they just reduce it to a level where it might be less objectionable.

> I wish I was more proficient at DSP programming -- I've learned all the theory, of course (not exactly quantum electrodynamics), But all I've actually made so far is some simple FPGA experiments.
There are plenty of purpose-built DSP chips that save you from experimenting with FPGA designs. Why re-invent the wheel when you can take advantage of decades of experience behind products like the Texas Instruments TMS320 family? There is also the Analog Devices SHARC family. Either of those families offer plenty of options for synth designers who don't want to lose any time creating an FPGA-level design just for some basic building blocks.

Texas Instruments even has $50 platforms that include DAC for audio and headphone outputs, so you can get started on synth algorithms without designing any hardware.

There are open-source libraries for the TMS320 that would give you an example to learn hard-code DSP programming. At least that's how I got my start for real (after flailing a bit with the Motorola 56000 so many years ago).

Brian Willoughby
Sound Consulting

More information about the Synth-diy mailing list