[sdiy] Hardware convolution box?

[Bruno Afonso]

I believe monome picked a blackfin for their aleph because you could use
open source tools to develop for it. It.may be only DSP platform where
that's the case but correct me if I'm wrong.

Sadly it's now effectively abandonware as far as hardware goes but there is
a chance someone may make some new ones. Dev on the software side was never
strong but some people are still pushing things forward. Sadly, convolution
has not been implemented but as far I can see you could do it .

The whole fixed point programing never really clicked for me but I'm
guessing it's just a matter of spending more time on it.

Axoloti could be a potential solution but there's no power for long impulse
responses.

On Thu, Feb 9, 2017, 09:26 Terry Shultz <thx1138 at earthlink.net> wrote:

> Hi Folks,
>
> Stand alone DSP’s are getting fewer intro’s as they are being added to ARM
> processors and hybrids.
>
> Analog Devices are still increasing their offering and a small amount from
> TI.
>
> Qualcomm has a Hexagon DSP platform in the Snapdragon family of processors.
>
> ARM by itself is not a replacement for DSP, but at some point it may be
> difficult to fine a standalone DSP.
>
> Cellular and Automotive drove DSP volumes but Multicore ARM are more
> attractive for those customers.
>
> Xylinx has ARM based products that are interesting and a few other FPGA
> devices are following in this path.
>
> Analog Devices has about the best offering for Audio applications over all
> . Tools are pretty good and development platforms are
> not too pricey,
>
> I too have looked at XMOS as a usefull tool, but I still look at the Sharc
> as a good platform.
>
> Having been the Core member of Motorola / Freescale audio DSP products, I
> have to abandon these over sustainable newer products.
>
> C compilers are generally ineffective for Fixed point devices, but
> Floating Pt. Devices are generally a better fit for C programming.
>
> I tend to still write in assembly to get the max performance and lower
> latency.
>
> Interesting to note that the AKM audio devices are now using embedded DSP
> in their designs and have tools to support programming.
>
> As many of you are not interested in ARM / DSP I apologize in advance. I
> prefer to go back to my roots in analog these days and build back at the
> transistor level my audio projects.
>
> Perhaps there is a balance point, Convolution / reverb is a different case
> and I prefer fixed point 24 bit DSP’s as I have over 20 years of experience
> in the DSP56k family of devices.
>
> just my long winded 2 cents.
>
> regards,
>
> Terry
> > On Feb 8, 2017, at 9:53 PM, rsdio at audiobanshee.com wrote:
> >
> > I LOVE XMOS, but it's not going to be efficient enough to handle long
> convolution.
> >
> > XMOS is great when you need to be ultra-responsive to inputs like
> buttons or other I/O, but only if there isn't a great deal of processing.
> Convolution is going to be more cycle efficient with a DSP.
> >
> > Brian
> >
> >
> > On Feb 7, 2017, at 7:26 AM, Mikko Helin <maohelin at gmail.com> wrote:
> >> This XMOS board designed (for stompbox use mostly) by an XMOS employee
> >> might be a good alternative for DSP chips and better than generic
> >> ARM's:
> >>
> >> http://www.diystompboxes.com/smfforum/index.php?topic=114354.0
> >>
> >
> > _______________________________________________
> > Synth-diy mailing list
> > Synth-diy at synth-diy.org
> > http://synth-diy.org/mailman/listinfo/synth-diy
>
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20170209/be3ab235/attachment.htm>