[sdiy] Hardware convolution box?

[Corey K]

I wrote a blog post on FFT convolution recently that is perhaps
relevant:
http://blog.reverberate.ca/post/zero-latency-convolution/

-Corey


On Feb 14, 2017 8:33 AM, <rsdio at audiobanshee.com> wrote:

Fair question. Convolution is very specialized, and it can't do everything,
but it's still a great tool to have in the arsenal.

One important thing about convolution is that it can't handle non-linear
aspects of amplifiers or other sound systems that we might desire to model.
But the good news is that a hardware convolution box doesn't have to be the
only thing that you have in the rack. In fact, once a DSP platform capable
of handling long impulse responses is available, it can also be
reprogrammed for other DSP effects besides convolution. I might even be
possible to combine effects. Of course, the longest impulse responses won't
leave any cycles for other effects, so those options would have to be
turned off for the convolution to complete. But for shorter impulse
responses they're surely be room in the code memory for some non-linear
tube simulation, transistor modeling, or other processes that convolution
can't handle.

Brian


On Feb 13, 2017, at 8:34 AM, Frédéric (Opensource) <marzacdev at gmail.com>
wrote:
> Hi everyone,
>
> I see you discussing this topic (the convolution box) for quite
> a while now, but my stupid and naive question is, what do you
> need this system for?
>
> I mean, is it only for reverb or rooms simulations?
> Or do you want to model some amps or other gear (like the
> Kemper Profiling thing does)?
>
> Because I personally see convolution approaches as very
> inefficient methods (in terms of computing resources) when it
> comes  to model a certain system response.
>
> Musically,
> Frédéric
>
> PS: Not trolling the list, just curious.
>

_______________________________________________
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/20170214/3d3d197b/attachment.htm>