Deconvolving a sampled signal?

On Friday, February 3, 2017 at 4:01:53 AM UTC+8, Evgeny Filatov wrote:
> 
> Deconvolution is indeed an interesting, but also one of the more obscure 
> topics.
> 

It is not obscure in seismic processing.
You have "deterministic" decon which take the horrible airgun signature
and tries to make it shorter, more like a ricker wavelet.
And "statistical" decon, invented about 50 years ago (Robinson-Treitel)
has been used ever since.

On Thursday, February 9, 2017 at 2:17:12 AM UTC+13, mbj...@y7mail.com wrote:
> On Friday, February 3, 2017 at 4:01:53 AM UTC+8, Evgeny Filatov wrote:
> > 
> > Deconvolution is indeed an interesting, but also one of the more obscure 
> > topics.
> > 
> 
> It is not obscure in seismic processing.
> You have "deterministic" decon which take the horrible airgun signature
> and tries to make it shorter, more like a ricker wavelet.
> And "statistical" decon, invented about 50 years ago (Robinson-Treitel)
> has been used ever since.

Indeed. I even saw a neat use of it when somebody made a digital violin. They needed the impulse response of the violin (an expensive one) and use a mechanical device to ping it. Unfortunately the device itself is not a pure impulse so its transfer function had to be separated from that of the violins using deconvolution.