[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: DSP Speaker




In the DttSP software, the sdr core for PowerSDR (Flex Radio) and 
others,  we have a block LMS approach to automatic noise reduction.  It 
does a really good job of doing noise reduction and automatic notch 
(your choice) and it definitely does not introduce nasty sounding 
distortion.  It is much more stable than standard LMS based automatic 
notch filtering.    When I developed the algorithm,  I was surprised 
that others had not done it.  I then learned of some subtleties which do 
need mitigation but were pretty straightforward.  You adjust the filter 
once per block rather than once per sample, so you must be careful not 
to introduce discontinues (distortion) at the block boundaries.  Once 
handled,  this is the only way this should be done.

That code is available for svn download (visit the Flex Radio web site 
or the dttsp-linux yahoo group).

Bob
N4HY


Tony Langdon wrote:
> At 12:04 PM 4/12/2007, Roger Kolakowski wrote:
>   
>> And the winning answer was....?
>>     
>
> That one's got me curious too.  I'm particularly fussy about DSP 
> noise reduction, and there are very few that I actually like.  The 
> one I use is made by Michels Engineering in Germany, which provides 
> an excellent degree of noise reduction and doesn't distort the audio 
> as severely as many others.  For me, that distortion interferes with 
> my ability to pull a signal out of the mud, and most of the DSP noise 
> reductions work best for me in the "off" position, as far as 
> readability goes.  The Michaels one, OTOH, has no impact on 
> readability at worst, and often a modest improvement, as well as 
> excellent noise reduction.  On very weak FM signals (ideal for SO-50 
> :) ), the readability is dramatically improved, and the DSP can 
> almost act as an FM voice activated mute.
>
> 73 de VK3JED
> http://vkradio.com
>
> _______________________________________________
> Sent via AMSAT-BB@amsat.org. Opinions expressed are those of the author.
> Not an AMSAT-NA member? Join now to support the amateur satellite program!
> Subscription settings: http://amsat.org/mailman/listinfo/amsat-bb
>
>   


-- 
Robert W. McGwier, Ph.D.
Center for Communications Research
805 Bunn Drive
Princeton, NJ 08540
(609)-924-4600
(sig required by employer)


_______________________________________________
Sent via AMSAT-BB@amsat.org. Opinions expressed are those of the author.
Not an AMSAT-NA member? Join now to support the amateur satellite program!
Subscription settings: http://amsat.org/mailman/listinfo/amsat-bb



AMSAT Top AMSAT Home