[hpsdr] PowerSDR/OpenHPSDR mRX PS v3.2.28 released

[Roger Rehr W3SZ]

Thanks for a great release, Doug and Warren!

The SNB in particular sounds like a great addition to PowerSDR.

Leif SM5BSZ, author of Linrad, has a long-standing interest in noise 
reduction.  He would be very interested in wide-band recordings of noise 
that are effectively dealt with by the SNB.

If anyone using the new version of PowerSDR encounters such noise, 
please send a wideband recording to Leif, or if you prefer to me, and I 
will forward to him.

If you also send a recording of the captured audio after processing, 
both with and without the SNB, so much the better.

But the wide band recording is the primary interest, as it can be used 
to compare the effectiveness of various noise reduction schemes.

Leif's email is leif@ his call [lower case] dot com.

Thanks in advance  for this help, and thanks again guys for this great 
new release!

73,

Roger Rehr
W3SZ

On 10/3/2015 10:54 PM, Doug W5WC wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> All,
>
> PowerSDR/OpenHPSDR mRX PS v3.2.28 has been released.
>
> This release can be downloaded from the openhpsdr.org website.
> http://openhpsdr.org/download.php
>
> This release contains the following changes:
>
> Radio Astronomy (RA) Joe, K5SO
> - added Rx2 to the RA feature to allow both Rx1 and Rx2 data to be
> simultaneously collected/viewed
> - added a new control to the RA display to allow the selection among options
> of
> 	1) simultaneous display of Rx1 data and Rx2 data,
> 	2) display of Rx1 data only, or
> 	3) display of Rx2 data only
> - added Rx2 data to the read/write-to-disk option from the RA window; data
> format for each line of the .csv file is
> 		time(secs), Rx1(dBm), Rx2(dBm)
> 	regardless of what option is selected for display in the RA window
> - added a compiler directive to try to resolve the EU vs US formatting issue
> of the previous RA read/write to .csv files
>
> Spectral Noise Blanker (SNB)
>
> Following the release of new Noise Reduction technology, NR2, earlier this
> year, this release includes another new tool for your "Noise
> Toolbox":  SNB, a Spectral Noise Blanker.  There are many different types of
> noise/interference and so having an assortment of tools at your disposal is
> of great benefit.  The Spectral Noise Blanker is of a very different type
> than our Wideband blankers, NB and NB2, and will address some additional
> impulse noise issues.  It can be used along with NB/NB2 when beneficial.  It
> also plays quite nicely with NR/NR2 to address both Impulse Noise (SNB) and
> Random Noise (NR/NR2) simultaneously.
>
> While NB/NB2 operate on wideband spectrum data and impact both the
> panadapter display and the S-meter, SNB operates differently and impacts
> only audio.  SNB uses Linear Predictive Coding (LPC) to evaluate the
> spectral formants of its input.  It then uses that spectral information to
> identify impulses and to synthesize replacement samples for the signal in
> regions where it has been corrupted by those impulses.  Yes, it does not
> just zero or interpolate across corrupted samples, it synthesizes
> approximations of the original signal using the spectral information and the
> samples before and after the impulse.  This technology largely does not
> depend upon amplitude to identify impulses; therefore it can often detect
> impulses that are equal to or smaller than the amplitude of the signal and
> can clean signals when other blankers fail.  Computing replacement samples
> requires quite a bit of matrix computation; so, to keep compute requirements
> within reasonable limits, SNB currently restricts audio bandwidth to roughly
> 300Hz to 5400Hz.
> This was a design tradeoff which will be revisited as faster processors
> become available.
>
> To turn this blanker on, simply click the "SNB" button on the main console.
> (Note that NB and NB2 have now been moved to the same
> button.)  In Setup, there are a couple threshold adjustments, "Threshold 1"
> and "Threshold 2".  However, the algorithm is self-adjusting and you
> probably will not be able to achieve much improvement by fine tuning these.
>
> Summary of your current Noise-Interference Toolbox:
>
> Impulse Noise
> * NB (Wideband)
> * NB2 (Wideband)
> * SNB (Spectral)
>
> Random Noise
> * NR (Variable-Leak LMS Algorithm)
> * NR2 (Signal/Noise Power Density Algorithm)
>
> Carriers & Similar
> * ANF (Automatic Notch Filter - Variable-Leak LMS)
>
> Noise / Interference Null
> * Diversity Reception [Requires a radio model with two ADCs]
>
> Peaking Filters
> * Audio Peaking Filter [for CW]
> * Dolly Filter [for RTTY]
>
> Expect future additions!
>
> Various Bug Fixes and features
> - added the ability to split the open collector outputs between VFOA (1-4)
> and VFOB (5-7). To use, click on the "4x3 Split" checkbox on the
> Setup=>General=>(Penny/Hermes/ANAN Ctrl) page.
> - added a 'mask' field to the static IP address feature. This allows using
> the correct broadcast address for the specific port the radio is on. This
> corrected a problem when using a static ip address when multiple adapters
> are present. See the Setup=>General=>Hardware Config page.
> - added multilingual for ToolTip comments
> - fixed the 6m panadapter display reading for the Anan-10E
> - fixed a problem with PowerSDR crashing on startup when the Disable PA for
> HF/VHF is enabled
> - fixed an intermittent issue with diversity
>
>
> You WILL need to reset your database.
> If you have release v3.2.28 dated prior to 3 Oct 2015 you will need to
> manually uninstall it before installing this release. The installer will not
> automatically uninstall it for you. This only applies to the pre-release of
> v3.2.28 and not prior version numbers.
>
> Thanks & 73,
>
> Doug, W5WC
> Warren, NR0V
>
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help: http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/