Customize RX filters (right click & configure)

David Decoons, wo2x

Hi All,

I was talking to the guys from Flex this weekend at the ARRL Centennial convention the subject was brought up about future ability to be able to right click RX filters per mode and customize and save them. PowerSDR had this feature. For example, edit the 1.6K button, rename it to 2.0K and change the low cut and high cut to 300/2300 Hz.

I'd like to see how much interest there is in this ability. I know I would like to have different filter buttons than the default six buttons.

Dave, wo2x

Ken - NM9P

I LOVED that feature in PowerSDR!  I hope this is included in the ever-expanding persistence in SmartSDR.

np2g

Yep that would be perfect Both low and high cut accessable with memory . 

W9OY

That was a suggestion of mine from the early days of the SDR-1000 and I think it was implemented by Eric.  I always wanted to have one spot dedicated to the narrowest bandwidth possible.  

73 W9OY

Charles - K5UA

This is a highly desirable feature, especially for contesting and CW operation. Another nice touch for the CW ops would be to save an IF Shift offset along with the high cut/low cut frequencies.

Ned K1NJ

  Thanks for that, Lee.  I made custom filters for cw and each digital mode
that I used.

Ned,  K1nj

G4BIM

Great suggestion Dave, looking forward to seeing that on later versions

Peter G4BIM

W4NTG

AMEN!

Al_NN4ZZ

Hi Charles,
On CW, I usually set the filter to 100 unless there is another signal very close and then 50 may be be better.  I tried some different settings even narrower (30 and 10) but didn't see or hear any improvement.   I also tried offsetting the filter but still didn't see any difference.   The 50 hz filter on the 6700 seems to do the job.  Maybe I just couldn't find a situation (i.e. contesting simplex pileup) where it would help.  Or maybe I'm not setting it correctly.  What has worked for you?

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com




In the snapshots below there aren't any close by signals, they are just to show the filter settings.  


 

Charles - K5UA

Hi Al,

A small IF shift below the center frequency allows me to adjust the "pitch" of the background band noise so that it is in contrast to the pitch I have set by the PITCH control in the P/CW panel.  This is a throw-back to the passband tuning days of the R4C where the passband tuning knob could be "swished" higher or lower.  Some ops prefer a high pitch background noise, but this is very fatiguing to my ears.  I like to listen to a low "growl" of the background with the CW signal standing out in contrast at around 400-450 hz.  Adjusting the IF shift becomes less and less effective (or even noticeable) as the filter width drops below 250 hz because the IF bandwidth itself is so narrow that the contrast between the pitch of the background noise and the CW note becomes negligible.

By the way, what program are you using for your screen captures?

Al_NN4ZZ

Charles,
Ok I'll play with it some more.   Agree, I don't like the high pitched background noise either.   I just couldn't get any noticeable change so far but will keep playing with it.   I've found the RX equalizer helps too.  Here is my CW setting.



I'm expecting the improved NR and APF will further reduce the background noise.   Just have to wait a while to see when we will get those DSP improvements and how they work. . 

I've been using HyperSnap for screen captures for a long time, probably over 15 years.  I first used it at work to help users provide better information when reporting errors.  We bought an enterprise license and it was worth it in time savings alone.   It has a lot of nice features.  They also have a nice tool for capturing video. 

http://www.hyperionics.com/

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com


Here is a snap of the Hypersnap screen and some of the HELP.  

Ned K1NJ

  Hmm...  I can't seem to get the filter zoom below 50 Hz.  I'm using version 1.2.1
(one version behind) and a 6500.
  Am I version limited or is the 6500 feature limited?

Ned,  K1NJ

Al_NN4ZZ

Hi Ned,
I wonder if the filter is really going down to 30 HZ and 10 HZ.   Although that is what is displayed, I couldn't hear any difference compared to the 50 HZ filter.    

The 6700 has that extra level of panadapter zoom that is not available on the 6300 or 6500. Notice the width of the YELLOW line on 50 HZ snapshot and the frequency values at the bottom.    

If the filter is not really that narrow, it could be a display anomaly.   Maybe Eric or Steve can comment.

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com.  

Stan VA7NF

At 10Hz the slope of the filter sides is the main factor.  On a solid carrier you can see the signal strength decrease as you pass the carrier out of the bandwidth. 

Oh the joy of 1Hz tuning and a 10Hz filter.  Quite some rig to have such control.

Now I notice the keying modulation on a CW signal (why I said a solid carrier above because the keying sideband affects the signal strength as I tune across an sending CW signal)

Paul Christensen, W9AC

> "I like to listen to a low "growl" of the background with the CW signal standing out in contrast at around 400-450 hz.  Adjusting the IF shift becomes less and less effective (or even noticeable) as the filter width drops below 250 hz because the IF bandwidth itself is so narrow that the contrast between the pitch of the background noise and the CW note becomes negligible."

Charles,

I do the same here for CW widths that exceed about 500 Hz. On larger CW widths, I always pull the lower skirt to the CW offset frequency -- in my case it's 600 Hz.  This ensures "single signal" reception so that close-in signals on opposite side of the center frequency are fully suppressed.  Also, by setting the lower skirt to the offset, the full CW audio background is maintained. (i.e., the growl you mentioned). 

Paul, W9AC 

Al_NN4ZZ

Ned & Stan,
I did some more experimenting and made a short video.  The setup is 2 slices with one set for a 50 HZ filter and the other set for a 10 HZ filter.    There is a carrier at 14.016.8

First - the 50 HZ slice will move up and over the carrier so you can hear when the signal first is heard and when it stops being heard.  It then rests at a frequency a little below where you first hear the carrier and the sound is muted.

Second - the 10 HZ filter is un-muted and moved up and over the carrier.  You will notice although the BLUE filter line is much narrower, the sound starts and stops at the same locations as the 50 HZ filter.  

Does this support the theory that the 10 HZ filter is still 50 HZ wide?  And that only the display is 10 HZ wide?    Thoughts?

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com


https://www.youtube.com/watch?v=M8PngDo6s9E

Ken - NM9P

I have no test data to prove it, but I have observed a difference in response on my 6500 when dragging CW filters narrower than 50. (Usually 25-30) But I only find it usefull on noisy bands like 160/80 under certain conditions. Usually 50 is enough. And I also use the RXEQ with 500 all the way up and all the rest at bottom. I use a 500 Hz side tone so don't need the 1k slider up at all. APF is helpful, but still needs a little bit of tweaking.

Stan VA7NF

Watch the S meter closely.  In both cases it rises as you approach the carrier, holds steady across the blue, drops as you move away from the carrier.  From much earlier conversations you can view the skirts of the filter and watch the AGC function.

The filter top (blue portion for 10 or 50Hz) is flat and the skirts are the same on either pass.

I am impressed with how well this filter functions.  Steve has said the filters could be made sharper at the expense of more delay.  Not necessary in my opinion.  Perhaps the mathematicians and coders can improve the skirts; but it can wait, at least after NB. 

W9OY

The issue is one of bandwidth vs data transfer.  A carrier has basically no bandwidth.  As soon as you start keying a bandwidth develops.  A 10hz filter therefore may be too sharp to be able to demodulate the information contained in the signal.  The bandwidth generated is a function of data rate.  A 100 wpm CW signal has a wider bandwidth than a 10wpm CWsignal.  Signal to noise on the other hand is also a function of bandwidth.  As the band width decreases the relative signal to noise increases.  So what you want is the least bandwidth capable of sustaining intelligent demodulation at whatever the data rate is.  My experience for most CW is about 25hz-30hz bandwidth.  Does this really matter?  If you are trying to receive a signal basically under the noise it matters.  It's not dramatic but subtle.  If you add things like diversity which again often adds a subtle improvement pretty soon you are copying what you could not copy before.  The point of course is you want to design filters for signals with bandwidth, and not just something that is amazingly sharp on signals with no bandwidth. 

73  W9OY 

Ken - NM9P

It is hard to imagine better skirts on these filters, especially the SSB "Brick Wall" filters.  The CW filters are nothing short of amazing in my book!  I understand that the Digi filters have a bit less slope performance, but that is to reduce latency.  Still, they are better than anything I have ever used before, and a generation better than PowerSDR's, which were my previous favorite!

Ken - NM9P

Right on point, Lee.  Combine 25-50 Hz filters, tweak the audio chain noise with the RXEQ, add a little APF, fine tune the AGC-T to find the "sweet spot" (or carefully adjust gain level with AGC OFF) and I am copying signals that I never would have even noticed before.  Add a super-zoomed panafall display that lets me see a signal before I even begin to tune for it, and "the hits just keep on coming!"   

I am really looking forward to customizable filter buttons.  I would gladly lose the widest one or two CW filters and add a 25 Hz at the other end.  I hardly ever use anything even as wide as 400Hz, but then I am seldom on the receiving end of a CW pileup, or running a frequency during a contest.  (This may change when my tower and Log Periodic go online!)

Ned K1NJ

  Ken and Lee,
      Just updated to 1.2.17 to see if it would make a difference.
However I still cannot drag the filter so that the number in the middle of
the bottom of the tuning trapezoid to less than 50.  Are you both able to
do that with your 6500 and 6300?

Ned,  K1NJ

Charles - K5UA

Hi Paul, That's what I was trying to say. Not a lot of readers are goint to know what we are talking about unless they are CW veterans of PBT or IF Shift. Thought I would just mention saving IF offset with the filters as a possibility, I doubt that this will get any traction among the faithful.

Ned K1NJ

      Figured out how to do this by grabbing the vertical filter edges instead
of moving the mouse up/down at the base of the filter center line.  All OK.

Ned, K1NJ

David

Is this on a feature addition list and in the high or lower priority range?

The ability to custom the default RX bandwidth filters would a great help. I would like to change the SSB low from the 100 default to something more in the 200-300 range. I find that to help improve intelligibility of many stations that have too much bass/lows in the audio. I have tried the EQ but find cutting at the source best.

KC2QMA_John

YES, YES, YES Customize RX filters! Seems so obvious to me!

These are the kind of simple features we are all waiting for.

Already looks like a lot of votes at the top of the page.

Peter K1PGV

Yes, please. Custom filter widths. It's hard to believe this isn't a basic feature, it's so easy to implement. Peter K1PGV

Steve G1XOW

yes, yes YES!

2 years have past, why is this obvious feature and clear user benefit getting knocked back in preference to other very minor tweaks?

KC2QMA_John

I don’t mind paying for great software but I just hope I wont have to pay to get simple features like these.

Please remember to vote at top of page if you like this idea.

Dan -- KC4GO

A test using Spectrum Lab software (http://www.qsl.net/dl4yhf/spectra1.html) and the Flex 6500 DAX audio between 50 Hz and 10 Hz  CW filter setting. Clearly it makes a difference to the band pass.
Video: https://drive.google.com/file/d/0B5gpcTVs7PtoYnVaUUxoY0RiSnM/view?usp=sharing

I have included the link for Spectrum Lab Software latest version.

Rich McCabe

Older thread but found it after taking delivery on a new M radio. I miss the ability to program filters as mentioned above.  The 250 filter is my most common used CW filter and its not available in the right panel and only available on the slice. I was able to set filter widths on my "non smart" Icom transceivers so seems like the SMARTEST radio should be able to do it.

That said what brought me here was functionality of my new 6600M. I probably should start a new thread but will add it here unless told otherwise.  Toggling through 8 filters on the BW select on the new radios (and Maestro) does not work well. That's to many filters and if you go past the one you want you have click 6 or 7 times to get back.  Typically 3 filters is enough and some might argue they want 4 of their choice.

I would like it see a press and hold on the filters so you can choose knob delectable filters.

In the photoshop below I added a tab on the corner of the filter but could be a star or anything else.

I am sue there will be a lot of changes in the future with the "M" interface as the amount of users is going to dramatically increase but I think thinks would be a great addition.

Rich