NEW Filter setting -- V 1.9.7

  • 2
  • Question
  • Updated 3 years ago
  • Answered
How does the filter settings work.  The questions below are for both the manual and AUTO features.  My focus is CW only but others may find the answers helpful for other modes as well. 

1 - are the manually set values only active when the AUTO button is not active? (i.e gray).  I see the manual values can only be changed when the AUTO button is greyed out and assume they are only used when not in AUTO mode.  Correct?. 

2 - Do the default values shown on the slider represent the previous filter shape from earlier versions of the software? (about 3/4 of the way toward SHARP)

3 - when the AUTO button is active the flyover says "automatically adjust filter sharpness based on bandwidth."    What is the intent (sharp or fast) and how is it changed and displayed as the bandwidth is changed?    It would be interesting to see the display change the skirt angles but that is probably a lot more trouble than it is worth.

 

4 - I didn't see an explanation in the release notes or changelog but in case I missed it, is there an update to the user guide with this explanation?

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10

Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes

Posted 3 years ago

  • 2
Photo of Eric - KE5DTO

Eric - KE5DTO, Official Rep

  • 913 Posts
  • 343 Reply Likes
The users guide is updated and does indeed cover these new features.  http://www.flexradio.com/downloads/smartsdr-software-users-guide-pdf/
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Eric,
Thanks, just downloaded the updated guide (page 144).   I must have missed the reference to an updated version being available.


One more question:  In the FLEX Insider Steve mentions that an increased filter depth is now available.  However the release notes seem to indicate the far right setting is equal to the previous software filter.  

How do you get the increased filter setting if the FAR RIGHT slider position is the max and equal to the previous SSDR filter?
 

*** INSIDER ***




*** RELEASE NOTES ****




Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10

  
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
After re-reading the release notes I think the answer to question 2 is that the default slider setting is a new setting compared to previous software versions.  It is a compromise with slightly less sharp filters.  The previous version software is equal to the far right SHARP setting.  

So if you want the filter to be like it was in previous versions of  SSDR, you have to disable AUTO and move the slider to the right.

Correct?

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10


Photo of Chris Tate  - N6WM

Chris Tate - N6WM, Elmer

  • 975 Posts
  • 272 Reply Likes
If I recall correctly the auto is not on the extreme side either way, but a reasonable middle ground setting..  so your assumption would be correct.
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
Auto changes filter latency/sharpness based on filter bandwidth.   Page 30.11 of manual.
Photo of Ken - NM9P

Ken - NM9P

  • 4239 Posts
  • 1351 Reply Likes
If I remember for my testing....Full right is the standard "Brick-Wall" filters we have been accustomed to using.  The three settings to the left represent different degrees of wider skirts but lower latency.  If you liked the former "brick wall" filters, keep it in manual and all the way to the right.

If you want lower latency and faster turn-around time, and some have reported, better QSK on CW, then experiment with other settings.  But know that the skirts get wider and wider as you go further to the left.  

It is a trade-off game!  But one that is fun to play!  The difference in the way the filters sound on CW and even SSB are definitely noticeable.

BTW:  when using lower latency settings in CW, the effect of APF is more pronounced, even at narrow bandwitdths.

I have found that this change, and a change in the TX SSB filtering have dramatically reduced the SSB latency when monitoring myself using the Full Duplex option I detailed in my tutorials about 10 months ago!

It looks like I will need to make a few new tutorials!

Ken - NM9P
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Before v1.9, we controlled the shape factor of the filter based on our "guess" about what you were trying to do.  These automatic settings are mentioned elsewhere in the manual.  So whether you increase of decrease sharpness depends on your exit mode/filter settings before you unchecked auto.
Photo of Gary L. Robinson

Gary L. Robinson

  • 254 Posts
  • 56 Reply Likes
I am mostly a digi operator and I can tell that the highest sharpness filter setting for DIGU is definitely better than it was previously.  I can see it in the waterfall.  All the previous versions of SmartSDR I have ran looked closer to the NEXT to the highest setting.  So I am tickled to see it sharper!

---Gary
Photo of Andrew O'Brien

Andrew O'Brien

  • 384 Posts
  • 44 Reply Likes
I'm mostly a digi op also.  What advantage does this offer ?
Photo of Gary L. Robinson

Gary L. Robinson

  • 254 Posts
  • 56 Reply Likes
makes it easier to block more OR all of a close or slightly overlapping signal and getting better or perfect copy. Especially weak signal modes - not so much for rtty and psk (depending on exact flavor of psk).
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
Lets take something like RTTY.  For the benefit of others, the ITU/NTIA formula simplifies to <baud rate> + 1.2 * <shift> where baud rate is in bits per second and shift is in Hertz.  Thus for RTTY with 170 Hz shift the "necessary bandwidth" in Hz is 204 + baud rate.  Let's round to 250 Hz.  If left as AUTO, the filter has always been as tight as possible even before 1.9.7 at 250 Hz.  There would be no difference in 1.9.7 filtering vs 1.8.  Now if you were using wider modes > 500Hz then disabling AUTO and leaving the slider far right would be beneficial at isolating the signal at the expense of latency.   Tighter means slower.  For Olivia rag chew, no one will notice.  But on RTTY contests, it may be an issue and widening the skirts would be preferred.  This is why this feature was added.
Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 323 Posts
  • 138 Reply Likes

Ken,

I notice that in CW mode, the filter slider setting that most closely approximates the old is position 2 (from left).  Probably not exact, but close.  Since owning a 6700, I've been closely watching the opposite sideband rejection when applying a strong carrier to the receiver. 

Until now, other SDRs on my desktop have allowed for shaper skirts and thus absolutely nothing heard after zero beat is reached and continuing into the opposite sideband.  With SSDR, I could always hear a bit of a response rise on the opposite side of zero.  However, none of the other SDRs have Flex's QSK speed.  There's a complex design trade-off between latency, filter sharpness, and T/R switching time.

In slider filter positions 3 and 4, the CW skirts are now sharper and at least equal to the other SDRs.  The other SDRs allow for a continuous setting of filter taps which is nice, but the 4-level slider approach is simpler in operation and probably all anyone needs.    

This said, I would like the new filter menu placed on the main SSDR screen -- somewhere, somehow.  It's definitely something that I would be adjusting as operating conditions change.

Paul, W9AC

Photo of Sergey, R5AU

Sergey, R5AU

  • 862 Posts
  • 117 Reply Likes
BTW Paul, I have tested QSK with all slider positions and for my self CW pretty perfect for my ears
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Hi Paul,
I need to experiment some more today with the new filter settings but it seems there are differing opinions on what CW filter setting is equal to the pre - V 1.9 filters.   I've re-read the user guide and comments from you, Steve, and others but still not sure.   




RE your suggestion: ...I would like the new filter menu placed on the main SSDR screen -- somewhere, somehow.  It's definitely something that I would be adjusting as operating conditions change.

Idea - Here is one option to consider: When you cursor over the filter and the "skirt" pops up, the appropriate slide based on the mode, pops up as well.  This option would not permanently use any of the main screen real estate which could be a plus. 

  


Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10
Photo of Stan - VA7NF

Stan - VA7NF

  • 469 Posts
  • 118 Reply Likes

I found the previous filters true brick wall.  Contesting - CW - 200Hz filter - If a station was just off frequency (75Hz) I could not hear it at all - Visual indication only.

The new filters add a slight passband (right to left) of 100Hz, 200Hz, 800Hz.  Tune into WWV, narrow CW and listen/watch the S meter as you tune off centre.  The differences are obvious.

Thank you FRS - At least now I will hear the station if it is strong enough, possibly enough to catch a call sign.

Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
The old filter settings are still there. Slide the slider all the way right when at 200hz.
Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 323 Posts
  • 138 Reply Likes

Al,

Starting from the left, I was referring to that as position 1.  Then sliding all the way to the right I referred to that as position 4. 

When listening to opposite sideband suppression of a carrier in CW mode, it seems to me that position 2 is close to the prior SSDR releases.  This really requires measurement to verify.  Possibly Flex plotted a family of before-and-after filter curves while they were testing?  It would be interesting to look at the overlays. 

Nice suggestion BTW on the screen pop-up.

Paul, W9AC

(Edited)
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Paul,
I usually set the bandwidth for CW to 100 HZ as my default setting.  And the filter setting of 4 (far right) is what you also get when in the AUTO mode for this bandwidth.  Although I've only been playing with this setting for a few hours, it works fine for me.  I'll probably just leave the filter set to position 4 in manual mode.

However, I can't tell whether this filter setting is sharper than what was provided in the earlier versions of SSDR or not.  

I like your idea of seeing the filter curves for the 4 settings and the curve for the older software.  That would certainly clear up any confusion about how they compare.  It would be easy to see which one of the 4 filter curves in V1.9 most closely matches the one from SSDR V1.8.  

@Steve H. -- did your team happen to plot the curves during the testing?  

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
There were no change in the filters.  You are just able to select it manually now independent of filter bandwidth.
(Edited)
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
That is a surprise. When I read Steve's comment in the Insider about "sharper" filters, it sounded like they were new and improved.

See snapshot below from the Insider.

Regards, AL/ NN4ZZ

(Edited)
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
No changes on RX filters, just the freedom to select.   Now on the TX filter side, latency was decreased by using one of the less sharp filters.
Photo of Mal

Mal

  • 47 Posts
  • 0 Reply Likes
Is there any way to change the filter settings from the Maestro?
Photo of Ed.G

Ed.G, Software Engineer

  • 103 Posts
  • 29 Reply Likes
Click the MENU at the top of the Maestro screen and then click the Phone/CW sub-menu. You should the filter settings under the label "Filter Options". You will have to scroll down some to see all the mode groups. 
Photo of Mal

Mal

  • 47 Posts
  • 0 Reply Likes
Thanks got it.
Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 323 Posts
  • 138 Reply Likes

Al, I should have mentioned that my opposite CW sideband testing has been done with much wider filters -- like 3.5 kHz.  That's where you hear the difference.   I normally run 2.7K to 3.5K during casual rag chews.

Paul, W9AC

Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
When on AUTO, SSDR is behaving just as it did prior to 1.9.7 with RX filters.  Below are the steps used based on bandwidth of the filter when AUTO is enabled for that mode.   The filters did not change in this release, just the ability to decouple the Sharpness from the Bandwidth if you so choose.  Anything else stated is a placebo effect!

Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Thanks, that clears it up.....I'm sure others also took  Steve's comments about "even better shape factors" as meaning the RX filters had been enhanced.

Maybe your explanation should be added to the notes and user guide.  




Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW.................. V 1.9.7.85
SSDR / DAX / CAT...... V 1.9.7.169
Win10
Photo of K6OZY

K6OZY, Elmer

  • 542 Posts
  • 212 Reply Likes
Ya after reading that again, I can totally see how you could interpret that way.
Photo of KC2QMA_John

KC2QMA_John

  • 784 Posts
  • 332 Reply Likes
I love the new Filter Options, I mostly rag chew and having the ability to soften the filter skirts in voice takes some of the “Q” or Ring-Effect out of the receiver audio and makes for a softer sound that allows for longer qso’s without ear fatigue.

This was something I has suggested a while back and now it’s done!

Thanks FRS keep up the great work!

Photo of dl9eri/oe9eri

dl9eri/oe9eri

  • 23 Posts
  • 12 Reply Likes
Today I took my scope and measured the receive latency of the new filter settings just for knowing how they work and found that I can now use the slider for equal latency in all choosen bandwidth and modes. The numbers are +- 5mS.

The numbers: left is 0 - 1 - 2 - 3 and auto

In 0 : all modes and bandwidth  55 mS
In 1 : all modes and bandwidth  75 mS
In 2 : all modes and bandwidth 105 mS
in 3 : all modes and bandwidth 165 mS

In the -auto- position of the sliders there are different latencies for the bandwidth in CW and DIGI
from 55mS on 3-5KHz, 75mS on 1KHz, 105mS on 800Hz, and <400Hz 165mS. On SSB in auto there is always 165mS latency but you can choose from sliderposition 0 to 3 to change.

note: it is the receive latency when your partner sends a -did- and you can hear it on the 6500 after the delay in mS

best 73
Photo of Paul Bradbeer

Paul Bradbeer

  • 170 Posts
  • 66 Reply Likes
Thank you! I found this the most informative and clearest post on the subject. Well done OM. Chuss! Paul M0CVX
Photo of Timo - OH5KW

Timo - OH5KW

  • 23 Posts
  • 0 Reply Likes
I just tested what effect the slider makes on digi modes with MMTTY. Seeing is beliving, hi.

73, Timo OH5KW
Photo of dl9eri/oe9eri

dl9eri/oe9eri

  • 23 Posts
  • 12 Reply Likes
Timo - what did you find with the different slider settings in RTTY - what ́s the result of your test in this mode ?
This would be really interesting for all digi-mode users !
Photo of Timo - OH5KW

Timo - OH5KW

  • 23 Posts
  • 0 Reply Likes
Use FFT window in MMTTY. If you have narrow filter you can see the difference just by moving the slider. In digi mode a narrow filter is important, I think, but CW contest is different story. Last weekend I worked 1000 QSOs in SAC contest using lower latency and little bit wider filter. Worked fine.
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9189 Posts
  • 3553 Reply Likes
The "auto" digital filters are progressive.  Narrower filters have sharper skirts (greater latency) and as you increase the filter size, they have lower latency (wider skirts) characteristics.