Horizontal Line for AGC-T

Ken - NM9P

For best normal sensitivity, yes, adjust from right to left until the noise JUST starts to drop.  This is the best overall setting.

But once one gets familiar with the interactions, a bit of "fine tuning" can be useful.  

On a noisy band, or when I am not interested in picking out the weak ones, then I will set it a LITTLE further to the left (about 1-4 additional "clicks."  using the mouse to click one-by-one.)

When trying to copy a weak one in the presence of rapidly changing noise or interference, sometimes I will bump it a notch or two to the right to eliminate any pumping caused by varying noise or strong nearby signals.

After adjusting WNB, NR, ANF, or changing RX EQ settings, I will tweak it up or down a tiny bit again to maximize signal-to-noise ratio.  Fine adjustments of one or two "clicks" can sometimes make a big difference in either comfort or readability.

I can't precisely detail an exact formula for this.  It has developed as more of an instinct as i have gotten used to the sound and "feel" of the rig.  It is the sort of "art vs. science" thing that drives engineer-types crazy!  

(I live a bit in both worlds - I majored in Math & Physics, then went to get an MDiv and became a pastor...My Myers-Briggs personality test rates me half-way between Introvert/Extrovert, and also halfway between "right-brain/left-brain" categories!)

My experience with WNB has been mixed.  it seems to work well, even spectacularly,  on some types of noise.  On other types of noise it is less effective.  Thankfully, the standard NB that was reintroduced in v.1.5.1 seems to work on some other types of noise that are not eliminated by WNB.  And on other types of noise the NR function is effective.  BTW, NB & NR seem to be better in v.1.5.1 than in other software releases.

My evaluation is that they have come a long way, but WNB still needs some work.

One suggestion is a mode for WNB that might be linked to the panadapter width, rather than the whole SCU.  This would help in cases, for example on 40 SSB, where radio Havana and some other broadcasters just above 7300 KHz cause the WNB to add distortion and display pumping.  If I could put these strong signals ff the panadapter and therefore out of the WNB sampling range, then they wouldn't be as much of a problem.

I have some noise sources that seem to be "highly correlated" but not universally wide-banded.  They only cover a medium frequency range.   If WNB is analyzing the whole SCU input, it might be missing something, or might not be as effective on the stuff that s limited in frequency range.  

Of course I give this feedback not fully understanding the algorithm used for WNB.

Cheers, Walt.

Ken - NM9P

DrTeeth

Sounds like an idea for a mini-video Ken ;-).

Ken - NM9P

It is on my rapidly-expanding list.  But I need to get through this weekend first.  And I do have my day-job to get back to after this week's "stay-cation" time is over.

Al_NN4ZZ

Here is the link to tge option to automate the AGC-T....but it still provides the ability to manually adjust it if desired.  As Walt suggested if there is an ideal setting that can be determined in software that would be the recommended automated setting.  Coupled with one of the horizontal line options it should give everyone something to like. 

https://community.flexradio.com/flexradio/topics/option-to-automate-the-agc-t-setting

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW......... V 1.5.1.70
SSDR / DAX...... V 1.5.1.152
CAT................... V 1.5.1.0
Win10

Al_NN4ZZ

Here is the link to tge option to automate the AGC-T....but it still provides the ability to manually adjust it if desired.  As Walt suggested if there is an ideal setting that can be determined in software that would be the recommended automated setting.  Coupled with one of the horizontal line options it should give everyone something to like. 

https://community.flexradio.com/flexradio/topics/option-to-automate-the-agc-t-setting

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6700 - HW......... V 1.5.1.70
SSDR / DAX...... V 1.5.1.152
CAT................... V 1.5.1.0
Win10

Barry N1EU

IMHO one of the few glaring deficiencies in SSDR is the lack of a simple onscreen display (just a line) of the AGC-T level like PowerSDR provides.  For a small investment in development resources, I believe the result would be 99% of the user base dancing in the streets.

We don't need dBm or s-unit values, just a line that we can set relative to the panadapter trace of signal and band noise levels.

Barry N1EU

Bob - W8FB

I couldn't agree more!

Bob, W8FB

Bill -VA3WTB

For me it's is fine as is, It is really easy to find the knee or sweet spot.

Barry N1EU

If I change bands and it sounds noisy, I could just look and instantly see if AGC-T is set properly by where the AGC-T line is sitting relative to the baseline noise level and other signal traces on the band.  Without it, I have to reach for my mouse, and start fiddling back and forth with the AGC control.  There is no doubt that adding that line on the trace is going to make operating a whole lot easier and minimize mouse fiddling with the AGC control.

Greg

I suggested this a LONG....time ago....same approach to a squelch level.

Ken - NM9P

@Barry, I agree that a simple line either all the way across the filter flag, or some type of triangular arrow indicator on the slice flag at zero beat would be very helpful.  While many of us have gotten very proficient at adjusting AGC-T, and other DSP functions like NR by ear, others may be more visually oriented and would benefit from the visual feedback.   it would indicate that "Signals above the line will be heard at full volume and limited by the AGC.  Signals below the line will decline in strength in proportion to their distance below the line...."

@Greg, Yes... another indicator could be used to set a simple level-controlled squelch.  Though eventually I would like to see a more sophisticated routine that recognizes the shifting pitch of the human voice and opens for speech, but not for other noise.

Ken - NM9P

Ken - NM9P

I just noticed that I hadn't comment on your drawings, Al.  I like #2 better, because it has a separate shorter line for each slice.  That would make for a cleaner, less busy display.  The number indicator could come up when adjusted, or upon mouse hover, and then fade after a few seconds.

DrTeeth

Sounds like a good idea. I am saying no more than that about it in this thread.

It looks like this thread has been taken the strain since the https://community.flexradio.com/flexradio/topics/smartsdr-mouse-wheel-tuning thread was closed, LOLZ.

Déjà vu:- Good idea proposed*, then people come forward agreeing. A short time later the 'Flex-can-do-no-wrong/it-does-not-generate-$s' pack will arrive. I'm going to really try hard to sit back and watch this one.

Seriously, there has to be a better way of suggesting ideas without the 'other side' getting angry/personal and a thread getting out of hand.

Greg

Its all Barry's fault...blame it on him.  He likes to stir the ****.

Just kidding....hi Barry! 

Jay Nation

Tim will eventually get annoyed enough, to remind us of the Feature Request #. Or maybe even create a number for it, If it's not already on the list.

It will get added to whatever release it get's added to. not the one before that, and likely will get fixed in a later version. For some reason it always seems to happen, just that fast, and in that order.

 SDRgadgets

73, Jay - NO5J

Tim - W4TME

It isn't annoyance as much as being informative.  So here we go...

The AGC improvement is feature request #2129

Kevin - I am not the product manager, so I cannot provide a yes/no answer to whether or not a feature will be added.  Since we do JIT (just in time) release planning, there are very few forward-looking features we talk about.  Yes, enhancements for setting AGC is a popular topic.

I'd personally rather see the AGC-T be set dynamically for optimal SNR based on the real-time band noise conditions so you don't have to muck with setting it (feature request #2134).   

Jay Nation

Tim 

Nevermind my last post, If it was seen. I removed it. I figured it out.
 
SDRgadgets

73, Jay - NO5J

Jay Nation

The post before your last in this thread wasn't intended to annoy. I was only poking fun, in your direction. Thanks for the Idea numbers, We'd all be lost without all the hard work you put into the Community. 

 SDRgadgets

73, Jay - NO5J

Walt

Big thumbs up for the #2134, the automatic AGC.

I have been getting tired of adjusting - it reminds me of my first radio, a knight kit span master, adjusting the regen knob for max sensitivity. 

After 50 years of technology - I hope you can make this function automatic.

Cheers !

mikeatthebeach .

I also agree
73 Mike
WB6DJI

mikeatthebeach .

Big thumbs up for the #2134, the automatic AGC.

Ken - NM9P

I will reiterate my statement fron 10 months ago...
I like the idea of an "automatic AGC" as long as I can still go back to the current method at will, OR if the slider could become a +/- adjustment to nudge the position a little above or below the nominal position to account for preference or band conditions.

So Manual AGC mode - the slider functions as it does now,
Automatic AGC mode - the slider becomes a center-detent +/- control.

Ken - NM9P

If they could implement this in conjunction with an "Active Hang AGC" that I proposed in another post, it would be heavenly.

Tim - W4TME

I do not suspect that an automatic AGC mode would be an all or nothing type feature, in that regard, so a way of providing a manual override would be included.

Tim - W4TME

No harm, no foul.