DSP functions ANF, NR & NB are basic radio functions that really need attention.

  • 14
  • Problem
  • Updated 5 years ago
  • Acknowledged
6500 on SmartSDR v1.3.0 Windows 7 Ultimate, I Intel i7 Processor, 16 Gig RAM.

I purchased the 6500 as a step up from my 14 year old Icom 756 Pro.  In many ways it is but the basic DSP functionalities of noise reduction, auto notch filter and noise blanker simply don't measure up to the Icom.  I don't regret buying the 6500 but PLEASE focus on these basic functions as soon as possible.
Photo of David H. W5CAH

David H. W5CAH

  • 13 Posts
  • 1 Reply Like

Posted 5 years ago

  • 14
Photo of dl9eri/oe9eri

dl9eri/oe9eri

  • 23 Posts
  • 12 Reply Likes
I am also comming from the Icom 756proIII after 7 years and many other Rigs before and I find the main functions of a Tranceiver should not to be in API and Remote Control - first I would like to see a good NB, ANF and NR and Memories. The persistance with profiles is vy bad, every time I switch from band to band I get other settings of qrg and power slider, somtimes 3 Receivers not set before etc. etc.. The general-receiver when used shows after usage the other bands with changed parameters than before. Keying CW without extra hardware would be much desired via CAT !

So there is for my 6500 some frustration, because i did expect much more and it may be a long path to go with flex to get main features of tranceivers implemented the way that they realy work.

I dont even understand, why they are not able to post a short note of how the profiles are stored and which of the parameters there are included in global or transmit profiles ? I can not imagine that this is much to write down and publish ?

As the basic setup and hardware is real excellent I will not be impatient now - and hope they will get the DSP-Proramming done in some time. It seems that this kind of SDR-Programming is much more difficult than doing the things inside the PC like others. And on the long run this should be the advantage of the 6000th concept - it will take more time...   dont worry with my comments please !
(Edited)
Photo of DrTeeth

DrTeeth

  • 1687 Posts
  • 389 Reply Likes
I have a very noisy environment and am new to the 6300. Been out of ham radio for over 25 years. I find that 2 of the three DSP functions do not appear to do anything at all, just the NR. I am looking forward to functional DSP sooner rather than later.
OT: My time is currently being spent trying to decide to which software (logging, data modes) to chose and commit to. Shame so many offerings are abandonware.
Photo of DK1EY

DK1EY

  • 177 Posts
  • 48 Reply Likes
Hey Guy,

Try N1MM+. I use it with my 6300 as well. How to setup, you can find here.

Trust me, a great software and does all I need here.

VY73
Tom
DK1EY

By the way: I also hope for better DSP-abilities!
(Edited)
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3784 Posts
  • 1637 Reply Likes
Guy

Like you I seem to try all the different logging software

For a logging.
Right now I use Ham. Radio Deluxe 6.2

AND DXlabs.

Both do things just different enough that at different times I tend to prefer one then the other

For Contesting I use Writelog... Mainly because I have used it forever

for PSK./RTTY MixW. Mainly because I have used it forever. Albeit DM780 is more effective

WSJT-X for JT 65/9.

CWSkimmer for CW albeit I rarely if ever do CW. So CWX if I ever need to XMT.
(I have not owned a CW key since 1965 when I tied my key to a lead brick and dumped it into Lake Ontario to celebrate the fact that I had the most advanced license and that I would never have to do CW again)

VE7CC and HRD for Spots.

DDUTIL to tie everything together

Parallels Access for remote ipad operations.

And a whole bunch more depending on my mood.
(Edited)
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Hi David,
For reference, here are links for 3 of the ideas for improvement.

NB
https://community.flexradio.com/flexradio/topics/nb_idea

NR 
https://community.flexradio.com/flexradio/topics/improve_nr_noise_reduction_request_to_add_another_d...

APF
https://community.flexradio.com/flexradio/topics/ssdr_v1_2_apf_could_use_more_gain_it_seems_diminish...

If you haven't already added your VOTE,  go to the link and click the LIKE button.



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

DrTeeth

  • 1687 Posts
  • 389 Reply Likes
Voted for all three!
Photo of Mike KD2CJJ

Mike KD2CJJ

  • 122 Posts
  • 34 Reply Likes
It does baffle me why simple DSP functions have not been a priority.  Its almost to the analogy that a car company comes out with all these new novel functions on how a car operates but forgets to implement Wipers and head Lights.  In my mind commodity functions should work as good as the next system - No worse and would not expect any better.  Of course as a future enhancement novel approaches to these functions would be great but could come later on and prioritized against other novel functions.  I am a prospect and will not lump down cash until these functions as good as my current radios.  I unfortunately live in a high noise environment (like most New Yorkers who live on Long Island) and these for me are must haves.  If you cant hear them you cant work them.  Does not matter how cool or how many novel functions the radio has!
(Edited)
Photo of Ned K1NJ

Ned K1NJ

  • 319 Posts
  • 83 Reply Likes
   Ironical, Mike.

    PSDR/5000 and SmSDR/6000:
  "We have met the enemy and he is us".

   Ned,  K1NJ
Photo of Stan - VA7NF

Stan - VA7NF

  • 469 Posts
  • 118 Reply Likes

There are many threads dealing with different aspects of DSP for noise reduction.
Since it has been a slow process to improve these functions I must assume that the resolution on these radios is difficult. 
Steve, please open the FRS kimono just a little and share the progression; it may/should go a long way to ease these types of posts.

Thanks

Stan Williams, VA7NF

p.s. I want this yesterday as well ;-)  

Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Official Response
We are indeed actively working these algorithms.  In the past, we tended to ask folks for a sample of what they are trying to overcome (noise file) and then we would tweak the algorithm.  This approach has tended to make it work better for one guy and worse for another so we are not happy with this approach.  Instead, we now have some folks doing some more in-depth analysis of the algorithms and suggesting changes.  We are also in the process of a long task of setting up automated software testing on all four SmartSDR radios with the intent of checking all of these algorithms against different stimulus each release to ensure changes don't creep in.  I'm hoping to have the first cut at changes in v1.4, but this is what we call a "science project" to pick the best algorithm and make sure we don't make things worse before making them better.
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1853 Posts
  • 672 Reply Likes
Good news Steve,
I'm glad to hear that they are all on the radar and being actively worked.  V1.4 would be be great!  

The ideas should probably have their status changed from "no status" to "under consideration"  or even better to "planned."    That way it's clearly apparent they are being worked and might help alleviate some of the questions. 


NB
https://community.flexradio.com/flexradio/topics/nb_idea

NR 
https://community.flexradio.com/flexradio/topics/improve_nr_noise_reduction_request_to_add_another_d...

APF
https://community.flexradio.com/flexradio/topics/ssdr_v1_2_apf_could_use_more_gain_it_seems_diminish...

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
Photo of Charles - K5UA

Charles - K5UA

  • 319 Posts
  • 89 Reply Likes
Hi Steve,
Among the smorgasbord of algorithms you are evaluating for the 6000 Series radio DSP functions, could you just use the basic algorithm used in the Flex 5000 that everyone seems to hold out as the gold standard? Is noise in the digital domain different in a PowerSDR radio than in a SmartSDR radio? Is it a matter of programming the algorithm in SmartSDR language, or does the PowerSDR algorithm itself not work on the digital domain noise detected in a SmartSDR radio.

I know nothing about DSP programming, but I suspect that a lot of others are wondering why this is such a mountain to climb.

Not complaining, just wondering out loud.
Photo of W5XZ - dan

W5XZ - dan

  • 571 Posts
  • 86 Reply Likes
thanks for the good news Steve....looking forward to the improvements.

73, w5xz, dan
Photo of Dale KB5VE

Dale KB5VE

  • 418 Posts
  • 56 Reply Likes
I can see where trying to use people's issues to creat the perfect dsp functions would create a merry go arround effect, the more you do you never get to the end. A lot of us flex owners have discussed why this has been such a issue from the beginning. I have owned probably 25 of the modern radios with dsp functions both if dsp and audio dsp. Also most of the after market dsp boxes and boards. In talking about it we all agree that most over 90% of these have vastly superior dsp NR and beat cancel. I have used the fact is they all have dedicated dsp chips with the "---floating " dual dsp chips ect. Is the reason these dsp function so well and are the benchmark we judge the flex by because these use allogarithims developed by vendors and purchased by icom, yaesu, kenwood and tentec, else raft?

To be honest I seldom have a dire need for the super dsp people want, but I do feel this is a rock arround the flex neck which will haunt it until it meets a certain standard. Reciever is the best I have ever owned and it has been worth the wait, but from all the request you have to admit the dsp nb,NR and anf is the one universal issue which keeps comming up.os
(Edited)
Photo of David H. W5CAH

David H. W5CAH

  • 13 Posts
  • 1 Reply Like
Well said Dan! Thank you!
Photo of WA6FXT Mike

WA6FXT Mike

  • 223 Posts
  • 40 Reply Likes
Steve,
Thanks for the update. I have a question that I'm sure has either been asked already and / or is pretty dumb... Why won't the algorithm used in PSDR work in SSDR?

73's de Mike
Photo of John n0snx

John n0snx

  • 229 Posts
  • 54 Reply Likes
I don't know how, BUT just fix it mentality...LOL...I have read the complaints since the first day I joined this community... I for 1 have seen the development of PSDR from ver 2.0 til now.... the improvements were outstanding... that took more than a couple of years to achieve.  Now we have SSDR... in it infancy and everyone wants it to be perfect for there own application. SSDR is not the same architecture as PSDR. It's not even close. So its easy to say... Make it like PSDR...but they can't.  Before I purchased my new Flex6xxx I read and KNEW before I purchased the radio(apparently some didn't) that the software was evolving and would get better with each new version. FRS has done a great job so far and I'm sure they will continue.....

Patience is a virtue... Good things are coming.... Maybe not fast enough for you, but they are coming none the less....

My hat is off to FRS for the fine job I think they are doing.  Including the software.

John
Photo of Jay Nation - NO5J

Jay Nation - NO5J

  • 16 Posts
  • 4 Reply Likes
Yeah! What he said.
Photo of Jay Nation - NO5J

Jay Nation - NO5J

  • 16 Posts
  • 4 Reply Likes
Little X, at the top, far right, click it. you have just located the real Noise Reduction button. works sorta like the BIG knob in the middle on legacy knobbed radios. Give her a spin see if it reduces the noise your talking about. Sometimes I find signals sort of down below the noise that are kinda interesting. Adjust the AGC-T lower and the slice volume up, with or without the DSP/NR button selected until you can hear it how you want it to sound. When you get it sounding the way you want it to sound your done. You don't have to settle for how I want it to sound. Yet!
Photo of DrTeeth

DrTeeth

  • 1687 Posts
  • 389 Reply Likes
If it were as simple as that, FRS would not be working on improving the algorithms.
Photo of Jay Nation - NO5J

Jay Nation - NO5J

  • 16 Posts
  • 4 Reply Likes
Guy

This rant is not directed at you, or anyone else that has only recently discovered SDR.

For some of the rest of you.

Try this without the quotes, in the Search box at the top of this page
(Hint: The " | " symbol is equivalent to "or", to most Search engine algorithms)

"Improved NR | DSP".

I'm betting Guy is correct, they probably are working on improved algorithms, it's been requested often enough. Steve and Gerald have even confirmed it's being worked on.
More than once. But my post was going to be about how to deal with the noise until we get the new NR.

Also take a look at:
https://community.flexradio.com/flexr...

That's the same method that made it work better in PowerSDR.

-Edit-
I removed a couple thousand misspelled words loosely arranged into badly punctuated sentences, in other words a Free Rant.
-Edit-

I'm stopping here before I get too silly. I'm going to write up a nice How-To article about NR, and how to be free of listening to noise. I'll post it when its been edited down to something that is pleasant correct and useful, while using the best grammar I'm capable of, and almost properly punctuated. I'll see what I can do about a translation to actual English. However, you may have to settle for Texanics. I tried using "Rosetta Stone" but it all sounds the same to me.

I tried this using screen captures but you can't hear the actual noise improvement.

I tried making a video but you can't see the improvement either.

So I will describe the process I use step by step.

Stay tuned ...

This conversation is no longer open for comments or replies.