How to MASSIVELY Reduce Your Noise level on SSB & CW.....For Free

Поделиться
HTML-код
  • Опубликовано: 4 фев 2025

Комментарии • 131

  • @ZiggleFingers
    @ZiggleFingers Год назад +10

    If you have an Nvidia GPU that Nvidia Broadcast supports you can do this locally. It strips noise just like this. ModernHam here on youtube has a video 'Using Noise Suppression AI on Ham Radio to Remove Static' showing how to do it with examples. He also shows other software if you do not have Nvidia

    • @gareth1500
      @gareth1500 Год назад +1

      exactly I use the Nvidia system nr broadcast and works OK. You can also use the amd nr software if using AMD gfx cards.

    • @KC7ZXYtube
      @KC7ZXYtube Год назад

      Thank you. Something seems kind of fishy here. I know this same tech can run locally. There is no need for internet/cloud processing. That's kind of silly at this point.

    • @noomnoom8469
      @noomnoom8469 4 месяца назад

      All, it is still a great endeavour that the gentleman has done. Plus it is experimentation on how to use ai systems to interpret signals and also reduce noise.

  • @bill-2018
    @bill-2018 Год назад +6

    Thanks for showing this Colin, It works very well but I don't want to be tied to a computer.
    G4GHB.

  • @gtretroworld
    @gtretroworld Год назад +1

    Very interesting indeed especially for the older HF radios…Thanks for sharing 73’s M3VWN

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Yes, works with any radio as long as you can get audio out which you can with any radio one way ot another.

  • @hb9vqq533
    @hb9vqq533 Год назад +2

    Thank you Colin, great Video as always, keep it up! Maybe you have encountered a term coined "listening fatigue" during extended periods of Radio operation like contests for example. With RM Noise the listening fatigue is significantly reduced.

  • @kuyakano8976
    @kuyakano8976 Год назад +3

    The clarified signal; that's how quiet it was on the bands live in the 80s and 90s of the previous century :)

  • @paulsengupta971
    @paulsengupta971 Год назад +8

    Impressive, but personally I found the SSB less intelligible with the NR on, similar to my 991. It works great if there's already a good SNR but otherwise I prefer to switch it off. For CW though, it sounds fantastic.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Its not a miracle cure and its not always needed. Its a very useful tool though.

    • @KC7ZXYtube
      @KC7ZXYtube Год назад

      Seems kind of like a snake oil cure. Sorry if that sounds harsh, but there's no reason to run something like this as a cloud based service.

  • @M0JBM
    @M0JBM Год назад +1

    It's a great bit of software I have been using it for a few months now. Keep up the vids Colin ,happy new yr

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Thanks John, great to hear others using it. Happy new year when it comes.

  • @z3r0c00l2
    @z3r0c00l2 Год назад +1

    Wow this is amazing.... If it's real time 100% then it rocks. I use my radio remotely... So this is so important for me to work in real time 😮. I will try it

  • @Tom_2E1FUE
    @Tom_2E1FUE Год назад +1

    it's fantastic, I use this all the time this will work brilliant with your noise down the road when you had the power lines noise :) it picks the weak contacts up like 5/5 fantastic.

  • @pcfreak1992
    @pcfreak1992 Год назад +2

    Did you ever try the QRM Eliminator a.k.a. X-phase? It’s a little device that also does wonders to noise on lower bands and it works (offline) purely on the radio signals. I haven’t tried it yet myself but the reports I’ve seen make a good impression.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      They do get a good name but only good for 1 source of QRM if you have multiple sources then you need multiple units.

  • @reedreamer9518
    @reedreamer9518 Год назад +11

    I put a lot of effort this year into building a stand-alone DSP audio device (running an M7 Arm and 16-bit ADAC) to process the audio output of my rig. Since I wrote the firmware myself, I can experiment with any algorithm I like. Currently I'm running custom spectrum noise reduction code, and it works quite well. Nonetheless, I'd be curious to know what method these guys are using.
    I'm not sure I like having to upload my signals to a server for processing, rather it would be nice if they released open source code so I can try running it on my device.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Sounds impressive. I've no problem uploading RX signals as its being transmitted regardless. Yes a local solution would be preferable.

    • @johnwest7993
      @johnwest7993 Год назад +4

      Exactly. This is a hobby. We are engaged in it because we like to learn more about it. We aren't consumers. We're not CBers. We're hobbyists and experimenters. Appliance radios and black-box software at remote locations hold little value and serve little purpose to someone like me who wants to learn about RF.

    • @hb9vqq533
      @hb9vqq533 Год назад

      In a nutshell RM Noise is based on a AI model derived from the training of a neural network. The NN is being constantly trained with real audio samples. The client programs allows to capture and submit such audio samples which then will be used to train the model. This way RM Noise will get better over time as more samples make it into the AI model. At the same time we are also collecting QRM samples, like OTHR, PV systems, SMPSU, LED Lights etc. in order to mitigate these kinds of QRM. 73 Roland HB9VQQ

    • @TheFattredd
      @TheFattredd Год назад +2

      I'd love to hear more. Have you put together a writeup somewhere, or a github repo?

    • @reedreamer9518
      @reedreamer9518 Год назад +1

      @@TheFattredd No, I have not posted my code or done a writeup, yet. But I do have a nice Fusion 360 model fly-through animation I made while working out the mechanical layout. I'll have to figure out how I can post it.

  • @jroosh
    @jroosh Год назад +3

    Impressive, especially if you can reduce the delay to a more practical figure.

  • @Ei2iP
    @Ei2iP Год назад +2

    Wonder why you did test on signals that were buried in noise?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Not sure I understand? did you mean didnt test on signals buried in noise? It wont detect signals below the noise floor as I beleive but will make the difference to make the QSO or not.

  • @G5STU
    @G5STU Год назад +2

    cheers Colin, the software really is great.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Absolutely, just need my packet loss issue resolved.

  • @robbannstrom
    @robbannstrom Год назад +1

    I might just give this a whirl to see if I can receive ANYTHING in this very radio-noisy hole I live in! Dunno if you've seen it, or heard of it at all, but sota-antennas, followed by a "d-o-t" and then the letters "C", "O" and "M" has some nice configurable antenna designers with propagation patterns, VSWR curves, antenna currents diagrams and Smith charts, mainly for portable antennas, if you're into building your own antennas.

  • @digitalanaloguehamradio
    @digitalanaloguehamradio Год назад +1

    Hi Colin, it’s remarkable frankly what it does. I have seen a few people talking about it. there is actually quite a few ways to do this.but this looks the simplest. Thanks for the demo.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Very simple Chris, as long as you have a windows computer and internet.

  • @dxscotland5901
    @dxscotland5901 Год назад +3

    I finally got mine working colin and it’s absolutely brilliant!

    • @johnratcliffe6438
      @johnratcliffe6438 Год назад

      How hard was it to set up? Is there an idiots guide?

    • @dxscotland5901
      @dxscotland5901 Год назад +1

      @@johnratcliffe6438 I’m rubbish with computers tbh and I managed no problem

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Especially for low bands.

  • @timg5tm941
    @timg5tm941 Год назад +2

    Wow!!! Could be a wee game changer this Colin

    • @dennis8019
      @dennis8019 Год назад

      game changer? for what?

    • @norrinradd8952
      @norrinradd8952 Год назад

      @@dennis8019 4:00

    • @timg5tm941
      @timg5tm941 Год назад +1

      @@dennis8019 improving quality of reception? My golf swing?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +5

      For us with cheaper end rigs especially.

    • @GM5DDX
      @GM5DDX Год назад

      @@dennis8019 for regular bowel movements (and QRM on 40). WTFV

  • @chuck_N4SHO
    @chuck_N4SHO 21 день назад

    Very impressive.

  • @MattBrunton1965
    @MattBrunton1965 Год назад +4

    If it were April 1st, I'd think this was a hoax. Astonishing results! Matt 2E0LNS

  • @Banjoman-p7t
    @Banjoman-p7t Год назад +2

    Every one keeps complaining about local man-made noise but not a single operator who has complained that I have talked to has even installed or tried a noise cancelling phase system? They work! The QRM won't fix itself?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      I know a few that have tried one. Unfortunately if you have multiple sources then you need multiple noise phase cancelling systems.

  • @pcfreak1992
    @pcfreak1992 Год назад +3

    It’s not bad but at times it’s quite unintelligible when the original was still sounding fine.
    Personally, I’m not a huge fan of internet-connected ham radio gear (and I’m saying that as a millennial) even though I am not a prepper or anything. I just think it’s magic to talk to people out of my own power with stuff I (at least partially) built myself.
    If you ever tried an Elecraft radio, their CW filters are really good and they can do pretty much the same on device what this AI model did. N7KOM has a short/video on his channel where he showed it off recently.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      I agree to an extent. My love is field radio. Some of these clips didnt need it. Its a good tool to have if needed.

  • @davidc5027
    @davidc5027 Год назад +3

    Sounds very similar to the noise reduction I get on my radio. Still, is very nice, and not saying anything to the contrary.

  • @geirha75
    @geirha75 Год назад +1

    Thanks for excellent information again, Colin. In your point of view Colin, are there any drawbacks using this software? I mean, how much will it cost to upgrade a desktop computer in order to gain the best results?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Not many drawbacks, just a slight delay by the time the signal comes back. You dont need any sort of fancy PC. I had it running earlier on my ancent laptop. All the heavy lifting is done on the £4K server.

  • @mrtechie6810
    @mrtechie6810 Год назад +2

    Nice. But needs to run locally on affordable hardware. Baby steps.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      In time, yes. Remember it's free and and the amateur radio community and contributing by submitting audio samples to make it even better. Its not ready to go in a radio just yet.

  • @yowsa52
    @yowsa52 Год назад +1

    Amazing. It is tiring listening to noise anyway whether it’s a signal in the noise or out of the noise. Whew boy. Take my money.

  • @cthoadmin7458
    @cthoadmin7458 Год назад +3

    Be interesting to know how much computing power this takes on the server, and whether it will be possible one day to run this from within a rig...

    • @dennis8019
      @dennis8019 Год назад +1

      lots of power.. but noise and hiss are part of the charm of radio ... either u are hearing it or not on radio waves. what's next a direct line to each ham?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      The way technology is going, you never know. That would be my wish.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Not without propagation there isnt! Its a tool in the arsenal for those who want it.

    • @hb9vqq533
      @hb9vqq533 Год назад +3

      The AI model runs on a Nvidia GeForce RTX 4090 GPU. Not something the average radio amateur has in his shack

    • @cthoadmin7458
      @cthoadmin7458 Год назад

      ... and you wonder what AI will be able to do next... Pull the SSB audio out of a noisy environment no human operator could ever distinguish. Speech recognition that "Repairs" the audio by guessing what half obliterated words probably were by the sounds and context. "You're 1 and 1 old man..."

  • @M0JHN
    @M0JHN Год назад +1

    Really surprised that the result is in fact not snake oil.

  • @DonzLockz
    @DonzLockz Год назад +1

    Thanks Colin. Why can't it take out Council Worker noise?😂
    Wouldn't it be awesome if this could finally be cleaned at a local location without internet but the filtering as crazy good.👍🤠

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      if it was only that good! Yeah hopefully it can be done locally eventually.

  • @SHTFchef
    @SHTFchef 9 месяцев назад +1

    If one needs an internet connection, it isn't radio.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  9 месяцев назад

      Thats whats good about radio, we all like different things. If we didnt then there would be no innovation.

  • @leroymay8156
    @leroymay8156 Год назад +3

    Seems there is no Linux support. Too bad, but Windows is not used by me.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Its been a 1 man show until recently so understandable.

  • @turinskiboy
    @turinskiboy Год назад +1

    Nice. Wondering if it is accessible for all amateurs globally?

  • @colinmcdonald7591
    @colinmcdonald7591 Год назад +2

    Does a good job. Still has that inherent digital loss that makes intelligibility of voice difficult, sort of the DMR effect. I think it's excellent on CW though and that is where it makes the most difference. I guess one positive effect is that it reduces or removes the hash noise that can cause ear fatigue over long periods of time. The TS590SG has pretty good noise reduction onboard, but it has that watery swishy sound whereas this RM noise software doesn't seem to have that at all.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +1

      Yeah its not perfect. Agree its better than the Kenwood's.

    • @z3r0c00l2
      @z3r0c00l2 Год назад +1

      I got Kenwood 590sg too ... I hate that watery sound ...

    • @stewartrv
      @stewartrv Год назад

      Yeah but still better then using DNR on my Yaesu 991a!

  • @michaelwatts8476
    @michaelwatts8476 Год назад +1

    Why would this not work on FT8?

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Its not been designed for FT8 but thats a completly different kettle of fish as its detecable below the noisefloor anyway.

  • @mikeramsey9747
    @mikeramsey9747 Год назад +2

    Impressive difference between the sound bites with and without the noise reduction.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад

      Its not miracle software but can make that importaint difference.

  • @AlvinMcManus
    @AlvinMcManus 10 месяцев назад +1

    So, basically eliminating my Heil Pro 6 Elite headset! I like my Heil's. NT 73

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  10 месяцев назад

      Just plug them into your computer, Thats what I do.

  • @charlieoscar09
    @charlieoscar09 Год назад +2

    BHI Will be very worried 😮

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +2

      They are doing it locally which is one bonus. No internet required.

  • @johnratcliffe6438
    @johnratcliffe6438 Год назад +3

    That's ridiculous. It seems much more effective than DNR on any of my Yaesu radios. A liferaft to save us from ever increasing QRM.
    I shall have to have a go.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +2

      I think it will get better and better. You can upload a custom noise sample too for the AI to use.

    • @johnratcliffe6438
      @johnratcliffe6438 Год назад

      @MM0OPXAmateurRadio I have got Station Master software a few weeks ago, but no real chance to use it. I shall give this a go.

  • @noomnoom8469
    @noomnoom8469 4 месяца назад

    Apologies the council are emptying the bins. Could that not be AI edited out?

  • @michaelcarey
    @michaelcarey Год назад +1

    Impressive

  • @johnwest7993
    @johnwest7993 Год назад +13

    I see the fundamental value of ham radio as being able to communicate independent from outside systems. That's why I don't bother with any internet-linked systems. If I'm going to have to use the web for radio work I might as well just use the Internet and not bother with the radio. It's a shame they don't have a system that can operate independently, as it could give my ears a rest to switch from I to AI. Finally, this is a hobby I engage in to learn more about RF communications and its various aspects, the physics, the software. So where's the source code? I can't learn anything about digital noise elimination from remotely located black-box software that just makes me dependent on unknown outside systems. So why is it being done this way, and once everyone is getting everything processed this way will it still be free? This reminds me too much of the free worms on fishing hooks.

    • @hb9vqq533
      @hb9vqq533 Год назад +1

      RM Noise is not an education program nor is it OSS. Neural Networks are well documented. Feel free to start your own project. It will be free a s long as volunteer radio amateurs like myself will carry the costs of purchasing the server hardware and the operational costs for the benefit of the entire community.

    • @yowsa52
      @yowsa52 Год назад +2

      I agree and fell the same way it’ll be a matter of time that perhaps radios will have the tech included or in the least have it on our computers independent of off site servers.

    • @kc8wvg
      @kc8wvg Год назад +3

      I have been using a bhi DSP Compact inline device. To reduce received noise on my SSB communications. It is quite effective. I recently told someone at DXE that it is almost like cheating to pull weak stations out of the noise.

    • @MM0OPXFieldRadio
      @MM0OPXFieldRadio  Год назад +3

      Nonsense, We all do what we want to. Did he also say FT8 is cheating?

    • @kc8wvg
      @kc8wvg Год назад +2

      @MM0OPXAmateurRadio if you're questioning my comment, I merely said that improving reception is a great idea no matter how it is achieved. The Bhi unit that I use makes such an improvement for me and my ears. I think that the idea is to make a QSO and enjoy the DX or talking to a friend. So, using a multiple element beam would also improve your SNR, but location, space, or money might limit that choice, whereas enhancing the reception is the way to go.

  • @anthonyspina4093
    @anthonyspina4093 Год назад +3

    Only runs on Windows -- too bad!

  • @JayN4GO
    @JayN4GO Год назад +1

    Won’t be long and the big 4 will have this internal capability

  • @youtubeaccount931
    @youtubeaccount931 Год назад +1

    ok ok this is really cool

  • @johnnorth9355
    @johnnorth9355 Год назад +1

    There was me thinking it was something to swat the idiots who disrupt the airways ? Ah well we can but hope.

  • @robbannstrom
    @robbannstrom Год назад +1

    Test