HISE Logo Forum
    • Categories
    • Register
    • Login

    Noise Cancelling

    Scheduled Pinned Locked Moved General Questions
    9 Posts 3 Posters 561 Views
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • C
      clumsybear
      last edited by clumsybear

      I'm trying to implement a routing system to eliminate the self noise of the ShapeFX Module.
      I know @Christoph-Hart, I know it's below a certain dB value and shouldn't be a concern, but I can't help it, I have to get rid of it ;)

      So what I want to do, is simple - route the shape fx noise (without a signal) to a separate path via the routing matrix, flip the phase and mix it back with the signal that runs through the shape fx module. In theory that should result in a distorted signal without the Shape fx noise.

      Now, I can't think of a way to route the isolated noise before the signal hits the module. So, the only way would be to run a second shape fx module without any signal, flip the phase and mix it with the first shape fx (which is used for the signal). But those two modules have different "noise profiles" I guess.

      Any suggestions?

      1 Reply Last reply Reply Quote 0
      • Christoph HartC
        Christoph Hart
        last edited by

        What self noise? The algorithm has no non-deterministic component it‘s just predictable input-output modification so there is no noise that you could reduce.

        It‘s way more likely that you mess up the phase and create comb filter artifacts at your attempt to remove noise which isn‘t there.

        1 Reply Last reply Reply Quote 0
        • C
          clumsybear
          last edited by clumsybear

          That's the analysis of the tanh function of the shapefx:
          Screenshot 2019-09-16 at 08.53.16.png

          That's the analysis of the tanh function of the result I'm trying to get:
          Screenshot 2019-09-16 at 08.54.36.png

          I'm not complaining about anything, just trying to figure out why the analysis differs so much from the second picture. Maybe I'm getting something wrong, sorry for being annoying :)

          1 Reply Last reply Reply Quote 0
          • Christoph HartC
            Christoph Hart
            last edited by

            How did you create the second signal?

            1 Reply Last reply Reply Quote 0
            • C
              clumsybear
              last edited by clumsybear

              It’s a distortion plugin from another developer for reference. Not sure how they achieved this. Maybe it’s another dsp algorithm

              1 Reply Last reply Reply Quote 0
              • Christoph HartC
                Christoph Hart
                last edited by

                It's actually the filter in the shaper that causes this noise floor (it's impossible for the tanh function to generate noise as long as the system library implementation of this function doesn't throw in a random generator for no reason).

                This might be related:

                https://forum.juce.com/t/dsp-module-iir-filter-causing-low-freq-noise/34416

                However I am also strongly on the -100db doesn't matter side. But please try scriptnode, you'll get a pure tanh function as node there.

                1 Reply Last reply Reply Quote 1
                • C
                  clumsybear
                  last edited by

                  I understand, thank you @Christoph-Hart

                  Yeah, normally I'm on the -100db doesn't matter side as well, but not while developing complex distortion fx with multiple stages. I will try the scriptnode tanh for that

                  1 Reply Last reply Reply Quote 0
                  • ustkU
                    ustk
                    last edited by

                    @clumsybear I am currently developing my own FX library using scriptnode, but for distortion, I don't know where to begin... Do you have any advice on what I should do/read? Again, I am using scriptnode only so I'm no C++ guy, unfortunately :)

                    Can't help pressing F5 in the forum...

                    1 Reply Last reply Reply Quote 0
                    • Christoph HartC
                      Christoph Hart
                      last edited by

                      If you don't go into analogue circuit modelling, the basic building blocks of distortion are oversampling, filtering and all kinds of wave shaping.

                      1 Reply Last reply Reply Quote 1
                      • First post
                        Last post

                      49

                      Online

                      1.7k

                      Users

                      11.7k

                      Topics

                      101.8k

                      Posts