Menu

#43 new distorsion types and edge rounding limiters

2.5.2
closed
nobody
5
2020-04-30
2019-06-09
friedolino
No

this three patches enhance the distorsion module with

tunable edge rounding for limiter and inverse limiter function
3 new distorsion types square, cubic and tanh
dc offset at the input of the transfer function
new function parameter to shape the function or the edge rounding

Discussion

1 2 > >> (Page 1 of 2)
  • Johannes Lorenz

    Johannes Lorenz - 2019-06-17

    Is there any order these patches would need to be merged, or are they independen?

     
    • friedolino

      friedolino - 2019-06-18

      they patches without prefix have to be merged in in ascending order 0001 - 0003
      The rtosc patch is needed to compile them. The mrz and osc_bridge patches are needed to reflect the new controls in the new gui.

       
  • Johannes Lorenz

    Johannes Lorenz - 2019-06-20

    I'll review and merge this.

     
  • Johannes Lorenz

    Johannes Lorenz - 2019-06-21

    friedolino: the rtosc patch contains

    #define DOC_IMP13(a,b,c,d,e,f,g,h,i,j,k,l,m) a b c d e f g h i j k l rDoc(j)
    

    Are you sure you did not mean rDoc(m)?

    Aside from that, the patch looks OK to me.

     
    • friedolino

      friedolino - 2019-06-21

      yeah. Thats the problem of copy/paste coding style.

       
  • Johannes Lorenz

    Johannes Lorenz - 2019-06-21

    First of all: Many thanks for the distortion patches.

    Looking at your first zyn patch (0001):
    While you can infer the meanings of the parameters, a quick description of each would likely not harm
    Can you give me the source for your formulae, so I can check easily whether they are all valid?
    Why is the polyblamp not used for e.g. the square wave? Is it specific for limiters, or did you just not implement it there?
    Can you please explain your change for distortion number 14?

     
    • friedolino

      friedolino - 2019-06-21

      in number 14 I added support for the DC offset parameter.
      Because the offset is added to input and subtracted from the output, it behaves like the distorsion function is shifted, making it asymetric. I did the same also to numbers 1, 7 and 10.

       
  • friedolino

    friedolino - 2019-06-21

    Ok, I will add more comments.

    Polyblamp could be applied to square wave, too. But on the effects it is more needed because often there is no more LP filtering afterwards.

     
  • Johannes Lorenz

    Johannes Lorenz - 2019-06-21

    About the square wave: Sorry, I confused wave and distortion.

     
  • Johannes Lorenz

    Johannes Lorenz - 2019-07-06

    Are you finished with adding the comments, so I can continue?

     
  • friedolino

    friedolino - 2019-07-09

    i am a little bit short of time these days. i hope, i get to it this weekend

     
  • friedolino

    friedolino - 2019-09-15

    I added a patch 0004 with more comments and source information.
    sorry that it took so long.

     
  • Johannes Lorenz

    Johannes Lorenz - 2019-12-09

    What is the master base commit for the osc-bridge patch?

     
  • friedolino

    friedolino - 2019-12-22

    fix for an error in the offset handling of the limiter function.
    and added the new waveshapers to oscilgen parameters (used in zest)

     

    Last edit: friedolino 2019-12-22
    • Mark McCurry

      Mark McCurry - 2020-04-26

      I'm seeing parts of this integrated. Are there still relevant pieces
      of this which need discussion prior to merging?
      If there are still open bits, can we consolidate this into a github
      PR, (it at least feels like discussions there go a bit faster).

      If everything relevant is merged, let me know and I'll shift this to closed.

       
      • friedolino

        friedolino - 2020-04-27

        all of theese patches have already been integrated on github in https://github.com/zynaddsubfx/zynaddsubfx/pull/41.
        This Ticket can be closed.

         
  • Johannes Lorenz

    Johannes Lorenz - 2020-04-27

    friedolino: Were the mruby-zest patches also be merged? I can't recall.

     
    • friedolino

      friedolino - 2020-04-27

      yes, at eb8fa13f4f2a8da42233179f1d6861bed4ec5a5d

       
1 2 > >> (Page 1 of 2)

Log in to post a comment.