Waterfall filtering? [added in v1.337]

edited November 2019 in Problems Now Fixed
What about implementing some simple vertical filtering for the waterfall? I mean something that vertically blurs the graininess. Several other SDRs (e.g. mchf, ELAD FDM, ...) have it, should not be difficult to implement.


  • One thing I've noticed with the waterfall, is just how much improved the view looks when changing the zoom level outwards.
    For example, if I'm at z9 looking at an AM signal, and I have a full waterfall showing in the window, and I go to z8 the visible waterfall (to me) has a better aesthetic while still showing all of the details that I've already been seeing. There's a very obvious border between the zoom levels with the new waterfall data appearing much more grainy than the old waterfall below.

    Maybe an option to have the waterfall data use the next zoom in, but twice the width, for the waterfall? Would that have much impact on the processing requirements either on the BB or the browser?
  • Today's v1.337 update adds waterfall filtering (averaging). The filtering methods and adjustments parallel those used for the spectrum filtering, although the range of slider values is slightly different based on practical experience. The waterfall and spectrum filtering are independently adjustable.
  • Hi John,

    I've tried the new filters, although I'm not sure if I like the end result, but that may be just me.

    To my eyes the waterfall looks very 'blurry' with even the mildest settings and weak signals seem to get lost among the 'smearing'.

    SDR Sharp has 'sliders' that allow you to set the waterfall and spectrum display 'attack' and 'decay' times.

    This works very well, as it allows you to catch short duration signals but they remain visible for sufficient time to be easily spotted on the waterfall display, yet the random background noise is still filtered out.

    I've varied a few SDR Sharp attack and decay settings as the waterfall scrolls down in the example below.

    Just my observation, your mileage may vary.


    Martin - G8JNJ
  • Yes, that is what I expected (nothing but blurriness) from applying simple averaging to the waterfall. The original post understood this when he said "something that vertically blurs the graininess". I find using the IIR with a gain of about 1.0 to be the most useful. I don't have the time to attempt anything more complicated at this point.
  • With the recent change (1.337) I notice that the noise graphs at the Northern Utah WebSDR (KA7OEI-1) on Grafana and WSPRDaemon.org have shifted upwards (noisier) by between 4 and 5dB.

    I presume that this has something to do with the change in the most recent firmware, perhaps having to do default waterfall averaging? If so, a uniform shift like this is not what I would have expected with averaging. Is this an expected artifact?

  • edited November 2019
    I can confirm that the s-meter on the web interface shows the noise floor also by about that amount higher
  • I see what's happened. Fix coming in a few hours..
  • jksjks
    edited November 2019
    Okay, v1.339 should restore things to the way they were.

    What happened was that the reduction in CIC filtering required to get 14-channel mode for the BBAI to fit in the FPGA leaked into the v1.337 build of the other FPGA configurations (4, 3 and 8-channel modes).
  • That addition is helpful, when I'm searching for NDBs the averaging makes the morse components stand out a little better against the background.
  • Hi,
    thank you for taking my suggestion into account!

    Unfortunately it does not seem to work here. The slider "WF Delta" cannot be moved. Am I missing anything? I use Firefox.

  • The slider min/max is adjusted to fit the averaging currently in effect. But if the averaging mode is disabled then it will not move. E.g. if the button is set to "WF Delta" but the WF menu is set to "off". Try another menu mode like "IIR".
  • Ah! Thank you! This is a very big improvement, the graininess goes away and the signals look much better to me.
    BTW what's the meaning of IIR/MMA/EMA ?
  • jksjks
    edited November 2019
    MMA/EMA: Explained in the 7th post in this thread: http://forum.kiwisdr.com/discussion/comment/6601
    IIR: An "infinite impulse response" filter as nicely explained by Alex, VE3NEA: www.dxatlas.com/rocky/advanced.asp
Sign In or Register to comment.