Announcement

Collapse
No announcement yet.

WinFilters - a new application for learning about filters

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • WinFilters - a new application for learning about filters

    Edit: The corrected program is ready to download. The errors were actually rather small.


    I've created a new application that is primarily a learning tool. While verifying the Bessel filter targets built into WinPCD, I found that there is more that is interesting about the Bessel filter. The implementation in WinPCD is currently for the typical Bessel filter. As Rane describes it "The textbook low-pass Bessel is often designed for an approximate time delay of t0 = 1/w0 rather than for the common -3 dB or -6 dB level at the design frequency used for crossovers." I followed this in setting up the Fc separation for the "types" of Bessel filter combinations in the program. The Bessel filters are all the same algorithm for any given frequency, the only difference is the Fc of each section. The "Bessel" filter is the one for the textbook time delay. The others are Fc "spreads" of this.

    The other two filter types are the typical Linkwitz-Riley and Butterworth. The Bessels only go up to fourth order while the others go up to eight. The group delay for the lowpass and the summed response are shown as well, though not the absolute magnitudes. Relative offset (excess-phase) can be added to the woofer (set farther back) or removed from the tweeter. The latter isn't really "closer" because the default phase is the minimum-phase for the filter. What it will show is that one, the summed response is the same with either driver moved and two, the phase will "curl up" as it would if you removed too much excess-phase from a tweeter measurement.

    One thing that I found interesting is just how close to flat (on-axis) the Bessel can be using the "flattest response" Fc spread. The off-axis will have a small lobe not on the main axis, just as an odd-order Butterworh does, however. After I update WinPCD for these new filters, it will be interesting to see the horizontal and vertical off-axis graphs.

    The link to the new page is at my site.

    One last note. You can export the LP and HP files to use as target files in other software. You set the export SPL (default is 90db), right click to bring up the context menu and select which filter to export.

    dlr Click image for larger version

Name:	WinFilters.gif
Views:	1
Size:	65.0 KB
ID:	1274469
    Last edited by dlr; 03-25-2016, 10:01 PM.
    WinPCD - Windows .NET Passive Crossover Designer

    Dave's Speaker Pages

  • #2
    thanks for this, an interesting app to play with.

    FWIW you don't seem to have a link to the page in this post, unsure if that is intentional but I had to hunt around a bit to find the download.

    Comment


    • #3
      Originally posted by 3ll3d00d View Post
      thanks for this, an interesting app to play with.

      FWIW you don't seem to have a link to the page in this post, unsure if that is intentional but I had to hunt around a bit to find the download.
      Here's a direct link: WinFilters

      dlr
      WinPCD - Windows .NET Passive Crossover Designer

      Dave's Speaker Pages

      Comment


      • #4
        Dave - Thanks, WinFilters sounds exactly like what I am searching for. But I can't seem to find the zip file on your website. What am I missing?

        Ed
        "Everything is nothing without a high sound quality." (Sure Electronics)

        Comment


        • #5
          Originally posted by ernperkins View Post
          Dave - Thanks, WinFilters sounds exactly like what I am searching for. But I can't seem to find the zip file on your website. What am I missing?

          Ed
          Somehow I deleted the download line when I was updating the page. It's there now.

          dlr
          WinPCD - Windows .NET Passive Crossover Designer

          Dave's Speaker Pages

          Comment


          • #6
            Got it and thanks again. The only hiccup is that the "Reference SPL" value doesn't seem to work - I always see values referenced to 90 dB in the frd file. But it still saves me a lot of time as I just pop into Excel and subtract 90 dB from all the values using a simple equation.
            "Everything is nothing without a high sound quality." (Sure Electronics)

            Comment


            • #7
              Originally posted by ernperkins View Post
              Got it and thanks again. The only hiccup is that the "Reference SPL" value doesn't seem to work - I always see values referenced to 90 dB in the frd file. But it still saves me a lot of time as I just pop into Excel and subtract 90 dB from all the values using a simple equation.
              It doesn't change the output relative peak? I'll have to fix that, it was a quick change, I must not have tested it completely. The normalized value is supposed to be adjusted by the value shown.

              dlr
              WinPCD - Windows .NET Passive Crossover Designer

              Dave's Speaker Pages

              Comment


              • #8
                If you're modifying the code: I also noticed that each file has a "0 90 0" as the last line. I didn't mention it before because REW ignores it.
                "Everything is nothing without a high sound quality." (Sure Electronics)

                Comment


                • #9
                  Thanks, I'll fix things. I'm not sure that anyone else has tried to use it to export to file.

                  dlr
                  WinPCD - Windows .NET Passive Crossover Designer

                  Dave's Speaker Pages

                  Comment

                  Working...
                  X