Navigation

    Duet3D Logo

    Duet3D

    • Register
    • Login
    • Search
    • Categories
    • Tags
    • Documentation
    • Order

    Intelligent filtering of Mesh Compensation datapoints?

    Firmware wishlist
    3
    5
    775
    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.
    • Yonkiman
      Yonkiman last edited by

      I'm using the IR sensor (great!) with a PEI bed (uh-oh…). I followed the instructions for painting and curing the back of the PEI, and while it took a few tries, I finally got it glued to my aluminium bed, and it looks black enough for Nigel Tufnel and it is flat enough to print on most of it with no bed compensation.

      But when I try to do do a mesh bed compensation and , I get some bad datapoints. I don't know if this is caused by flaws in the paint or scratches on the surface, but I get some bad datapoints:

      I imagine my immediate solution is going to be sanding the top of the PEI and losing that nice, smooth finish. But I was thinking that a median filter (https://en.wikipedia.org/wiki/Median_filter) should be able to do a really good job of throwing out the outliers and making even that data usable. It might even be beneficial on bed maps not as pathological as mine.

      My other (related) suggestion is an option for a single G30 probe command to probe multiple points and average them.. Ideally you'd add NUMBER_OF_POINTS and RADIUS to the probe command, and the command would then probe that number of points around the given X,Y. I think that would be helpful (increasing accuracy and repeatability) no matter what kind of probe you're using.

      Should I add either to the firmware wishlist?

      1 Reply Last reply Reply Quote 0
      • dc42
        dc42 administrators last edited by

        Those both look Iike good ideas to me, worth adding to the wish list.

        Meanwhile, you could import the heightmap.csv file into a spreadsheet, apply a median filter to it, and re-export it. Then load it using G29 S1.

        Duet WiFi hardware designer and firmware engineer
        Please do not ask me for Duet support via PM or email, use the forum
        http://www.escher3d.com, https://miscsolutions.wordpress.com

        1 Reply Last reply Reply Quote 0
        • Yonkiman
          Yonkiman last edited by

          @dc42:

          you could import the heightmap.csv file into a spreadsheet, apply a median filter to it, and re-export it. Then load it using G29 S1.

          This is why you're my hero. 🙂

          1 Reply Last reply Reply Quote 0
          • dc42
            dc42 administrators last edited by

            Bear in mind that in the height map file, plain 0 with no decimal point means that the point was not probed. So don't change those values, and don't use them in the median filter calculations.

            Duet WiFi hardware designer and firmware engineer
            Please do not ask me for Duet support via PM or email, use the forum
            http://www.escher3d.com, https://miscsolutions.wordpress.com

            1 Reply Last reply Reply Quote 0
            • roboduet
              roboduet last edited by

              My other (related) suggestion is an option for a single G30 probe command to probe multiple points and average them..

              The new is a well-forgotten old.

              +1 for this feature. The same problems with IR sensor forced me to stop using it.

              1 Reply Last reply Reply Quote 0
              • First post
                Last post
              Unless otherwise noted, all forum content is licensed under CC-BY-SA