After much heartache, gnashing of teeth and hair pulling, we have come up with the following work around to dealing with the grid and the volume of points produced.
We choose a contour line interval of 0.25’ and save the contours as a dxf file. Then we just open a new drawing, import the contour lines and use them to create a new surface. Then we just cut and paste the surface only into the drawing we created for our project. Once we have the surface we can choose any contour interval we want.
My logic is that the small contour interval picks up all of the irregularities in the surface no matter what the grid density used. I believe it to be a much better way to transfer the data. We have field checked dozens of surfaces and generally find the accuracy to be under 0.10’. A grid does not do well with curb and gutter details. My method picks up every small change in elevation. We have tried using an even smaller contour interval but I don’t believe it really helps the surface any.
Our only headache has been the overlap between drone flights. We shoot a lot of roadways for topography. We use a lot of GCP (250’ interval pairs) and try for at least 500’ of overlap. We usually fly and process in about 4000’ segments with 500’ of overlap. We then try to find the best fit of the contour lines that match between flights and trim each contour and surface to match. Then we just combine it into one large surface.
Hope this helps.