I experienced the same phenomenon on .040" inconel 718 square groove butt weld 1g position. X -ray indication was a straight line down the middle of a weld that was obviously penetrated. The suspect weld was then sectioned in seven areas where the indication appeared assuming we could better characterize the indication at 100-200x . After metallographic prep including electrolytic etch using 10% oxalic acid, guess what? there was no indication of lack of fusion or hollow bead, only grain boundaries that were thick and of the typical columnar arrangement in a weld bead and that were perfectly alligned in the center of the bead. The main obvervation of the weld was that it was welded with a faster that normal travel speed. It is still hard for me to imagine how this caused enough change in density for it to cause a distinct RT indication ,but it happened . My best advice is to take a look at travel speed, if ripples are elongated or v shaped slower speed should be used