Presence of knots and associated fiber deviation are crucial for engineering properties of sawn timber. Yet, there is a notable absence of a thoroughly calibrated and verified mathematical model for fiber directions. This gap is largely due to the lack of comprehensive and detailed experimental data on growth surface geometry and 3D fiber orientation. Such data, ideally extracted at the sawn timber level, should include diverse information related to single knots, multiple knots, knot clusters, and both live and dead knots. This study presents a comprehensive laboratory examination of a full-size Norway spruce timber board. The extraction of knots, growth surfaces, and full-volume 3D fiber directions was successfully achieved, yielding highly detailed experimental data. The method developed comprises X-ray computed tomography for 3D knot and growth surface geometry, and optical scanning utilizing the tracheid effect for in-plane fiber directions. A limitation was identified when the normal vector of growth surfaces and the normal vector of the optically scanned board surface are orthogonal but a sensitivity analysis revealed that an angle error introduced to the in-plane fiber directions has limited impact on the computed 3D fiber vectors when the angle between the two normal vectors is below 60 degrees. The 3D knot, growth surface geometries, and fiber patterns observed in this study clearly align with the patterns revealed by a previous micro-CT study. The method and data obtained are valuable for the subsequent development of a more refined and rigorously calibrated fiber angle model than those currently available.