The increased demand for timber as a sustainable alternative to other construction materials has resulted in a shortage of natural resources available for sawmill conversion. This pushes industries to employ more rational and efficient technologies to increase production yields and minimize residuals. Several scanning technologies are integrated with state-of-the-art production facilities, although the majority determine log suitability based on dimensions and/or surface defect detection. It has always been the ultimate desire of sawmillers to see the interior of a log before the first cut to optimally determine the sawing pattern and estimate product quality/value early. Such technology, industrial CT, is now available but only to large companies that can afford the very high investment costs. A more affordable alternative is to use classical X-ray radiometry to scan logs in one or more cross-sections perpendicular to the feed direction. An alternative configuration is to tilt the direction of the X-ray beam away from perpendicular. The radiogram of a log generated with such a setup is not as intuitive for interpretation but can be a highly advantageous addition to an automated image analysis system. Assuming some simplifications and implementing appropriate machine learning algorithms, it may be possible to offer a scanning solution similar to CT at a fraction of the cost. The goal of the research presented was to build a digital simulator of the X-ray radiometer to test the feasibility of different configurations and to identify all limitations of the simplified tomography solution.