This paper presents a novel transformation we call Subspace Supported Rational Transformation (SsSRT). Despite the structure of this transformation seems to be similar to Mobius transformation, SsSRT differs from Mobius transformation due to its enhanced applicability to vectors, matrices and folded matrices. This transformation has been developed during the studies of digital image processing via Tridiagonal Folmat Enhanced Multivariance Products Representation (TFEMPR) in order to increase approximation quality. In this work, Transformational Tridiagonal Folmat Enhanced Multivariance Products Representation (TTFEMPR) enhanced by using SsSRT is taken into consideration as a truncation based approximation method. The SsSRT flexibilities have been determined via an optimization at the constancy level TFEMPR approximation.