Face verification in the simultaneous presence of age progression and changing face-image quality is an important problem that has not been widely addressed. In this paper, we study the problem by designing and evaluating a generalized Q-stack model, which combines the age and class-independent quality measures together with the scores from a baseline classifier using local ternary patterns, in order to obtain better recognition performance. This allows for improved long-term elms separation by introducing a multi-dimensional parameterized decision boundary in the score-age-quality classification space using a short-term enrolment model. This generalized method, based on the concept of classifier stacking with age- and quality (head pose and expression) aware decision boundary compares favorably with the conventional face verification approach, which uses decision threshold calculated only in the score space at the time of enrolment. The proposed approach is evaluated on the MORPH database.