In a strongly competitive context, one of the main stakes for the actors of the audiovisual services offer is to guarantee to the spectator an optimal Quality of Experience (QoE). Nowadays, QoE is often limited to the perception of the audiovisual quality (AVQ) received by the system. It is mainly measured through the collection of rates given by testers onto quality scales, after visualization and listening of the AV sequences treated by one or several technologies to be evaluated. These subjective tests are following procedures recommended by the International Telecommunication Union. However, the restored quality can affect some factors of QoE (fatigue, effort, etc.) which are not reflected by the quality scores. A method considering the evaluation either of the AV quality only received but of the quality of experience, widely considered, could allow to report better the influence of the sound and image quality on the spectator. The present work is centered on the research of an alternative method to current methods of quality assessment for multimedia applications in a context of viewing/listening of 2D or 3D AV contents. The proposed approach addresses the QoE in terms of analysis of subjective indicators and tonic physiological (electrodermal activity, heart rate, peripheral cutaneous temperature, blood volume pulse) and oculars indicators (PERCLOS, duration and frequency of the blinking of the eye, number of saccadic movements, pupillary diameter). Physiological and ocular measures have for advantage not to be subjected to the biases of the subjective measures (representativeness, scales, etc.) and to reflect phenomena such as fatigue or mental effort, potentially induced by the presence of audio and/or video degradations, which may be critical in terms of QoE. Two protocols were tested to study the relevance of this approach. The results showed that QAV modulates the subjective measures and have putted forward the insufficiency of quality rates to reflect faithfully this effect. The impact of the quality on the physiological and ocular measures is less obvious. A set of factors in particular connected to certain attributes of the test contents, as the dynamics or the luminosity, would have been able to mask or decrease the quality effects observation on gathered measures. However, two of the physiological indicators reacted to the presence of audio and/or video degradations when these were accumulated to the detrimental effect of other factors (3D video or test period effect).
