It is a widely spread statement nowadays that electronic and online technologies offer surprising and unlimited prospects for the mankind. However today’s distance learning courses do not always meet modern requirements. Their evaluation by the participants of an educational process is contradictory. The article presents materials of the research conducted by its authors in Ural Federal University in spring and autumn of 2018. The results of the survey held among the students and lecturers of the Institute of Natural Sciences and Math as well as students and lecturers of other Institutes of the University offering majors in engineering served the basis for the research. There were obtained data on peculiarities of perception by students of online lecturers as well as data on forms of representation used in such courses. It allowed to reveal the existing problems and to prove the need of developing distance learning taking into account psychology of perception. The authors offer such a direction of development which will increase the level of visual and sound perception of E-courses and will allow to raise not only content, but also forms of electronic learning process to a new level of adaptability and openness. The article shows that machine learning technology aimed at active modification of a visual and sound image of online lecturers can become an instrument of improvement.