The traditional format of scientific journals cannot provide an adequate description of modern research techniques and this creates a critical problem of knowledge transfer that severely impacts research and education. Recent experiments indicate that 70% to 90% of studies published in scientific peer-reviewed journals are not reproducible. Even if partially true, these findings present difficult questions about the future of scientific research. We believe that this phenomenon is due to the way scientists communicate. The solution is to provide scientists with a new method of communication and the growing world of video publication does just that. A video-based approach to scientific publishing provides systematic visualized publication of experimental studies. This presentation will discuss the development of scientific video publication, its technical challenges, implications for scholarly communication, acceptance in the scientific and library communities, and first-hand accounts from users of this medium. Results from recently conducted case studies will be shared, such as the experiences of research groups at Purdue University and University of Alaska which indicate that video publications can save a lab up to $15,000 per experiment.