A mechanism is presented for editing a video stream as the stream is received, allowing real-time displaying and storing of the edited video stream without requiring the received video stream to first be recorded. The mechanism allows the selection of editing options prior to starting or during the record process. The mechanism can monitor the input video stream for specific characteristics and edit the video stream according to predefined editing options based on those characteristics and/or user inputs. Special effects can be intelligently selected based upon internal and external information about the video stream, for example, user actions, input source, and content information.