Lecture 26AnnouncementsPlan for TodayVideo Conferencing DiscussionDigital SignalsDigital SignalsDigital Signal ProcessingDigital CommunicationsCourse SurveysLecture 26The Digital World of MultimediaProf. Mari OstendorfEE299 Lecture 2612 March 2008Announcements Final Project Upload your presentation to CollectIt!!  Graded by prof (0-5) [points deducted for late submissions] Presentations on Friday 3/14 in extended class (3:30-6pm, EE403) 8 min presentations (6-7 min talk, 1-2 min for questions) Peer grading (0-5) Written reports (2-3 pages) due to CollectIt by 3/16 5PM. Graded by prof (0-10) Summarize key points of the paper and identify open questions Include a bibliography if you refer to other work Grading: technical quality (5), organization and completeness (3), style (i.e. no typos, grammatical errors, etc.) (2) Grade checking: Grades will be emailed Thursday night. Corrections needed by Monday noonEE299 Lecture 2612 March 2008Plan for Today Video conferencing discussion Course highlights Course surveysEE299 Lecture 2612 March 2008Video Conferencing Discussion As a speaker: What was challenging? Would you organize your presentation differently for the in-person presentation. As a listener: What presentation strategies are helpful? How is the technology distracting? Either:  Do you think that it was better than a conference call? What sorts of changes do you think are needed before video conferencing is more widely used?EE299 Lecture 2612 March 2008Digital Signals Digital signals:  Represent time/space in terms of discrete points (samples/pixels) Represent amplitude in terms of a set of discrete values  Digital signal processing: keys to success Shannon’s sampling theorem Leveraging imperfect human perceptionEE299 Lecture 2612 March 2008Digital SignalsIdea: signal building blocks - Works for different types of signals (audio, image, ECG, …) - Works for other component signals (besides cosines)x(n)Time ÅÆ FrequencyX(f)x(n,m)X(f1 ,f2 )EE299 Lecture 2612 March 2008Digital Signal Processing Using signal building blocks Work in the domain that’s best for the problem  Divide and conquer! Signals can be: Generated Manipulated (to create nicer sound/looking signals, or just something new) Compressed for more efficient storage Encrypted and/or watermarked Communicated Interpreted or understoodEE299 Lecture 2612 March 2008Digital Communications Transmission: noise happens Adding redundancy makes it easier to detect and correct errors Signals that are farther apart are easier to distinguish Networks: delay, congestion happens Need to code for loss as well as errors Ad hoc networks: collaboration happens Need to code for dynamic topologyEE299 Lecture 2612 March 2008Course Surveys Repeat of the initial course survey Help us understand what you learned & what you value Help you appreciate how much you’ve learned Standard UW course