AN OVERVIEW OF DIFFERENT APPROACHES FOR LECTURE CASTING Wolfgang Hürst, Wolfgang Waizenegger Institut für Informatik, Albert-Ludwigs-Universität Freiburg, Germany Georges-Köhler-Allee, Building 51, 79110 Freiburg, Germany [email protected]
, [email protected]
ABSTRACT Podcasting is a special approach to distribute multimedia (usually audio files) over the internet. Users can subscribe to a particular podcast which is then delivered automatically for replay on their mobile devices (MP3 players, laptops) or desktop computers. Podcasts are not restricted to pure audio files but also include slide shows with synchronized audio replay and video files for mobile multimedia players (such as the new video iPod from Apple). In this paper, we address one particular usage of podcasts for education, i.e. the production and distribution of recorded lectures via podcasting. We describe different realizations of such lecture podcasts, compare their characteristics with established requirements from the general lecture recording scenario, and discuss different usage scenarios. KEYWORDS Lecture Recording, Podcasting, Lecture Casting, Video Casting.
1. INTRODUCTION The term podcasting describes a special way for the distribution of multimedia data (usually audio) over the internet. “Pod-” is a reference to the Apple iPod (Apple 2006). However, the reflection of the iPod’s name in podcasting is only motivated by the dominance of the respective hardware in the MP3 player market. In general, podcasting can be done with any kind of device which enables audio replay – including mobile MP3 players, laptops, as well as traditional desktop computers (Windows or Linux PCs and Macs). The ending “-casting” is derived from broadcasting. However, in contrast to broadcasting, podcasting usually downloads the respective content to the user’s device for local or mobile replay. These downloads can be initiated manually but more often, users subscribe to so called podcasts, i.e. document collections or “shows” that are normally focused on one particular topic. New documents or “episodes” are then downloaded automatically and uploaded to, for example, a mobile MP3 player as soon as this device is connected to the respective computer. The concept of podcasting dates back to the end of the year 2000 but really took off around 2004 (Wikipedia 2006a). In fact, the New Oxford American Dictionary declared “podcasting” the 2005 word of the year and, according to Wikipedia (2006b), defined it as “a digital recording of a radio broadcast or similar program, made available on the Internet for downloading to a personal audio player”. However, today podcasts are not only limited to audio content but include multimedia data, such as slide shows with synchronized audio replay and video for mobile multimedia players (such as the new video iPod from Apple) as well. In the same way as podcasting is not limited to any particular replay device, its realization is not based on proprietary technologies but builds on common, standardized approaches for media creation and delivery, such as RSS 2.0 or Atom for the subscription of podcast shows. The main reason why podcasting has gained such a tremendous popularity is the integration into a whole concept together with tools that support easy production, hosting, and distribution services on the producers’ side, as well as easy subscription, automatic download (to a computer) and upload (to a mobile player), and easy organization on the consumers’ side. In relation to e-learning, Podcasting offers significant potential for several reasons. Students are already using it, not only for entertainment purposes but also to gather information, for example, via podcasted radio news shows. Therefore, it offers an effective way for the delivery of multimedia learning material, such as audio
snippets from a foreign language course. In addition, new learning scenarios can be applied, for example, the exchange of acoustic annotations among students or between students and teacher. The required hardware for podcast production is often already available or can be purchased at rather low costs. Various educational podcasts already exist (see EPN 2006 or Productive Strategies 2006, for example), but their actual benefit for learning is still controversial. Different scenarios in which podcasting might be useful for education are described by Meng (2005) and Campbell (2005). In this paper, we want to focus on one particular application of podcasting in an educational context: the delivery of recorded lectures to students via podcasts. Both Meng (2005) and Campbell (2005) mention recorded courses as a possible usage for educational podcasts. Several lecture podcasts already exist (see Albert-Ludwigs-University Freiburg (2006a) or Productive Strategies (2006) for an extensive but by no means complete overview). Yet, there are also critical voices which question the usefulness of podcasted lectures (e.g. Reynolds 2005). Hence, it is not clear (yet) if podcasting recorded courses makes sense at all, and if it does, how it should be done to achieve the highest possible benefit for both teachers as well as students. Although there are already several podcasts available which contain recorded lectures, few actual data about usability or a reasonable (i.e. effective) production process exist. In this paper, we want to address this issue, that is, discuss if and how podcasted lectures might be a valuable complement to existing teaching services provided in higher education. First, we summarize the general lecture recording scenario and highlight respective requirements that have to be fulfilled in order to guarantee a reasonable production of lecture recordings as well as a gaining usage for the end users (Section 2). Then, we describe different options to realize lecture podcasts, i.e. podcasts containing recorded lectures, and evaluate if and how the requirements from the general lecture recording scenario can be met (Section 3). We conclude with a short description of our current work and an outlook to possible usage scenarios (Section 4).
2. LECTURE RECORDING Approaches to automatically record and distribute instructional talks such as classroom lectures have been the focus of research since the mid 90s (Müller and Ottmann 2003, Abowd 1999). Today, several commercial tools are available that enable teachers to automatically record their lectures, and post-process, and distribute them to the students with little (if any) need to manually interfere into this production and distribution process. While some early approaches for lecture recording started with video taping of the lecturer, it soon became evident, that capturing the presentation graphics, i.e. the presented slides together with annotations made on them, is a far more useful choice. They usually contain the (visual) essence of a talk, whereas a video recording often just represents the “talking head” of the lecturer or a rather low quality capture of the presentation graphics. Today, we generally assume that the audio stream with the voice of the lecturer and the presentation graphics cover the main information and essentials of most presentations. For example, Gemmell and Bell (1997) describe these two data streams as “critical media” for any (tele)presentation, and Lauer et al. (2004) present some empirical data supporting this statement. Therefore, nearly all successful systems for automatic presentation and lecture recording produce a recording that contains the presenter’s voice and the presentation graphics stream, whereas the video capture of the lecturer can often be added but usually is not necessary (and often omitted to save up bandwidth and disk space). Figure 1 presents an example of a recorded lecture with different media streams that are synchronously relayed in a media player. The lecturer’s view: Automation of the recording process. Although it is still controversial if recorded lectures can replace single classroom sessions completely, it is commonly agreed and has been verified in various evaluations that such documents make a good complement to existing teaching services (Hürst et al. 2006) if they are produced in a reasonable way, i.e. if the recording process is completely automated and the produced files fulfill certain characteristics which guarantee a high usability. From the presenters’ point of view, automation seems to be the most important issue. Lots of the earlier experiments with lecture recording produced very good, high quality documents by investing a tremendous effort into manual recording and post-processing. In contrast to this, several research groups (including ourselves, see Ottmann and Müller 2000) have always argued that due to limited resources (i.e. required time and man power) it is essential to keep the amount of required manual work to an absolute minimum. Reality has proven us correct. Most of
Figure 1. Example of a recorded lecture replaying the audio with the voice of the lecturer, the presentation graphics stream (right), the lecturer video (bottom left), and thumbnails of the slides for easy navigation (top left).
these approaches that relied on massive manual processing were only one-time projects that could not get established as a regular service. In contrast to this, lecture recording has become a routinely used service at many universities which have emphasized the automatic aspect of the recording and production process in the past and many of these institutions now provide recordings of several courses to their students on a dayto-day basis. For example, at our faculty, courses are routinely recorded, resulting in a huge collection of recorded lectures which can be accessed and replayed not only by our students but by anyone with an internet connection (see Albert-Ludwigs-University Freiburg 2006b). For a detailed overview about how lecture recording can be automated, including post-processing and delivery, we refer to Hürst et al. 2006. The end users’ view: Quality and usability. From the end users’, i.e. the students’ perspective, usability is a key issue, and system designers have to provide the required characteristics necessary to guarantee it. These features include, for example, the possibility to easily navigate and browse the files. Several studies (e.g. Lauer et al. 2004, and Zupancic and Horz 2002) confirmed that students normally do not review lecture recordings as a whole but use them very selectively, accessing only parts of particular interest and skipping areas that are less relevant for their individual information needs. Research responded to this request by, for example, developing interfaces that enable users to easily skim a file’s content, for example, by jumping from slide to slide (or sub-chapter to sub-chapter), scroll along the timeline at random speed, in any direction (Hürst and Müller 1999), and on different granularity levels (Hürst and Götz 2004), etc. Another important aspect is the quality of the recorded media. While this seems an obvious requirement, its actual realization is not easy or straightforward. In the beginning of this section, we already mentioned that the audio and presentation graphics are usually regarded as the most important media of a lecture recording. Therefore, they should be recorded in the most reasonable quality. Audio recordings generally only differ by bit rate and recording quality (which in turn is a result of, for example, the degree to which the used recording equipment is able to filter out side noise, etc.). However, significant differences exist for the presentation graphics depending on the method used for capturing. Normally, it is agreed that capturing the respective signal directly at the source, i.e. the computer or data projector presenting the information, is preferable compared to a pure video capture of the actual presentation. For such a capturing, there are
generally two approaches: vector- and raster-based recording. With vector-based recording, the presentation graphics are directly recorded at the source, i.e. normally directly by the presentation software tool used during the lecture. The result is an object-based representation of the presented material. Because the data is kept internally in a symbolic representation, the recordings generally have a low data rate and a high quality because vector-based formats can, for example, normally be scaled to different sizes without a significant loss in quality. Therefore, they are particular useful for scenarios with low bandwidth (due to the reduced data size) and if the resulting documents are replayed on a variety of devices (e.g. displays with different resolutions). One significant disadvantage of such an approach is its dependence on particular recording software. In contrast to this, a raster-based recording is usually done independently of any kind of presentation software but instead uses screen- or VGA-grabbing to capture everything that is visible on the lecturer’s screen during the presentation, including but not limited to the presented slides and annotations. The advantages and disadvantages of this approach generally complement the ones of vector-based recording approaches, i.e. the data volume is usually higher and quality might be reduced (esp. when downscaled to a lower resolution), but recording is not limited to a proprietary system and captures “anything that goes on” on the screen. Summarized, we can conclude that lecture recordings are a useful complement to other teaching materials, if the produced data has a high quality, interactive navigation is supported, and if it can be produced (i.e. recorded, post-processed, and distributed) automatically. In the following section, we will discuss if and how the scenario of lecture recording, delivery, and replay can be transferred to the podcasting scenario.
3. LECTURE CASTING Before we discuss the production and distribution of podcasts containing recorded lectures, we first describe different versions of podcasts and give examples for lecture podcasts with the respective characteristics. Podcast variations. Although podcasting is generally not limited to audio data, the term podcast is most often used to describe delivery of pure audio files, for example in the prominent MP3 audio format. This is due to the fact that as of today, the main portion of podcasts is still purely audio-based and, obviously, because of historical reasons since audio was the first and originally only media type delivered via podcasting. The majority of lectures delivered as pure audio recordings (i.e. just containing the voice of the lecturer but no visual feedback besides lecture title and the presenter’s name) comes from liberal arts, social sciences, and similar areas where classroom lectures are mainly oral presentations with limited usage of additional, visually presented material. Examples of such recordings can be found at Stanford (2006). So called enhanced podcasts are still based on audio data but include additional information. First of all, audio files can be segmented into single chapters, thus enabling users to quickly skip particular parts or to easily re-listen to specific segments. These segments can be labeled, for example, with a headline for the particular chapter. Such labels are displayed when users navigate through a podcasting episode thus enabling them to quickly identify the content of the respective segment. Finally, enhanced podcasts can be synchronized with additional artworks, i.e. still images which are synchronously replayed with the audio showing, for example, a different picture for each new segment. Although this technique was most likely introduced to represent information such as the respective album cover of a song that is currently replayed, it can also be used to show a still image of the presentation graphics stream (e.g. a slide) which is synchronized to the voice recording of a lecturer. Examples of enhanced podcasts containing lecture recordings that are segmented at slide transitions, labeled with the slides’ titles, and enhanced with still images of the respective slides, can be found at Albert-Ludwigs-University Freiburg (2006c). Screenshots are displayed in Figure 2 (left and middle). With the advent of mobile multimedia players (such as the new Apple iPod, Apple 2006) so called video podcasts or vodcasts gained increasing attraction. Vodcasts generally contain a video recording, i.e. an audio stream together with a synchronized video that is replayed on the small display of a mobile media player (or on regular desktop computers). Examples of video recordings of lectures which have been used to produce video podcasts can be found at Harvard (2006) and NC State (2006). A special case of vodcasting is
Figure 2. Enhanced podcast with jump labels for navigation (left) and synchronized thumbnail views of the slides (middle) and vodcast of the same lecture, i.e. video of the slides with related annotations (right). Notice the difference in size between the slide thumbnail shown in the enhanced podcast (middle) and the respective video podcast (right).
screencasting (Wikipedia 2006c), i.e. video podcasts that contain a video which was not recorded with a camera but produced via screen- or VGA-grabbing. Screencasting is often used for software demonstrations (e.g. vodcasts containing training courses for particular products) but, obviously, can be very useful for lecture recordings, i.e. the replay of the recorded presentation graphics stream. An example of such a screencast containing a recorded lecture can be found at Albert-Ludwigs-University Freiburg (2006d). A respective screenshot is illustrated in Figure 2 (right). It should be noted that the area of podcasting is a very vital field that builds on common internet standards but is not standardized itself. Hence, different names for different characteristics or versions of podcasts are frequently being introduced and often, people use the same name to describe different developments or use different names to describe the very same specification. For example, in relation to educational podcasts, names such as learncasting (Wikipedia 2006d), podagogy (Podagogy 2006), or profcasting (Profcast 2006) (for recorded lectures) are often used as well. However, the terms introduced above are commonly accepted in the community and therefore used throughout the rest of this paper. For podcasts containing recorded instructional lectures or presentations, we will subsequently use the notion of lecture podcasts. As we can see from this overview about different specifications and characteristics of podcasts, there are a variety of ways to distribute recorded lectures via podcasting to your students. Which one of them (if any) is the best and most useful is not clear yet. In fact, the answer which one will have the most positive impact on learning can only be answered in the long run based on actual usage data over several months or maybe even years and thus can (and will) only be given over time. However, based on our experience with lecture recordings and their usage and acceptance by the students, we are able to draw certain conclusions and make predictions that go beyond the level of pure speculation. Automatic production. In the previous section we summarized the most important requirements for the production and usage of lecture recordings in a traditional scenario, i.e. when replay is done on desktop or laptop PCs and Macs. We identified the automation of the recording and production process (including postprocessing and distribution) as one key issue from the producers’, i.e. the lecturers’ side. Because of the fact that the podcasted documents are usually encoded into some general media format, such as MPEG-1 Audio Layer 3 (.mp3) for audio or MPEG-4 for enhanced podcasts (.m4a) and videos (.m4v), it is generally possible to include the creation of podcasts into the normal, automatic production process as well. Implementing such a podcast export feature is by no means a trivial task because it involves significant advanced multimedia programming. However, roughly speaking: If we are able to automate the production and distribution process with lectures recorded for replay on desktops and laptops, we should be able to do it for lecture podcasts and replay on mobile devices as well. This argument is confirmed by the observation that several companies which offer tools for automatic lecture recording already started to extend their existing products with features for automatic podcast production (see Camtasia 2006, for example).
Delivered media types. In the previous section, we identified audio and the presentation graphics stream as the most important media types of a lecture recording. Video of the lecturer is sometimes added, for example, to create the feeling of social presence, but in most cases omitted in order to save bandwidth and storing space. This is particularly true in case of instructional lectures where the video often just contains the presenter’s talking head. With a mobile multimedia player, we can basically replay all of these streams, i.e. audio and video recordings. However, replay of video is usually restricted to a single stream due to the limited display size of mobile devices. Solutions such as “picture-in-picture” (e.g. a small version of the lecturer video in the left bottom corner of the presentation graphics video) are possible. However, due to the reduced relevance of the lecturer’s video most likely a pure replay of the presentation graphics should be a better approach. For devices which are not able to replay video files but offer a color display, enhanced podcasts with slides added as still images might be a reasonable alternative. Such a realization will not be able to reproduce the whole classroom experience if massive freehand writing is used and a lot of annotations are made on the slides. However, in a lot of situations, such an audio recording enhanced with timesynchronized still images of the slides might be a good alternative if no video playback is supported. The usefulness of a pure audio replay of a lecture recording clearly depends on the actual case, i.e. how much information is communicated via the audio stream vs. the presentation graphics. However, since audio is normally considered to be the most important information source of a lecture, we can think of a lot of scenarios where pure audio replay might make sense. For example, while the loss of the visual information might be critical for students who have not attended the original class, replaying just the audio signal might be sufficient if a user just wants to repeat a particular lecture when preparing for an exam. Production of podcasts and respective quality. Based on the targeted user community (or rather the devices used by them), one might consider producing lecture podcasts as pure audio, enhanced podcast, or vodcast, as described above. Due to the variety of devices available, it might actually be useful to offer all three different versions for download in order to enable users to subscribe to the versions that are most suited for their personal needs and best match their particular usage scenarios. While the production of audio is usually straight forward, creating the visual information for a video or enhanced podcast in a reasonable quality involves several issues. Generally, transforming video into different formats or down-sampling its resolution involves the risk of reducing its quality. Therefore, we would assume that approaches for vectorbased recording would yield to a higher quality for podcasted videos than a raster-based recording. However, we tested different versions, both vector- and raster-based, with various formats on an Apple video iPod and generally were surprised of the high quality of the visual output, even for data captured in a raster-based format via screen-grabbing. While some text on the slides became illegible due to a small font size, most information could be clearly recognized. As a rough rule of thumb, we can recommend font sizes larger than 24pt for reasonable readability on such a small display (resolution 320x240) whereas fonts smaller than 20pt generally become unreadable. However, the situation becomes critical in case of enhanced podcasts because in its current implementation, no full screen visual arts are supported but instead, just a square is represented which unfortunately does not take use of the whole display space being available (cf. Figure 2). This significantly reduces usability of the visual information due to a reduced readability of the respective slides. Navigation and browsing. Since users generally only replay selective parts of a lecture, it is important to support some sort of navigation or browsing in order to enable them to skip irrelevant parts and quickly identify parts of particular interest (cf. previous section). However, most of the examples for lecture podcasts given before do not support any kind of navigation except for the standard features which are provided by the media player and which are normally not sufficient for this scenario. Therefore, we experimented with different versions based on the possibility of enhanced podcasts, i.e. by providing markers or jump labels for easy navigation through different segments of an audio file. In our final implementation, we use the slide transitions to segment the recordings. Hence, users are able to navigate through a lecture recording based on the used slides by a simple button click. Each respective segment is labeled with the slide title thus giving further information to the user. It should be noted that all this information can be extracted from the data automatically, thus fulfilling the first requirement specified above, i.e. the need for automation of the production process. In addition, we display a snapshot of each slide synchronized with the respective audio segment from the lecture. Although a comparable segmentation with jump labels can be created for a video as well, it is unfortunately not supported by the current replay mechanism of the Apple iPod. We hope and
expect this feature to be included into future updates of the system software, and we assume it to be a very useful one with a significant influence on usability.
4. SUMMARY AND DISCUSSION In the two preceding sections, we described the general automatic lecture recording and distribution scenario, introduced different variations of podcasting, and finally discussed how lecture podcasts can be produced given certain restrictions and requirements. It turned out that not all of these requirements (e.g. a comfortable and easy navigation) can be supported due to limitations of the current technologies (e.g. missing support for jump labels in videos) but also that to some degree, we can exploit the existing features to improve the situation, that is to produce podcasts which go beyond pure audio or video recordings but instead are optimized to the lecture recording and replay scenario (e.g. by using enhanced podcasts to include jump labels for easy slide-based navigation). To evaluate the actual usability of these different variations of podcasts, we set up a web blog with podcasts for two courses currently offered at our university. These are routinely recorded using common lecture recording software and offered for replay and download to the students in the usual way through our e-learning portal (see Albert-Ludwigs-University Freiburg 2006b). In addition, the students are now provided with three different versions of podcasts, i.e. pure audio, enhanced podcasts with markers and slide snapshots, and screen-captured videos with synchronized audio (see AlbertLudwigs-University Freiburg 2006c). Based on this setup, we will perform a detailed evaluation with the students who are actually using these podcasts (or who will not use them for whatever reasons) in order to identify useful setups, problems with the current technology, reasons why they do not use them, etc. As said before, the answer to the ultimate question if lecture podcasts are useful at all, i.e. have a positive impact on learning can only be answered in the long run. With the planned evaluation, our aim is to get a step closer to this answer. In addition, we expect to identify useful usage scenarios for lecture podcast in order to disprove critical voices questioning their relevance and usefulness (or verify their arguments if it turns out that lecture podcasts are indeed useless). Although we see some truth in those criticisms, there are a lot of examples where lecture podcasts seem to make sense. Think of a student who just wants to refresh a lecture before meeting with two of his/her friends to do their homework together but does not have the time to go through the whole lecture. With a mobile media player, quickly reviewing parts of it can easily be done during the ride on the bus to the university. Think of the three students sitting together doing their homework, one of which remembering a quote the teacher made during the lecture and which might be of relevance for a particular exercise. Using the lecture podcast, one of them can quickly look it up and re-listen to the respective part of the lecture1. In addition, we agree with the often heard statement that no one wants to listen to a podcast containing a long, boring lecture given in a monotone voice. However, assume that you really have to listen to such a lecture, for example, because it is the only way to get the respective information; where would you prefer to do so, in the classroom or sitting on a park bench with a related text book on your lap and headphones plugged into your ears? After all, it is not the podcast which makes this lecture boring; it is the lecture itself that seems rather unattractive and not very supportive for an interesting and absorbing education. So even podcasts or audio recordings which we all agree are not very engaging might have a certain value2. Again, which scenario will prove to be really useful and therefore, will be adopted by the students can only be figured out in the long run. The work we did so far and presented in this paper brought us one step closer to the answer to this question and only time can tell, if some of the intuitive arguments give for or against lecture podcasting will hold true. 1
However, “quickly look it up” is only possible if reasonable navigation and browsing support is offered by the system and it is not clear yet, if the features offered by the current technologies are sufficient to support such a task, as we described before. 2 As an interesting side note, one of the lecturers of the two courses we are currently offering as podcasts is using them to re-listen to parts of his lectures in order to verify and improve his presentation style. While such a self reflection could have also been done with traditional lecture recordings, he never found the time to do this before. Now, with a mobile media player, he can easily listen to his lectures while taking the bus to work and, hopefully, at the same time improve his teaching skills.
REFERENCES Abowd, G.D. (1999) Classroom 2000: An experiment with the instrumentation of a living educational environment. IBM Systems Journal, Special issue on pervasive computing, Volume 38, Number 4, October 1999. Albert-Ludwigs-University Freiburg (2006a) List of Lecture Podcasts, see http://ad.informatik.uni-freiburg.de/blogs/podcasts/?page_id=9 Albert-Ludwigs-University Freiburg (2006b) Lecture Recordings of the Faculty of Applied Sciences, see http://electures.informatik.uni-freiburg.de:8484/catalog/courses.do Albert-Ludwigs-University Freiburg (2006c) eLectures Podcasts, see http://ad.informatik.uni-freiburg.de/blogs/podcasts/ Albert-Ludwigs-University Freiburg (2006d) Manoli, Y.: Entwurf analoger CMOS-Schaltungen (lecture in German), see http://electures.informatik.uni-freiburg.de:8484/catalog/course.do?courseId=CmosSS2005# Apple (2006) Apple iPod, see http://www.apple.com/ipod/ Camtasia (2006) Camtasia Studio and Screencasting, see http://www.techsmith.com/camtasia/screencast.asp Campbell, G. (2005) There’s something in the air: Podcasting in education, EDUCAUSE REVIEW, Nov/Dec 2005, Volume 40, Number 6. Available at http://www.educause.edu/ir/library/pdf/erm0561.pdf EPN (2006) The Education Podcast Network, see http://epnweb.org Gemmell, J., Bell, G. (1997) Noncollaborative telepresentations come of age. Communications of the ACM, Volume 40, Number 4, April 1997. Harvard (2006) Malan, D.: Computer Science E-1: Understanding Computers and the Internet, see http://www.fas.harvard.edu/~cscie1/?page=podcast&type=static Hürst, W., Müller, R. (1999) A Synchronization Model for Recorded Presentations and its Relevance for Information Retieval. Proceedings of ACM Multimedia 99, Oct./Nov. 1999 Hürst, W., Götz, G.(2004) Interface Issues for Interactive Navigation and Browsing of Recorded Lectures and Presentations. Proceedings of ED-MEDIA 2004, AACE, Lugano, Switzerland, June 2004. Hürst, W., Mueller, R., Ottmann, T. (2006) The AOF method for automatic production of educational content and RIA generation. Int. J. Cont. Engineering Education and Livelong Learning, Vol. 16, Nos. 3/4, pp. 215-237. Lauer, T., Müller, R., Trahasch, S. (2004) Learning with lecture recordings: Key issues for end-users. Proceedings of ICALT 2004, Joensuu, Finland, August 2004. Meng, P. (2005) Podcasting and Vodcasting – A white paper, University of Missouri, March 2005. Available at http://edmarketing.apple.com/adcinstitute/wp-content/Missouri_Podcasting_White_Paper.pdf Müller, R., Ottmann, T. (2003) Content production and the essence of simplicity. Proceedings of the 4th International Conference on Information Technology Based Higher Education and Training (ITHET 03), Marrakech, Marocco, July 2003, IEEE Education Society. NC State (2006) Hindi 102 @ NC State University Video Podcast Lectures, see http://fltc.chass.ncsu.edu:16080/Hindi/ Ottmann, T., Müller, R. (2000) The “Authoring on the Fly”-System for Automated Recording and Replay of (Tele)presentations. ACM/Springer Multimedia Systems Journal, Special Issue on “Multimedia Authoring and Presentation Techniques”, Vol. 8, No. 3. Podagogy (2006) Podagogy – Where Podcasting meets Teaching and Learning, see http://www.podagogy.com/ Productive Strategies (2006) List of Academic Lecture Podcasts, see http://productivity.strategy-blogs.com/2005/10/list_of_academi.html Profcast (2006) Profcast: from the University of Mary Washington, see http://www.profcast.org/ Reynolds, R. (2005) Right ways and wrong ways of podcasting in education. XplanaZine, August 2005. Appeared online at http://www.xplanazine.com/archives/2005/08/right_ways_and_1.php Stanford (2006) Stanford on iTunes, see http://itunes.stanford.edu/ Wikipedia (2006a) The History of Podcasting, see http://en.wikipedia.org/wiki/History_of_Podcasting Wikipedia (2006b) Podcasting, see http://en.wikipedia.org/wiki/Podcasting Wikipedia (2006c) Screencasting, see http://en.wikipedia.org/wiki/Screencasting Wikipedia (2006d) Learncasting, see http://en.wikipedia.org/wiki/Learncasting Zupancic, B., Horz, H. (2002) Lecture recording and its use in a traditional university course. Proceedings of the 7th Annual Conference on Innovation and Technology in Computer Science Education, pp. 24-28, ACM Press.