Today, I presented the ‘Glasshouse Birdman’ prototype in a MCD studio postgraduate session. Many people present where not aware of my research topic. I decided not to provide any context to see what people thought I was researching. Initially there seemed to be some frustration with how the prototype was presented, especially if as I described the plan was to present VD for use in real-life scenarios. People suggested that the prototype looked to polished and finished rather than being a work-in-progress. A version that looked more like a sketch (drawn in crayon) with other supporting visual diagrams was the suggested alternative. Something that showed the process and behind the skin structure.
With the floor open for what this specific practice was researching there was some discussion about exploring non-linear narrative with a point made that this was not necessarily a new area of research. Therefore, where was the research going to be taken? A question that instigated some ideas on what the practice was exploring in terms of new territories. Classification, Tagging, Folksonomy emerged as one tangent - social media. The semantic web another. The idea of looking at how narrative is being compressed led to thinking about new types of audiovisual literacies. I mentioned that one thing I found myself doing is comparing production processes. These are the differences between how I would have recorded and post-produced a work like this for a linear montage edit. There was a suggestion to look at how this content would be structured in a linear version which I have been planning.
I pointed out the difficulty of working with so many new fields of study like interaction design, design research, software development and how none of these fields are where I come from in terms of prior knowledge. One idea was to find a specific theoretical model to bring all these together as way of critiquing and theorising the research. Experience Design was one suggestion. The session was useful particulary in terms of thinking about a revision of the research focus and questions. The original research topic is very broad but has influenced the resulting practice as a form of critique on online video practice. But, the topic could be revised along with the research questions in terms of shifting the focus at this stage to a focus on a new emerging territory.
Reference: Bill Buxton, Sketching User Experiences,: Getting the design right and the right design, San Francisco, Morgan and Kaufmann, 2007.
The Glasshouse Birdman prototype is coming together as a work-in-progress. It has been a useful experience to learn how this type of production is recorded, authored and structured.
Montage editing: In comparsion to the usual broadcast production there seems to be little need to use montage editing. In most cases the granular nature of videodefunct works with scenes as short duration single continuous shots. Even dialogue with the style of approach suits being uncut. I was surprised that this would be the case in a documentary style prototype. It is amazing how much conversation and observation is presented in small self-contained fragments. In fact montage editing seems strange in this environment not only from a production perspective but also as clips. The multi-channel composition asks for scenes to be realised as shots played back concurrently rather than edited into montage sequences. The breaking up of time in a montage sequence seems very odd after working with single continuous shots. Although, I think sharp short montage sequences could be used to enhance single continuous shots in the triptych. Also, longer edited montage sequences, I found due to edits and change in shots information ended up being big file sizes with compression. I
Recording notes: I did find myself shooting with the zoom and handheld camera moves I would not normally use to get coverage as part of aiming for one continuous shot. Working solo I experimented with a radio mic because I knew this work would rely on Terry telling stories as he did things. Also, I had to dress in Terry’s clothing as the birds freaked out when I was around on the first day, so a audio person would have made things even harder. In retrospect, the radio mic sound brought up ideas about online video audio where in comparison to television and cinema the subtleties and complexities within reason are not so important or even lost through most computer speakers. The radio microphone recorded the audio close to the Terry including his dialogue and sound associated with his actions. These sounds in miniature video frames are what is needed to make things clear for the viewer, the sounds that help communicate what is happening on screen. But, I did have some technical problems with the radio mic on movement which could be refined. Also, more time could be put into mixing in atmospheric underlay to make things richer and make up for all the audio being recorded close and narrow through the radio mic. Post-production processes do need to be simple though when you are working with so many singular clips.
In the hypertext environment where the clips are highly granular, self-contained and inter-linked, a lot of the content can be used, with all sorts of supporting narrative threads. I even considered a behind the scenes category for example. There is real opportunity to be self-reflexive and work very informally compared to the restrictions of a single montage edit in terms of camera coverage, the narrative structure and what has to be discarded. In this example it is really surprising the ratio of content that is usable. A highly sustainable way of recording.
Categories and Tags: I am starting to make narrative structures work using categories and tags in hierarchical levels. The user enters categories then brings up tags. The narrative can be contained within a category by keeping the tag only within that category with no cross-references which leads to a type of dead-end for the user. Introducing some tags that cross-over into other categories utlises the hypertextual nature of the interface and makes the multi-channel viewing more dynamic. The trick seems to be to keep tags to a minimum and make the cross-over choices with some consideration. I did find myself re-thinking categories as I got to know the material better.
Interface design: There could be some changes to the way the interface is modified for this content. For example I like the three audio tracks running with the louder being moused over. But with the single person focus there is no real need to hear three lots of dialogue across the triptych - the volume could be lowered on the clips not being watched. More later on other revisions with the categories and tags.
Chris let me know recently about a online documentary that he made titled ‘Thumb Candy’ on SMS text culture in the Philipphines that he put together within a blog. He gives Videodefunct a plug on the More about the project page as being an influence on using tagging and a blog to classify the video content.
1. Log material and get a sense of the content
2. Loosely think about categories by making connections across the material (try sketching a map)
3. Start with individual stand alone clips as a way of determining what might work as montage sequences. (export into temporary category folders)
4. Work out which clips should be edited into montage sequences to help develop specific narratives within the overall content. (think about maintaining the shortest durations possible)
5. Review all the exported clips and emerging categories.
6. Think about the titles and tag - keywords. (Look for associations across tags with the aim of keeping the amount of tags applied to each clip at a minimum)
7. Think about how categories can be used like chapters or sections for developing particular narrative tangents. (These are sections that offer less choice across the content with the aim to steer the user down particular narrative paths.
8. Review the tags by testing the player. (Look at user testing if needed)
I put this summary together for a in-house post-grad feedback session next week. The portrait angle helps with clarifying my objectives with the prototype.
Birdman Portrait: I am in the process of making a Videodefunct (VD) prototype that explores the structuring of an online video portrait. I am interested in making a prototype that reflects more of a traditional documentary approach compared to the earlier abstract VD prototype ‘Pedestrian’. The content is personal and intimate with the objective to produce an example that is accessible to a broader audience. The portrait is of a wildlife carer named Terry Dale who specialises in looking after native birds in Queensland. We recorded together his daily routine of caring for native birds, which included in a very informal manner the documentation of numerous stories of his experiences. What I am looking for in terms of feedback on a micro-level is ideas on how the material should be archived, with the objective of using this example to clarify the broader objectives of my doctorate research. The structuring of this content is affected by decisions that incorporate not only narrative but also classfication and interaction design.
I travelled up to Queensland recently to record some video material for a VD prototype. I was interested in making a prototype that reflected a more traditional documentary approach compared to the earlier abstract nature of pedestrian. A work-in-progress prototype that is accessible to a broader audience. Content that demonstrated an intimate type of human story. My questions to myself – How would VD as a player work structurally with content that viewers can connect with on a personal level? How would that content need to be prepared in comparison to the pedestrian experiment?
So, I was on the lookout for a personal profile that reflected a personality and a passion. I had the wonderful opportunity to work with an Uncle of mine. Terry is a wildlife carer who mainly looks after injured birds. It is an interest that he has had since he was very young. Over a period of few days we recorded some material together and it was a new experience to work with someone I knew really well in comparison to winning the trust of a stranger for television broadcast.
With the VD application in mind I was thinking constantly about self-contained fragments, what I could capture in most cases in one scene, within a short duration of time. An approach that has similarities with recording material for factual television, but in this case I was not really thinking about coverage for linear editing. The coverage was about providing an overview of the story that gives viewers an insight into the character and his particular interest through an observational process. Now, I am logging the material and working out whether some scenes may be edited together into sequences.
PAD.MA - short for Public Access Digital Media Archive - is an online archive of densely text-annotated video material, primarily footage and not finished films. The entire collection is searchable and viewable online, and is free to download for non- commercial use.
We see PAD.MA as a way of opening up a set of images, intentions and effects present in video footage, resources that conventions of video- making, editing and spectatorship have tended to suppress, or leave behind. This expanded treatment then points to other, political potentials for such material, and leads us into lesser-known territory for video itself… beyond the finite documentary film or the online video clip.
The design of the archive makes possible various types of “viewing”, and contextualisation: from an overview of themes and timelines to much closer readings of transcribed dialogue and geographical locations, to layers of “writing” on top of the image material. Descriptions, keywords and other annotations have been placed on timelines by both archive contributors and users. At the moment, PAD.MA has approximately 160 “events” on video, mostly from Mumbai and Bangalore. This adds up to about 100 hours of fully transcribed video footage, which we expect to grow to more than 400 hours by early 2009.
Notes: Pad.ma uses a lot of text-annotating to provide extra information on video content across the timeline.
Q 4.1: What is annotation?
A 4.1: Annotation is the adding of textual information, in this case to parts or the whole of a video event. This is similar to the general concepts: comments, commentary, or marginalia.
Q 4.2: Who has put in the current annotations?
A 5.2: The first layer of annotations have been put in by the original contributors of the video event.
They also have a map function as an overview which utlises google map functionality. The videos are kept at full length and have a scrolling feature for accessing the timeline at any point. Once a point in the timeline is accessed an overview of where you are on the timeline is provided along with a lot of annotated text, including keywords. In a way the features normally hidden in video editing software are revealed in the interface. The application seems to rely on Python and JavaScript programming with the source code available https://wiki.pad.ma/wiki/Source. They are not bothering with supporting IE and there are plans to support Ogg theora.
Q 2.1: Which browsers do you support, on which platforms?
A 2.1: We currently support Firefox and Safari, on Linux, MacOS and Windows. We do not support Internet Explorer. However, if you wish to endeavour to make the site work on IE, please appeal to IE to support web standards in their next version.
Q 2.2: Do I need to install anything to view the videos on pad.ma?
A 2.2: Yes, currently you need to install either the VLC plugin (ensure you tick the “Install Mozilla Plugin” box while installing) for Firefox/Safari on Windows or OSX, or the OggPlay plugin for Firefox on OSX or Linux machines, to play the videos. In the near future, Firefox plans to support Ogg Theora (the open source video codec we are using) natively, and you should not need to download anything to view videos on pad.ma.
Notes on their position in terms of software development:
PAD.MA is not intended as a software product, but you are free to use the code to create your own instance, if you like. Obviously it would be more interesting if these instances fed into each other, if people’s annotations could layer and combine rather than exist in artificially separated environments.
An example of using a mouse over tool that shows metadata attached to the video clip. This is similar to the snap shots plug-in. A tool that increases the text element of adding information to audiovisual data before it is viewed. It also works towards breaking and keeping blogging chronology in way as the date/time of posting is included.
The SIAB people have released a number of wordpress blog themes customised for publishing video:
The Video Producer theme series will feature WordPress themes that are designed from the ground up to break the “blog diary” format and allow video creators to display their work differently from what is usually seen on most blogs, while retaining the flexibility and functionality that the WordPress core already provides.
After some recent writing on videodefunct about what it is, I recognised even more the work VD is doing as part of a growing movement towards customised video themes for blogs. I see in the distant future a plethora of choices as this development becomes more and more accessible.
Geert talking to Tom Sherman about his Video Vortex presentation including some views on his notion of vernacular video. This YouTube clip was made by the Masters of Media group.
I use this blog for PhD research on Internet video practice. I also have a teaching blog that I use in the Media Department at RMIT University.
