Conference here at RMIT in November. All about filmmaking as a form of academic research. Details available: http://www.rmit.edu.au/mediacommunication/sightlines
I’m running a Korsakow workshop as part of the Australian Screen Production and Education Research Association annual conference. It’s in Newcastle this year. If you’re at all interested in seeing the software, and you’re attending, please come along. There’s a bunch of us also presenting papers at an interactive documentary panel during the conference too.
These are notes from my teaching in 2011, and are the various strategies and things I’ve learned through using Korsakow that make using the program simpler, or at least easier. They are a mix of tech and conceptual things. Am dusting them off for the new subject and realised worth reprising here.
- Dense nodes. One way to make connections between clouds in your k-film is to use dense (dense in the sense of thickly linked with keywords) nodes that work like ‘hubs’ that join or connect different clouds together. I’ve written about this in more detail nearby.
- Export as you go (the quick way). Korsakow now lets you do a low res export, or the usual export to web. A tip. Ignore low res export because if you choose to do this it exports everything, every time. You are much better off setting your compressing settings to the lowest quality, doing an export for the web, and then as you add more media do the web export. That way you get the dialog box that asks you if you want to redo everything or any the new additions. Choose new stuff and things export much much faster. This is also how you can easily test interface changes, and make minor changes to keywords and so on. These changes don’t require video to be added or recompressed, so are exported very very quickly and you can then look at the work to see what differences your changes have made.
- NAME YOUR PROJECT. Yes, that is yelling. Why have no title for your project? Why publish it as ‘untitled’. It is your work, give it a name. Also the title of a work can do.
- Transcoding. Transcoding is what Korsakow does when you export a project. It takes the media you have added to the project and recodes (recompresses and changes the file format) to FLV using what you have selected under File – Project Settings – Export. This takes ages, literally if you have a lot of video it can take hours. So be prepared for this.
- Export a project but nothing appears. If you load up a project with video, but have not added any keywords in or out (‘snuified’) the clips, then when you export your project to the web and go to play it there will be nothing visible. Your video will be transcoded, but nothing appears because it needs clips with rules (keywords) attached to build anything.
- Compression settings for export (File – Project Settings – Export). H.264 medium currently provides the best compression options (file size versus quality). If you want to be low rez (compression artefacts, etc) then use the low resolution H.264 project settings.
- Thumbnail videos. Don’t use the video you add to provide thumbnails. They are large files and it is very (very) inefficient to use these as thumbnails – people will be downloading something the size of an elephant when it should be a mouse. This matters because we all pay for bandwidth, so please don’t expect people to have to pay for elephants to buy a mouse. If you want video thumbnails then compress your video a second time to the size of your thumbnails.
- Start simple. Complex structure is created by the repeated application of simple rules. Use this when thinking about keywords. Don’t use 20, use 4, build, export, test. Then think about what changes you might need.
- Korsakow is only for authoring. This means you make content – video, still images, audio, outside of Korsakow using whatever software takes your fancy. What you bring into Korsakow is your media already to go. You don’t make video edits in Korsakow (well, in my terms you make edits possible via the keyword link structure, but that’s a different kettle of fish entirely to editing a movie in the usual sense of the word). So all the media your bring into your project should already be the right size for your project, the right compression, and in the format and data rate that you want. You NEVER bring in something bigger and then make it smaller within Korsakow (for instance a video that is 1280 x 720 that you then present as 640 x 360). This slows things down since heaps more is being downloaded to be played than necessary, and is, well, the digital equivalent of picking your nose at the dinner table (seriously, that is how uncool it is).
- Use this as part of an iterative process. This means do not think you can plan your Korsakow project on paper, or somewhere else, and then you will just import your media and link according to your plan. You can try to work like this, but that is a bit like thinking you can write music by writing each instrument and not listening to them all together, and not actually playing music while you’re writing music. Or it is like writing something in Word, then exporting it as HTML, and then telling yourself that you are writing hypertext. No. You need to work within the medium to be working in that medium. And you need to work in that medium to learn how to work in that medium. You don’t learn how to play a piano by reading about it, watching others, watching a DVD, or writing out instructions on paper. You have to play. So it is with Korsakow. But there is more to it than just having to use it to learn it. There is a logic to the tool which you can only use if you work in the tool. So the way to make a project is to add some media, add keywords, export, and play the work. How’s it going? What happens? Does it do things you didn’t expect? Good things or bad things? This is really much easier to do when the project is small, when you can try things out easily, and then as you develop an understanding of the shape,patterns, and behaviour of your project you can add more media, and keywords, and then export and evaluate the work. This is what I mean by an iterative process. You add some stuff, you evaluate, you make changes. It is built up through small steps, small increments. This might be unusual for mamy of you. In TV you are taught to script, storyboard, and then use that as your plan. In essay writing you prepare an essay plan and use that to build from. Here it is much much more organic. Personally I’d describe it as more like writing a song. Add a line, sing it, fiddle with it a bit more, back to the music. You sort of have an idea of what it is about, or where it comes from, but there is no strict plan or map.
- Do low res exports. Part of this iterative process is to add, keyword, add SNU ratings and so on, and then export to test and see what is going on. (Again, think of this exporting and viewing as what you would do in radio or TV, you do an edit, then you look at the shots, or listen to your piece, with your new edit. You don’t just keep editing to the end then check that the edits are OK, same process for making a k-film.) So, exporting can be slow since the Korsakow engine will take your video and audio, and then transcode and recompress it. This is slow work. So for work in progress when you export go into File – Project Settings – Export and choose FLV Low. This will export much more quickly, and yes it will be low quality but this is drafting. The final export or for publication of work in progress you will change the Export settings to a higher quality and re-export. But remember, if you have a lot of material this will take a lot of time. (And this is also why you must always have your original media, if you have moved it, deleted it, renamed it, then there is no way for Korsakow to export and render a better quality version because it can’t find the source media.)
- Complexity comes from the iteration of a simple structure (or rule). You build complex structures (patterns) in a k-film not by using a lot of keywords but by defining and applying a simple set of rules, consistently.
- What happens if the source videos are different dimensions? A mess. In a word, well two. The software is designed to support video that, more or less, is at the same dimensions. Being smaller might not be a problem, it might just be presented larger, or you can put a background behind it to make it fill the frame. I suspect (but haven’t tested) that the default behaviour will be to fill the video screen you have defined. So if it is a different aspect ratio (which is a much bigger problem than a different resolution) then it will be stretched, or squashed to fix. Work arounds would probably be to use Compressor to crop the video to get it to the right aspect ratio and/or dimensions.
- Read the error messages. Yes, Korsakow is low budget software so you can’t export the sort of design, support and so on you get for either very large open source projects (WordPress for example) or commercial software. However, pay attention to error messages you get, while cryptic they usually point to what is going on. For example if you get an error message about a file type being not supported or corrupt it will tell you which file. Remove the file from your project. If the export now completes there is something wrong with the file. There are a lot of examples like, this, so read the error as it does give good clues as to what might be going on.
- Thumbnails can be video or jpegs. Video thumbnails are very cool, because when you mouse into them they play. But this seriously increases the bandwidth demands of your project. If you are going to use video thumbnails then it is essential that they are recompressed to a suitable size and data rate. This means they get recompressed to the size of the thumbnail. You NEVER compress video to, say, 320 x 240 if it is going to be displayed at 160 x 120. Why? Because the larger video requires four times as data as the smaller. It is about bandwidth which becomes about how much data has to arrive at your viewer’s computer before they can do anything. The more you can minimise this (the less data that has to arrive) the better off you are.
Hannah Brasier has started her PhD, and I’m very happily one of her supervisors. Late last year she successfully completed her PhD confirmation, and she’s put her slides and some excellent notes on her blog. Towards Affective Knowing in the Diary Film and Interactive Documentary.
This is a small iBook project that came out of the nonfictionLab symposium held in December. Been working on it over Christmas, amongst cleaning the pool, presents, family and visitors.
Rezine 01: Research Notes Toward Critical Nonfiction Practice (iBook, 157MB).
From the introduction:
The first nonfictionLab symposium kicked off at RMIT University in December 2013 bringing together a sampler of the scholarly work being undertaken by the lab. Or, as we pimped ourselves:
From the essay, film-making, poetry, documentary, vernacular media, digital archives, memoir and design, nonfiction is increasingly a site of creative, theoretical and analytical interest. With keynote speakers, Ross Gibson and Jeff Sparrow, this inaugural nonfictionLab Symposium 2013 seeks to place some markers, critical and adventuresome, across the interdisciplinary domain of nonfiction studies.
Panels sessions include: Guessing games: Interpreting surfaces, subverting perceptions, Experiments with experience: Negotiating memory, observation and imagination, (Dis)placements: Locating perspectives: spectral sites and designs, Patternings: Generating rhythms, rituals and the accidental.
This rezine is the first transitory, possibly ephemeral, quick and dirty research sketch, or field notes, of our work. The intent is to show things in progress, a snap shot collage list of small bibs and bobs that are all on their way to becoming something else. A chap book come digital pamphlet that is an opportunity to begin to describe and argue for the sorts of theoretical and critically engaged creative nonfiction we do. Let’s open the black box of research and scholarship and rethink scale, practice, documentation, the rational and poetic.
This is the Bunnings of research, welcome to rezine 01.
Rezine 01: Research Notes Toward Critical Nonfiction Practice (iBook, 157MB).
Interesting…but fiction is also insecure. I really like this essay by Vivian Sobchack (1999) ‘Toward a Phenomenology of Non-Fictional Film Experience’ because she suggests that there aren’t fixed boundaries between fiction and documentary – that it’s about spectatorship- and depends on the viewer’s experience of a film , how we might view, feel, interpret changing moments in any given film. Thus her famous quote: “One viewer’s fiction may be an other’s film-souvenir; one viewer’s documentary, another’s fiction”.
Nevertheless, your post has got me thinking – last night I watched “The Outlaw Michael Howe” on ABC TV; it is a tele-movie, a historical period drama about a convict in Tasmania. I have a range of issues with this film which I am grappling with how to address…then today I read it was (part?) funded under Screen Australia’s National Documentary Program’s Making History Initiative. Well, it may be based on a true story but this does not make it a documentary! and why it received government funding as a documentary concerns me. So obviously I do think there are significant differences in the fiction-non fiction modes of address.
Sobchak’s essay is in Collecting Visible Evidence (1999), ed. Michael Renov and Jane Gaines, Minneapolis: University of Minnesota Press, 1999, 241-254.
The insecurity I have begun to think about (in two different ways, one is about centres of indetermination, the other the one that Jeni’s responding to) is a way to think about the sorts of willingness of documentary, historically, to play with form and technology. While fiction has done this, documentary seems to be a richer site of technical experimentation, as well as more complicated modes of address. Fiction, in film, might play with story and plot, it might ‘break’ the fourth wall (Godard of course springs readily to mind), but not that many play with modes of address that, say, Marker did way back in Letter from Siberia (direct address, irony, animation, literal repetition of footage).
Translating this to online, and I think the evidence is showing the documentary is doing much richer things in regards to networked practice than fiction. There is more variety of approaches and work, certainly more experimentation in relation to content, style, form, platform, and so on, than I think has happened in fiction film making. So I was wondering why. Why would nonfiction be more willing than fiction, which after all is celebrated as the place of ‘creative’ practice, be more conservative in relation to these things.
So my tentative answer comes out of possible worlds theory and those elegant definitions that I Iike (for their pragmatic exactness) where fiction are works about a world, while nonfiction are works about the world. In the former the world has to be internally consistent, and true. So it is true that Darth Vader is Luke Skywalker’s father, and that Jedi Knights have light sabres. Just as it is true that Ethan Edwards is a returned soldier from the American Civil War who undertakes a quest to, perhaps, kill his kidnapped niece. These are truth claims, and verifiable as truth claims, but they are truth claims that can be verified because they come from fictional universes with internally coherent rules. Nonfiction, on the other hand, makes claims about our world. They can be contested, but the evidence doesn’t come from the fictional world, it comes from outside the text, from the world.
Hence fiction is very secure in itself. Once I set the rules I can do what I like. Unreliable narrator, in a universe where people sprout limbs as required, photosynthesise, and reproduce like fungi. Where if you fall in love you die. It really isn’t a problem. So narrative is sovereign, in that solar, regal, absolute way that the idea of sovereign demands. It lets you do whatever you wish, just keep it internally consistent. Nonfiction on the other hand can’t do this. The world is always there, bearing witness. I can make all sorts of truth claims, sure, but here narrative is not fiction, narrative is telling and claiming about a world that is external to itself. Here, in spite of how arrogant any nonfiction work wants to be in its claims for certainty and concrete absolute factness, it’s test is not internal coherence but the outside. Which is unbounded. This is the reverse of fiction, for fiction always has clear edges – there is no scientific breakthrough that will suddenly render the universe of Star Wars wrong – and so nonfiction can’t invest in narrative as sovereign. The world is sovereign here and so narrative becomes unsure of itself. AKA, insecure. As the world is sovereign, and outside, and unbounded, in relation to any nonfiction work, then what I say, and how I say it, can never have the security of fiction, and this insecurity opens up the form to, well, wonderment and experimentation.
So, nonfiction is much more willing to break things, play with things, question. It finds itself having to, because it can never pretend to say enough to create the sort of hermetic universe that is fiction’s right.
Was a beautiful day Saturday. Mid 20s, ms 7 y.o. having a sleep over. Ms A. taken over the dining room with all the furniture moved out and a quilt laid out, mix of new modernism meets Amish. Very impressive piece of work. Real coffee. Spent some time working on the sonnet Korsakow film. It’s tricky, this one.
A sonnet is all about structure. That’s part of the point. It’s a modern sonnet, which means it’s free verse, but still following the 3 stanza’s of four lines and a closing stanza of two lines structure. I have written the lines, well, more or less as they get edited, fiddled and worried with pretty regularly. I have the video. The video’s are very very simple, and repeated – that’s just part of the poetry really. But the structure, a sonnet needs to be fourteen lines, not more, not less.
A Korsakow film is all about structure. That’s part of the point. It’s shape is, essentially, musical (or poetic) and it is music that gives us the richest and simplest vocabulary to deal with repetition, rhythm, chorus and its close friend, the hook. You return, leave, come back again. Repetition, in different guises, is a fundamental architectonic principle here.
And therein lies the trickiness. I could make a simple HTML based work that you progressed through, a line or even a stanza at a time, to its close. However, in using Korsakow I’m interested in something a bit different, where each stanza’s line could be read in any order, and so letting the lines of each stanza change in their order each time you view the work. Each stanza is marked by a change in interface, but the problem of how to shape this film so that it works as a sonnet is problematic.
One option is to fake it, and provide a fixed path through each line and video. The last line of the first stanza taking you to the first line of the second stanza, and so on. That is hardly what I’m after, and I can hand code in HTML so don’t really need the generative engine that Korsakow provides. So the other option is to let the lines and videos of the first stanza be connected, in any order. This is pretty easy. I could give each clip one life, let it join to the other stanza’s, and each time you view it the order would vary simply because each line has the same chance of being connected to another as any other. The problem though is that in Korsakow I can’t write Boolean conditions, so I can’t provide a rule that would in effect say “if all clips viewed provide a link to stanza 2″. So I can provide a link from one clip to stanza two, as a bridge across, but if I wanted to constrain it to only appear as the fourth line, then my concept of letting each line appear anywhere is broken. If I let it appear at any time then you could arrive there after only one or two lines, and then find yourself in stanza two, and then the four line structure of the sonnet disappears.
Similarly, I could let clips be able to appear more than once (what Korsakow describes as ‘lives’), but since it doesn’t count in accessible ways I also can’t write a rule that would say “once any four of these clips viewed, move to stanza two”. Now, this would be a better rule, as the rhythms I like and value in these works would be more apparent. With this rule you might read the same line in stanza one twice, and never see one of the other lines until you read the poem a second or third time. I’m very good with that. That is what should happen in these sorts of self organising systems, you read and return and in these changes between readings you learn the shape of the work (and therefore what it’s about). But I can’t write such a rule.
So at the moment building it is feeling like a bit of a clunky hack. Any of the first four lines are set as start films, so we can begin anywhere. Thinking it through the solution is to have the four lines of a stanza all linked equally to each other (they all share the same in and out keywords in Korsakow). They have only a single life and the interface only allows for one thumbnail to be shown. This means a clip loads in the current stanza, you only get one choice of where to move to next from this clip, but this choice is only constrained by being any of the remaining clips for the current stanza (remaining as with clips only having a single life the current clip cannot be returned as a possible connection). Then every clip in the first stanza also contains a second keyword which links to the second stanza. This second keyword is listed on the second line of the out keyword, and so if the first keyword doesn’t match then the second is used. In this way the film is able to cycle through the four lines of the stanza, in any which order, and once four have been viewed it then links to anywhere (as all four clips in the second stanza contain the same keyword that all four clips in the first stanza are pointing out towards) in the second stanza.
(Brief note, the second keyword search is on a line by itself because if you list them like “keywordOne, keywordTwo” on the same line then Korsakow treats this as an ‘or’ search – so it will search for keywordOne or keywordTwo. This means you end up in the second stanza quickly, whereas listing the keywords on individual lines means the first search is performed, if a match is found, it is selected, if a match does not exist then it performs the next search – I only found that out by testing both options.)
So, that’s my current solution, which I think achieves the desire to have a sonnet which has stanzas, where the lines of each stanza can be viewed in varying orders, where the four lines in three stanzas and two lines in a final stanza can all be realised, while still allowing multiple passages. Yes, you have to start again each time, if you want, but, much like the strange prohibition on repetition, poetic networked objects can only be understood through reviewings, the old model of a single, comprehensive (start to finish) reading is not merely redundant here but hermeneutically wrong.