This is just a quick list of things that were noted from the first k-film projects and so are useful to keep in mind for the final project. They're listed in no particular order but these are problems that should not be repeated in the final projects.
- Video thumbnails. Using video for thumbnails can be very cool. Why? Well, unlike jpeg thumbnails, they change when you mouse over them (they start playing), that by itself is just decoration, but from a user experience point of view it provides enough further information to clearly indicate that it is a different sort of object to just an image. This matters because you might have a film that has a jpeg on screen somewhere, but it is not a link or a button, but users can't actually tell because it will (almost) behave the same way as your thumbnails. Simple, but confusing. Video thumbnails can also provide a bit of additional context because when they are moused in to they play so before clicking on it your viewers can have a sense of what is there. This could also be a problem - why choose to view this by clicking on it if I can view it as a thumbnail? If you do use video thumbnails then the problems are simple. Do not have a first frame that is the same colour as your background, as it will be invisible. And you must recompress your video to your thumbnail size, otherwise you are wasting an enormous amount of bandwidth (which does cost money), slowing down your film (since it has to download the main video and video for each visible thumbnail to work) and generally being a poor online citizen by not recompressing them down to the thumbnail size. Recompressing them also gives you the chance to be creative. Why not drop the frame rate so when your viewer mouses into them they stutter a bit? Perhaps change their colour too, so they might be black and white, or have a blue hue. This helps identify them as thumbnails (they are a different tint to the other things on screen) but also means that when your viewer clicks on them they don't see the same thing only bigger, they see the same thing larger, at full motion, and in colour. A small, but valuable reward for having made a choice.
- Image thumbnails. Image thumbnails are good because they're very easy to make, they require very little bandwidth to load and play, and because you can be selective about what you use. Which is also the most common mistake made. The first frame of your video is not, necessarily, the best frame from which to make a thumbnail. Move through the video to find something more relevant. It might be more relevant because of the content, or it might be more relevant because you can choose an image that helps set up visual patterns in your thumbnail navigation (see the following point).
- Patterns are visible in the thumbnails. We have spent a lot of time talking about the importance of patterns to being able to interpret things. There are two places where patterns happen in your k-films. One is in the relations created between your clips. This then this then this, and then trying to think about what the connection between them might be. However, the more immediate and important place where patterns happen is in whatever style of navigation you provide your viewers. The thumbnails (which is why a k-film with only one thumbnail at a time for navigation could be a deeply dull experience) are where you can immediately see connections between parts, the patterns. This is because there are multiple clips appearing here, as thumbnails, and this is where I make the choice as to what I will view next, and this is where I can see the consequences of my choices. Yes, I click on one to view that, but as a consequence of that choice I don't just see that clip. I see that clip and then, via the thumbnails, I also see that clip in relation to several more - I can see this relation - not just try to remember it in terms of what I think I saw first, second, third. So, it is in your thumbnails that the structure of the work becomes apparent.
- The film has no title. Simple. It is a film, it is not called "untitled" it should have a name. Why? Because you made it and it took time. Because it is yours and you and others need to be able to refer to it (did you watch "Six Ideas for Darkness" and not "did you see that film at h-t-t-p-:-forward slash-forward slash" etc). Because a title can provide an enormous amount of context that orientates people to how to interpret the film, think of it as a song or poem title (because the works are relatively small, they're not feature films).
- Consistency of layout. There is a reason why in your blogs if your design has a sidebar on the right then it is on the right on the front screen, each category screen, archive screens, search screens, and so on. If your blog has a sidebar then it is using a grid (columns create grids). So think of your layout, your interface, as being on graph paper and so things should all fit in consistent numbers of squares. Edges need to line up. It makes your film look and feel much much better. Look at a newspaper, there is a lot of visual information there, it is quite easy to understand at a glance because of grids and the use of size and scale (within the grid) to indicate importance.
- Text size and location. Text can matter to your film. It does not have to have text, but what is written in relation to what we see can be very powerful. It can explain, contextualise, prompt. Offer commentary, create distance by being unreliable. Emphasise a film's poetic qualities by writing in a particular voice and way. And text is then also a visual element in relation to your film. Think of it not so much as something to just add some commentary but more like a credit sequence where text is always recognised as highly visual. So how big should your text be? Where on screen should it be? Should it have some autonomy, some visual authority quite apart from the video? Or should it quietly sit as secondary?
- Compression artefacts. This refers to the visible square blocks that appear in video that is at a low data rate. There is nothing wrong with such artefacts, think of them as like being able to see the brush strokes in a painting, but you might not want them. They appear because your source material is too low in quality and/or resolution so when it is recompressed by the Korsakow software it is going to be bad. Compression can't create data that isn't there originally. Or it is there because you are compressing at too low a quality setting. Compressing video is very slow, it takes an enormous amount of computing power (when I first started doing work like this to compress a video clip of a few minutes length would need to be run over night since it took so long and made trying to do anything else impossible), so be aware of this when making your projects. But if there are artefacts then they should be there because you want them, not because you don't know what you're doing. Perhaps you want it to be abstract, to be clearly digital, low res work. But this is a decision you control and you should not let this be dictated to you.
- Timed links. This is one of the key things viewing your projects has taught me. Timed links, in general (not always) suck. Why? Because if the main place in which patterns become present and visible to your audience is in your thumbnails then not being able to see them all at once constrains this, negatively. Because if a link (thumbnail) does not appear how do I, as your audience, know whether this is because a) the work is ended, b) there are no more links available (it is just a dead end, which might be different to the being finished), c) that I just have to might because there might, maybe, be a link. Finally, most of our k-films rely on loops and repetition. When I return to a clip that I've already viewed I now have to wait what feels like an unnecessary amount of time simply to be able to go somewhere else. You need me, the viewer, to want to hang around to experience your work. This is treating my time, my attention, as secondary to yours. That is not an economy or equation that survives very well here (how many times do you sit through the splash screen on a web site, every time you visit the same site and how soon do you click the 'skip intro' link? well waiting for a link to appear is the same thing).
- Watch your own work, let others watch it. Road test your film. View it, invite someone else to and do not tell them what to click. Just watch what they do, maybe ask them why they do it, but do not sit there telling them what to click. The aim of this is to actually see what real people, who don't know the structure, do with your film, what they understand about it, how they decide to 'drive' it. As makers you know its structure, you know all the content, so you are not able to judge. (In the first films I would click something and I'd be told not to keep clicking that, you don't get to sit with each viewer telling them what to click, and you should not have to.)
These pages are written by Adrian Miles ('official' and 'real') as part of the course material for Integrated Media One. This is a second year subject within the Bachelor of Communication (Media) program at RMIT University, Melbourne, Australia. These pages are notes to support students in this course and are best regarded as an informal aside to the lectures and workshops that constitute the subject. Views expressed here are not necessarily those of RMIT University.