vlog 4.0 [a blog about vogs]

Backend Work

So, been busy building a template to use as a basis for quite a few works. I’ll talk about why another time. But also wanting to start a blog that just has the video, nothing else. So been looking at the templates at press75.com where there are some video specific ones. What appealed are some of the gallery view options in some of the templates – just like the idea of that wall of images with video behind them – and the thumbnail and video plugins they have. Unfortunately they aren’t quite what I need, which is probably symptomatic of how I approach this video caper.

First of all the gallery stuff, I’m about to play around with it but I often make work that have odd shapes. For example the diptych that I’m using is over 1600 pixels wide but only 360 pixels high (it places two 16:9 videos alongside each other). But the galleries all seem to be assuming either a 4:3 or possibly 16:9 aspect ratio so how the heck will it cope with long skinny, or narrow wide thumbnails? Of course in the interests of neatness I might just have to crop and cut and hack the images into the default gallery size so the template works…

Secondly it seems the video plugin is a simple FLV player. Well, simple because all you need to do is a) upload your thumbnail (or point to its URL), and b) put in URL of the video (from a variety of service or self hosted) with width and height. Hey presto. Except since ti relies on FLV it means interactive QuickTime doesn’t work since interactive QuickTime requires QuickTime as its runtime engine. So what I thought would be a neat system actually doesn’t work at all. So, I’ll experiment a bit more, but right now I think the easiest way to embed video in WordPress is to use Pageot to write the embed/object code and stick that in your blog post.

Why you ask? (Well, I know most of you don’t but humour me.) Because then I can have a poster movie which when clicked loads the actual movie and can target a new window, QuickTime Player or the current window. Being able to target QuickTime Player through the embed code, without having to write an interactive movie, is good because when you make a movie that is 1600 pixels wide it tends to break in a lot of browsers so just making QuickTime Player actually display it means it will display properly. Also means it tends to run better since you’re using a dedicated app to view the work rather than the plugin.

So, back to the experimenting.

Recent Developments in the Diptych Department

Well, the riveting coding work continues. So I think it is time to pop out a work in progress so that things make more sense. The diptych here is not how things will be in the final work. I do want to speed up and slow down the video, but what I was doing was constraining this so that after, say, slowing or speeding it up by a factor of 5 (so 5 x faster or 1/5 of normal) I’d flip the track back to normal playback speed. After building it and experimenting and playing with it it just doesn’t make sense. As you’re clicking the thing slows, then all of a sudden after another click, Hey Presto!” we’re back to normal speed. Nah. So this work in progress is just to see this in action. What hasn’t been done yet, well, both soundtracks remain, these will be removed and one soundtrack will be added as a separate child movie so that it plays normally independent of the image track speeds. This means the effect you get when mousing into either video so that the volume for that track changes will be removed (though in this demo it does create some pretty weird aleatory outcomes). So that also means I’d better let you mute, or perhaps pause?, the soundtrack.

Next Iteration?
For the final version clicking the slow and fast buttons (‘-’ and ‘+’) will just keep speeding up or slowing down the video. QuickTime from memory doesn’t like to drop below 1/10 of normal speed (normal is whatever the video frame rate is) while I think acceleration probably depends on your CPU. Mousing into the other button (eg if you’ve sped the thing up, then mousing into the ‘-’ button) will restore the current video to normal speed. Then add a pause button for the soundtrack, and perhaps see about making the buttons semi-transparent so they’re not quite so there. That should do it. I think.

Coding Development

Well, been scripting, sketching and testing. The pathway to date has been:

1. make a diptych
2. mouse into either video increases volume of audio for that track
3. you should be able to pause either video
4. mouse out returns video to original (approx 10%) volume
5. include time based links in each of the videos

Then it also became

6. let either of the videos (in either pane) able to be accelerated or slowed

This has raised all sorts of interaction questions and issues. For example, mousing into either video to change soundtrack is not intuitive, and depending on the video it may not be clear that mousing actually does anything. Now I prefer mouse over and mouse enter to mouse click for making things happen. It makes the videos more like touch movies, things you have to caress, rather than these bloody clicky things. However, in adding the buttons to speed up and slow down the movies it seemed that a mouse click made more sense since I can imagine people just not getting what goes on when things happen just because you mouse into something. Of course changing content or style as a direct response to the mouse in helps (like the way you can make links change when you mouse over them in HTML using hover in your CSS), but that is different to actually having the work do something qualitative (eg change speed) because you mouse into something. The bigger issue that this then produces is inconsistency – you mouse into the video to effect a volume change, but you have to click other things to make other things happen. This is not good.

Then to compound it once you start speeding up or slowing down the video then mousing to change the volume becomes irrelevant. Once the video is going faster you just get very fast sound, and conversely when it is going slow you don’t get much at all – who cares and what does it matter if you can, or can’t hear it?

So right now I think I will remove the sound change sprites. I will also remove the soundtrack from the video and actually run that as a separate, third childmovie. What this means is that the sound will play, at normal speed and volume, regardless of the playback speed of the video windows. I haven’t implemented this yet, but particularly like it (well, until I test it). This way you can slow down, accelerate, stop the videos but the sound track (and possibly commentary?) will continue. Of course I then might need to stick a mute button on there too. mmm.

As usual with me the scripting is slow. I’m not what you’d call a programmer. I got the faster and slower buttons working fine, basically it counts mouse clicks and the faster does an argument where the film speed is set to multiply by the number of clicks, up to a maximum of 5. Therefore it plays at normal, twice, triple and so on up to five times normal speed. The slower button I divide normal speed (what QuickTime thinks of as a rate of ’1′) by the number of clicks: so 1/1, 1/2, 1/3, 1/4 and 1/5 and after 5 the variable (mouse clicks) is returned to 1. This though causes problems, since you can speed the film to 5 times normal, then 2 clicks in the slow and the movie is now 1/2 normal speed, not 1/2 of what it was before. eg If I click 4 times the video (lets assume it is 24 fps) will not be running at 96 fps, so clicking on the slow button you’d think should say, half the current speed.

The problem with all of that is that if I’ve made the video faster and it seems that this button should make it slower, then that’s what it should do. So, the first attempt at this I counted mouse clicks and saved them to a variable and used this in both sprites. Quickly figured out that was pretty stupid, clicking the faster one 4 times to get a video 4 x normal speed would mean clicking once on the slow button the video is not 1/5 of normal speed. That just seemed rather a large change. So I spent a lot of time trying to work out ways to do all of this, and the solution I’ve finally got to is not ideal, but is a bit clearer from a user experience pov. Basically for each sprite I now check for current film speed. For the go faster sprite if the film speed is less than ’1′ then it has been slowed down, so I then make the film speed 1 then start counting mouse clicks to accelerate it. This means if you’ve got the video at 1/5 normal speed and you go to click the go faster button then it restores playback speed to normal, then starts going faster.

Similarly for the go slow sprite it checks to see if the film speed is more than ’1′. If it is then the user has been clicking the go faster sprite and so again if you click the go slow button the first thing it does is restore the film speed to normal and then start counting mouse clicks to slow the video down. I tried to work out a more elegant solution, for example simply multiplying current speed by mouse click to accelerate the video, but the maths doesn’t work. For example if the video is playing at 1/4 normal speed and you then click the go faster sprite, then I need the video to now be 1/2 normal speed, the calculation to get that would be 1/4 x 2, and I could get the mouse count to 2 for this, but then click go faster again and the calculation becomes 1/2 x 3 so instead of getting to normal playback it is 1 1/2 times. This might not matter that much (though once you start playing with the work you’d never get normal speed again), but then when you start slowing it down (and so dividing it by the mouse count) you just get all sorts of variations – 1.5 normal speed divided by 1, then 2, then 3…

Ah, but you see I’m such a poor programmer that in writing this out I see that I’m doing this all wrong! All I need to do is to divide or multiply speed by a constant (say 2) and just make sure it doesn’t work past, say 5 iterations (5 times acceleration) and things would work. Sheesh, I really struggle with this stuff because now I can see just how bleedin’ simple and obvious that is. OK, let’s return to the kitchen and try that one out.

UPDATE FROM THE KITCHEN
Bugger. The two actions available are SetRateTo(x) and SetRateBy(x). The first is an integer that just sets the playback rate (1 for normal, 2 for double, etc). The second you’d think might be a multiplier, eg if you entered 2 then it would double, but the ‘By’ in SetRateBy just adds, so if you put in 2 then it just adds 2 to the current playback speed. If it were at 1, then you’d get 3, 5, 7 etc. This makes it good for the go faster sprite (SetRateBy(1)), but for the go slower sprite there aren’t any numbers I can add to 1 to end up with less than 1! Also if the number becomes negative the video plays backwards, which is not the effect I’m after.

Link Movies

I’m working on some new videos. The videos themselves are simple and straight forward, but the architecture for them is taking some trial and error, or what in academic talk I’d describe as ‘iterative development’. The basic structure is a diptych, with most likely the same video in each video pane. They will play alongside each other, as child movies (think Chinese boxes, complete movies inside other movies) so they can have independent play states. One will be pretty hi-rez, the other lo-rez. I’ll talk about why another time. At the moment they have their audio set quite low, and when you mouse into either video pane the audio for that panel goes to full volume, but I also want links in the video. These are links to external sites. The issue is that these links will be time based and so appear, then disappear. So, when you mouse in to such a link what should happen? And when you click what should happen?

The first experiments I did I just had the sprite (the image you mouse into) change the mouse to a pointing hand and changed the image to indicate that it was a button come link. That was OK, but it does not tell you anything about what the link come button does. So the next iteration I replaced this with a 50 x 50 image of the destination, for example a screen grab of the web site that the link is to. It is too small to tell you where you’re going, but it is recognisable a web page so you should be able to work out pretty quickly that it will take you to a web page.

But when you click should the movie stop, or continue? It sort of makes sense to stop, since you’re being taken somewhere else and once you’ve gone there you can always return (if interested) and continue where you left off (the click creates a play/pause effect, click once you stop and load url, click again and it resumes playing). But remember there are two videos playing alongside each other so the click should then probably pause both videos. Alternatively I could provide simple controllers for each video which lets you play and pause each video pane and so you could then stop the video, and follow the link, and then it would make more sense to just let the link be a link without pausing the video (it is not a good idea to make that stop and link when you already provide a sprite that play and pauses the video). The problem with this one is that now things are getting pretty busy. I can play the video. Stop it. Times each video pane. And I can click on a link which loads a URL and if the video is not stopped it will keep playing, or if the video is stopped then it stays stopped.

I like the idea of the link stopping the video, but then you’ve got no way to stop the videos if that’s what you want to do, which is generally a pain in the arse. The other issue is that since the links are time based, i.e., they disappear, do you need to be able to scrub or return to follow the link because you didn’t earlier, or there was more than one to follow? I think what I should do is:

• let the videos be slowed down and accelerated
• links should not pause either videos
• clicking the link again should resume
• a simple play/pause button should be provided (that stops both or one for each?)

That way you can slow them down if you like, so playing with their speed in relation to each other. You can follow the links, and you’ll quickly learn they don’t stop so if you want to stop then stop then click the link, and we’ll see how that goes. Oh, and I think I’ll install a new WordPress blog which will just host the video content (perhaps using one of these templates), stuff is just so lost in here and it is perhaps time to just have a single place to collect the video work.

Vogging

Have started sharpening the mouse for my coding skills in LiveStage Pro (which still runs happily, after several system updates, moving to another computer – all in spite of the company now apparently bust). I’m making diptychs where mousing in changes the sound level, and I’m also playing with adding links in the videos. At the moment the links are URLs out to pages around the place, but I am thinking through making the links also go to other video works. The problem with this is that I neither need to treat these videos as an individual project, since earlier videos aren’t going to have links, or just use tags and links in the text of the accompanying blog post to do this.

This is one of the issues with serialised authoring, as while you can (if you can be bothered) link to earlier content you don’t generally make a habit of going to earlier posts and linking to more recent ones. Hence the internal link structures in your own blog (or other serialised form) tend to be only backwards looking. One solution to this is the development of ‘related’ or ‘more like this’ plugins in blogs so that stuff out of the archive can be joined, backwards and forwards, through your material. To date I don’t think there’s a very good solution for video in this context. Related links sort of work, and I imagine a series of gallery pages (think rows of thumbnails) that are the result of tag searches would also be useful for this.

Systems like Korsakow for video (and Tinderbox for text) solve this since they are database structures where you add new material and then there are rules you have written which define how things link to or relate to each other and how they are to be presented. The advantage of this sort of system is that relations are just that, and so can join old to new and new to old.

YouTube Reader

Following hot on the heels of “YouTube: Online Video and Participatory Culture” there is Snickars and Vonderau’s “The YouTube Reader“. Anthology, seems to have some big name authors in there.

Academic YouTube Text

Jean Burgess and Joshua Green’s “YouTube: Online Video and Participatory Culture” is out. Haven’t actually seen it yet, but on the list. Not really quite my specific field, more about participatory culture I suspect than video per se, but will be an important early book in the field.

Emergent Funding

http://buyacredit.com/ where a British pound will get your name on the credits of a feature film. This is similar to a simple videoblog project that Jay helped initiate where people would donate a small amount to establish a pool of funds that could then be awarded to applicants. These sorts of things can work as long as the scale works. The long tail is good for these sorts of initiatives if you have a community large enough to sustain it.

Closures

Just received an email: Jumpcut is closing. Part of the prioritisation efforts at Yahoo! which I guess translates into it is not making any money (is anyone who hosts a ton of video content?). This comes hot on the heals of the disappearance of the Ezedia suite of QuickTime tools. Video software development, outside of edit systems for traditional practice, remain a poisoned chalice. LiveStage Pro, eZedia, VideoClix. All gone.

What does this tell us? That we are perhaps at a similar point to where we were with hypertext systems in the early 1990s. Storyspace, Intermedia, even possibly HyperCard. Some were brilliant, so far ahead of everything else (HyperCard for example) that only a very few ‘got it’. Some argued for why hypertext mattered, but most outside of industrial style systems (eg help systems for computers and custom systems to manage things like military documentation) could see little use. Then along came HTTP, HTML, and, well, here we are. So I remain sadly optimistic that something will appear, somehow (and it certainly is not YouTube), and software will start popping up all over the place to let us make video that is deeply, joyfully intertwingled and network savvy.

Video Vortex 4 Schedule

Straight off the email:

VIDEO VORTEX 4, SPLIT-CROATIA, MAY 21-23 2009

OPENING EVENING: THURSDAY 21 MAY 2009

17:30 Afternoon Screenings Preview at Kinoteka Zlatna Vrata, Split

19:00 Opening evening at Multimedia Cultural Center, Split
Word of Welcome by Geert Lovink, Miranda Veljačić and Dan Oki
Introduction speech by Lev Manovich (to be confirmed)

20:00 Exhibition opening (with food buffet)

21:00 Performance: Emile Zile – Post-It Kino

(Buses to the hotel at 22:30 and 23:00)

DAY ONE: FRIDAY 22 MAY
Conference at Multimedia Cultural Center, Split
(Buses to the Conference at 9:00 and 9:20)

9:45 – 11:30 SESSION 1: Tele-image Research Strategies
Moderator: Sabine Niederer
Presentations by:
- Andreas Treske
- Nathalie Bookchin
- Dalibor Martinis
Discussion

COFFEE

11:45 – 13:45 Session 2: The Database
Moderator: Tomislav Medak
Presentations by:
- Maarten Brinkerink
- Kuros Yalpani
- Albert Figurt
- Alejandro Duque
Discussion

LUNCH

14:30 – 16:15 Session 3: Video Art meets Web Aesthetics
Moderator: Leila Topić
Presentations by:
- Vera Tollmann
- Vito Campanelli
- Sarah Késsene
Discussion

COFFEE/TEA

17:00 SCREENINGS
Presented by Dan Oki
- Lemeh42, Study on human form and humanity #1, (2′00″).
- Cornelius Onitsch, *AV*, (5′00″).
- Ivana Runjic, Show me your hard disk and I will tell you who you are, (7′00″).
- Nathalie Bookchin, Parking Lot, (15′00″).
- Shelly Silver, In Complete World (53′00″).
Q&A

20:00 CONFERENCE DINNER
(Buses to the hotel leave at 22:15 and 23:00)

DAY TWO: SATURDAY 23 MAY 2009
Conference at Multimedia Cultural Center, Split
(Buses to the conference leave at 9:00 and 9:15)

09:30 – 11:15 Session 4: Online Video Theories
Moderator: Geert Lovink
Presentations by:
- Jan Simons
- Gabriel Menotti
- Amir Soltani
- Stefan Heidenreich
Discussion

COFFEE

11:30 – 13:30 Session 5: Online Video Narratives
Moderator: Brian Willems
- Jasmina Kallay
- David Clark
- Valentina Rao
- Paul Wiersbinski
Discussion

LUNCH

14:15 – 16:15 Session 6: Politics of the Moving Image
Moderator: Petar Milat
Presentations by:
- Sasa Vojkovic
- David Teh
- Ana Peraica
- Antanas Stancius
Discussion

COFFEE/TEA

16:30 – 18:00 Session 7: Social Cinema
Moderator: Dan Oki
Presentations by:
- Perry Bard
- Evelin Stermitz
- Dagan Cohen
Discussion

Evening: 19:00 CONFERENCE DINNER

21:00 PERFORMANCES
- Cym, 30′00″
- Surprise Act

22:00 VIDEO VORTEX PARTY
(Buses to the hotel leave at 24:00 and 02:00)