Just finished defining a process for recording, editing, and posting to the web video recorded during KJZZ’s weekly live radio call-in show, Here and Now. All on a budget of next to nothing.This was done using two consumer grade digital cameras, two tripods, Adobe Premier elements, a cd of the show recorded off the air, and some finishing touches using Windows Movie maker. It was inspired by Talk of the Nation’s interview with Michael More.

To do this, we used two digital still cameras, a Canon PowerShot A540 and a Canon PowerShot A640. These cameras are not video cameras, so they save video as mjpeg avi files (just a bunch of .jpg photos all grouped together) and not real video files. This is important to note, in that the files should be converted to a true avi with some kind of video compression (I used Indeo compression). I used a program called MPEG Streamclip to do the conversion. Converting the file seemed to require less demand of the CPU during the editing process.

The beauty of simply using two tripods: one trained on the host, the other on the guest, means that all you have to do is start the camera before the show and let it roll. The other benefit is that these cameras are really small and that they seem less intrusive than full blown television cameras, so people seem more authentic, like they’re not performing for an audience.

The other detail is that the audio that the cameras captured were quickly thrown away. The audio that you hear in the finished video came directly from a recording of what went out on the air. It was a little bit of a bear to sync up the audio. When previewing the video on a television screen, the voices looked a bit out of sync, but when the same files were exported to YouTube, the voices seemed very much in sync.

The video segments were put together using Adobe Premier Elements. (I couldn’t use Windows Movie Maker because I needed to work with two video clips on a timeline). As I previewed the video, to switch between the guest and the host, I spliced the guest’s video and turned the opacity to 0. This allowed the video of the host to be viewed. I don’t know if this is the right way to go about it. But this was my first afternoon working with Adobe Premier.

I exported the edited video, then brought it into Windows Movie Maker to add the credits and the music. I’m sure this isn’t the most refined way to go about it, but perhaps you might find this useful or encouraging. One other thing that you might find interesting is that, to find Creative Commons licensed music to use, I went to jamendo.com to grab the first instrumental track with an Attribution-NonCommercial-ShareAlike license that struck my fancy.

Throughout this process, I found that there was no shortage of opinions. Everyone I talked to had an strong views on how to do this. And everyone’s opinion was different. I’m not sure what to say about that, but I hope you found something in my description of this process enlightening.

The beauty of this is that by simply creating an xml file with links to images and their captions like this:
http://kjzz.org/news/arizona/archives/200708/borderbard/bardonborder/slideshow.xml.you can easily implement this player on the server side.Unlike a program like SoundSlides, you do not need to have software on the desktop to build the slideshow.

To do this, we used the XML Flash Slideshow program from flashnifties.com :

http://www.flashnifties.com/xml_slideshow.php

You can purchase this for $10, or they also have a free version.

While it’s not as integrated with the actual story, as described in the the Transom interview with Ben Shapiro here:

http://transom.org/guests/review/200706_ben_shapiro/

And while it is similar to the “Sound Slides” flash slideshow player

http://soundslides.com/

which NPR uses in any of the stories listed here:

http://www.npr.org/search.php?text=%22Audio+Slideshow%22

I think this will prove to be a useful tool for creating flash slideshows.

Tell me, what do you think? Is this merely a stop-gap technology towards creating videos?

What about the viral possibilities of this? Any ideas about the possibility of turning something like this into a “Slideshow Widget” to allow people to place on myspace or blogs?

What about the impact that this has on the page design? While I think it’s important to a photo with each story (which could also be distributed as part of an RSS feed) along with additional links, etc., how much do we place on a page? And how might this all be brought together into a more coherent whole?

Tell me, does anyone else have any experience with putting together a Slideshow using flash? What can you add to the discussion about your experiences?

In looking at the developer’s forum, I found that cetra3 created a version which includes a Seek Button and play bar, which allows people to view where they are in an audio file as it is playing – as well as allows people to select where in an audio file they would like to listen.

I am not sure how close this is to WNYC’s audio player. I had an issue with their player not recognizing the xspf playlists as they were dynamically generated for each story (some weird 404 error in the header — likely more an issue on my end than theirs).

I’ve also put together some handy javascript to allow people to copy the code and embed this on their blog or myspace page. I’m also using the xspf file to serve random audio promos and images for the public radio talent quest.

Thought you might find this interesting.

For example, here’s an audio story from KJZZ’s recent Chandler Roundup Series.

we must promote open standards for aggregating content – consistent tagging protocols at the station level, heavy use of RSS to pull content together, distributed content modules that can exist simultaneously on local and national websites, etc – to allow all us to mix and mashup these resources so they can surface at the local and national level.

Additionally, Craig Rosa’s comment about microformats had me thinking… how does the taxononomy/lexicon for describing objects using microformats inform how we might structure a database of web resources? While I know PBCore is noble and vast and everything, it’s primary goal is to be used for digital asset management – not for web sites (please let me know if I am wrong in my thinking about this). Are there content management systems out there which use microformats as a naming standard for metadata right out of the box? And how would this help facilitate a consistent “tagging protocol?” Also, in looking for how we might do this, I wonder if RSS is only the tip of the iceberg. Does this also mean a standard API for searching (and retrieving) RSS… if so, does this imply a standard REST protocol for querying a station web site and getting information back? I did a quick search for “common api standard REST” and came up with this modest proposal: Atom Will Change the World which contains a sprinkling of all the best buzzwords: Atom, GeoRSS, Dublin Core, etc. etc. But what it also says is that Atom is not only a syndication format, it’s also a publishing protocol:

In case you are not familiar with Atom, the syndication format provides a standard format for saving blog content in XML and the publishing protocol provides a standard API for clients to read, create or update Atom documents stored on servers.

This article goes on to say:

By combining a simple data model, a standard data exchange format, easy extensibility and a common API and simple specifications, Atom offers a great foundation for building web services.

and that Atom

provides a foundation on which anyone can build. Its much easier to build a Web service by adding some custom content to an Atom feed than it is to create a new XML exchange format and API from scratch. Thus, I believe Atom will become the de facto way of building web services. The first place to look is to Web 2.0 sites. Many currently expose their data via proprietary web service APIs – I’ll wager over the coming year most will move to Atom.

Shoot! Should I, as a webmaster at a public radio station, be implementing Atom feeds and moving towards an ideal of a standard API? And is the API already out there? Should I be learning about the Atom publishing protocol?

He’s done a great job with this! It’s interesting to note that he started with the radio story, and overlaid video that he captured during his interviews. He’s also done a good job at getting interesting visual angles and ability to think about what visuals he needs while out in the field.Among other things, we’ve also been discussing how much of the visuals could be done using still photos, and how to compile a list of “B-Roll” footage that you will want to get (visuals that you can know ahead of time that will help you tell the story). I’m sure Thomas will have more to say about his process and about the possibilities of thinking visually for radio as we discuss this here. I’m sure that through thinking about this issue there will be learning enough for us all.

I’ve have testing the ability to map KJZZ’s news stories here:

http://kjzz.org/news/map

http://kjzz.org/map

using, for example, the following rss feed:

http://kjzz.org/search/georss?keyword=immigration&size=5

http://kjzz.org/search/rss?keyword=immigration&size=5

I’ve also started testing this with Yahoo Pipes.

It would be interesting to see what a world map would look like if NPR was geotagging their stories, or what a national map would look like if we aggregated stories from public broadcasting stations across the country.

Are there other stations doing this, or considering doing this, with their stories? If so, should there a dialog between stations about aggregating stories and about the possibilties and best practices for mapping stories?

Is there some kind of national initiative for at stations across the system to start serving their content in this way. Is there support for Andy Carvin’s Election 2.0 Task Force where he says:

we must promote open standards for aggregating content – consistent tagging protocols at the station level, heavy use of RSS to pull content together, distributed content modules that can exist simultaneously on local and national websites, etc – to allow all us to mix and mashup these resources so they can surface at the local and national level.

What are your thoughts and experiences?

I love the way the text was used – like little breaks for narrative in silent films.

It was a real democratic way of telling a story. I’m always looking for democratic ways to quickly use technology to achieve an end and I think this technique really fits the bill. Think of it as Radio for YouTube!

I tried to find examples of this at NPR, and I didn’t see this exactly. I saw slideshows sure:

The World of Maurice Sendak
The Streets of New York
The Partisans of Ali
Emily Haines and the Soft Skeleton

But nothing really jelled. If there was narrative, there was no music or ambiance. If there was music, no narrative.

And only one video that I saw used text in a similar way. It was at the beginning of the absolutely chilling All Alone in the World from A War Photographer’s View of Iraq .

Imagine a lighter subject, imagine music, a narrator, interviews, photos and text. That’s what I’m talking about.

Andy Bailey, in the Filmmaker review says that Tarnation was made with a cheap video camera, consumer editing software, troves of home movie footage and less than $300. He says that it’s ironic “that a film originally created for less than the price of a plane ticket now has to obtain thousands of dollars in music clearances in order to move forward for distribution.” Okay, so edit the film with creative commons licensed music from the start, and you’re on to a real fast and democratic way to generate radio for YouTube.

Really. NPR, if you’re listening. In the absence of a YouTube for Radio. Make Radio for YouTube.

After all, we specify – within our Podcast feed (not our plain rss just yet), using the <license> element, that this feed is made available under a creative commons license that specifically prohibits derivitive works.

Is a mashup a derivitive work? Seems like one to me.

I did a cursory search of the web and found an interesting article by Jonathan Bailey which suggests that rss feeds are, by definition, public domain protected, syndicated property and should be respected as such. (thanks for the correction).

He also describes three approaches for removing a feed from the yahoo pipes service.

I know this is not the time for locking up our content, but maybe it’s prudent that we change the license for our feeds to something that is more reflective of its actual purpose and use (although perhaps retain the NoDerivs part of the license as a sub-element of each item within the feed – as opposed to the feed iteself).

My question was this, how do we make it easy for people to talk about our work, to cite it, to “copy it” to display it. Is it through an embedded flash player that we can make it easy to include the appropriate attribution information, the station’s call letters/url, the author’s name, the link back to the story. If there is a standard practice out there, please let me know about it.

For instance, I got an email from The Herberger College of Fine Arts at Arizona State University, they wanted to use an interview that we did with one of their professors, Jazz Studies director Mike Kocour. I babbled on in technicaleese about attribution etc and they simply said, “just give us the audio.” For all the slickness and polish of ASU’s site, when they finally post the audio that they said they just wanted, what will the presentation be like? Will this be accurately and equitably presented?

How do we as public broadcasters have this conversation? If we start with how to protect the rights of our own content, then maybe it will be easier for us to think about extending similar protections to user generated content.

I found out recently that Google’s calendar supports the iCal format (as does Apple’s iCal program, and Mozilla’s Sunbird). (Upcoming.org also supports an iCal export). So it would be possible that a station’s schedule could be edited in sunbird, saved to a DAV server and displayed on a station web site, or subscribed to within the google calendar, and distributed via google to people’s web pages as a widget. KQED has already done this here.

As a case in point, if this schedule:

http://kjzz.org/programs/schedule.ics

were actively updated, listeners (using Google Calendar, Apple’s iCal program or Mozilla’s Sunbird) could have up to the minute information about the KJZZ program schedule and copy information about any program into their personal calendar.

Additionally, this means that national organizations, such as NPR, if they were to subscribe to each individual stations’ feeds, could develop (or adapt) a parser to bring these ical files into their database, and thus actually display individual station’s schedule information at NPR.org. Other organizations and businesses who could benefit from this are publicradiofan.com , radiotime.com, and the individual program producers who could show which stations are airing their program, and when.

While this is not as ideal or complete as stations exchanging schedule information in xml, it is perhaps a first step in the right direction, and perhaps more useful to listeners in the end. As a matter of fact, this doesn’t have to be some prohibitive, technical rocket science… given the proper username and password, a designated staff member, can even edit their station’s schedule at their desktop (or at home) within Sunbird, and publish this to the station web site (or to any web server which supports the WebDAV protocol) – or they could use a hosted service like the iCal Exchange.

For those of you interested in displaying ical from within a plone site, Nate Aune (from plone4artists) has done some interesting work with iCal here :
http://plone4artists.org/products/plone4artistscalendar

A cursory search yielded a

PHP ical parser here:
http://sourceforge.net/projects/phpicalendar/

A parser for .Net:
http://sourceforge.net/projects/icalparser/

For Python:
http://codespeak.net/icalendar/

For Perl:
http://www.perl.com/pub/a/2005/08/18/ical_dot.html

Ruby too!:
http://icalendar.rubyforge.org/