Monthly Archives: June 2008

Augmented Reality – blurring the line between real and virtual worlds

I’ve been keeping an eye on various blogs and panlists relating to Second Life and virtual worlds, and will probably be sharing interesting tidbits here from time to time. Today, I wanted to briefly mention Augmented Reality, a new 3D technology that’s been causing a bit of a buzz on one of my panlists. Augmented Reality is the real-time integration of virtual 3D content into real-world environments. It’s pretty useless to try and explain it without a demonstration, so check out a YouTube clip (there are plenty of others on YouTube if you search for them):

http://kr.youtube.com/watch?v=O2i-W9ncV_0&eurl=http://arsecondlife.gvu.gatech.edu/index.html

The videos look like someone did some cool CGI on a home video or something, but actually, all of the imagery was superimposed on the scene by a computer in real-time, with no behind-the-scenes human animator. Given a special Augmented Reality headset, you could walk around a real room and interact with both other people and life-sized SL avatars! I know other research has followed similar lines, but this was the first case I’d found where instead of creating a new virtual reality specifically for their technology, someone created technology to integrate in with existing virtual content.

Livescribe Pulse smartpen

I just found a draft of this and realized I never posted it – I spent an hour or so a few weeks ago trying out the Pulse smartpen from Livescribe, and I wanted to write a bit about it here. For those who weren’t here to see me test-driving it, the smartpen (according to the Livescribe website) is “a computer within a pen that captures handwriting and simultaneously records audio and synchronizes it to the writing. Users can simply tap on their notes to replay what was recorded from the exact moment they were writing, so they never miss a word they hear, write, or speak.”

The pen comes with a special Livescribe notebook, which at first glance appears to be of the normal spiral-bound college-ruled variety, but upon closer inspection proves to have a faint pattern of microdots covering each page. This is how the pen keeps track of position. You can write something while recording, then go back and tap any word, or even an individual letter, or even an individual part of a letter (the cross of a t, the middle of an l) and a tiny speaker on the pen will play back the audio recorded at the exact moment that word or letter was originally written.

One of my first concerns was the quality of the audio recording — would it pick up a professor lecturing a hundred feet away, and if so, would the professor’s voice be audible over the much-closer background noises? — but a few informal tests showed that the audio capabilities are actually quite good, producing a relatively true-to-life recording. As an added bonus (although I couldn’t see any practical use for it,) the included earphones have tiny microphones embedded in them, allowing 3D audio recordings and playbacks.

Aside from the audio-recording-synchronized-with-your-notes feature, the biggest practical benefit from using the pen — and one that is not mentioned in any of the promotional materials — is the automatic digitization of anything you write with it. The pen has a tiny camera near the tip, and using the microdots on the paper as a guide, it keeps a running record of everything you write and combines them into an image of each complete page. The pen’s tiny USB device lets you call up these images on the computer (in the Livescribe Desktop, which can be downloaded from the company website,) as if you had run the pages through a scanner. However, they are not simple images – you can click on any word or doodle or mark to hear the corresponding audio, just like on paper.

Furthermore (and I found this to be the most helpful feature of all,) the program has Search capabilities. You can search the (handwritten!) notes for words or phrases. Anyone who’s ever searched through pages and pages of a notebook looking for a single fact or specific lecture will appreciate this feature. It also was a bit of a surprise to me, as I did not realize that the pen was “understanding” what I was writing and converting into computer-readable words. It was very cool, the first time I typed in a word and saw the program highlight parts of my scribbles.

The side of the pen has a small, elongated computer screen, from which you can explore a number of interesting features – for instance, a “Translate” option, where one selects a language, and then can write any English word and hear the pen speak the translation (!). (The pen comes equipped with demo-translators in a few languages, capable of recognizing about 20 words each, but full translation packages can be downloaded from the website.)

Some of the neatest capabilities arise from the nature of the “paper-based computing” platform. At times I felt a bit like Harold with his Purple Crayon – you can doodle a piano on paper, and then tap the keys and play the doodle as if it were a real piano!! Or instead of using the little menu buttons on each page to interact with the pen’s computer, you can simply draw your own buttons and use them.

Concluding thoughts: extremely nifty, and very fun, but perhaps not the most practical of tools. Having an audio recording of a lecture synchronized to your notes could be useful, as could having search capabilities on handwritten notes, but with the growing prevalence of videotaped lectures and typewritten notetaking, these might not be necessary. The other features strike me more as cool toys than as needed capabilities.

Shareable Metadata

Jen Riley

http://www.dlib.indiana.edu/~jenlrile/

http://www.dlib.indiana.edu/

http://www.openarchives.org/

The idea of shareable metadata is slowly taking shape in Library cataloging. My connection wiht Matthew Beacom on the metadata team led me to this lecture. It was a broad overview on how the Univ. of Indiana library is approaching adding metadata to digital and non-digital resources. The idea is to create a record that is broadly distributed to aggregators. Below you will see my notes from the lecture.

How in IT can we put this to practical use. IT is in a unique position to offer faculty an entry level cataloging structures that either ingest this shareable metadata from library resources for use in publishing and research or capable of mapping to the library if it is a personal collection that the library may absorb.

Indiana is a user of OASIS open archives initiative uses dublin core xml schema and outputs to google image search, artstor, open worldCat http://www.worldcat.org/, OAISter http://www.oaister.org/, CIC (not sure what this is), IMLS http://www.imls.gov/, GEM, and other collection registries

Metadata is useful even when objects are not digitized. Quality metadata fits it’s purpose, promotes search interoperability and is human understandable outside of its local context.

A metadata record is a single view of that resource. For example art history data vs. history data speaks to a different office.

The 6 c’s of shareable metadata:
1. Content – how element values are structured, define the audience and choose the vocabularies. Content standards include granularity of description, version of the resource to describe and elements to use. Do not put in empty fields or non-useful info (scanning info) because the technical processes that sort can have trouble with this.
2. Coherence – the record should make sense on its own, place vlaues in appropriate elements, repeat elements instead of adding several to one field, avoid local abbreviations and codes, ensure mappings from local shared metadata formats.
3. Context – allows user to understand a resource based on the record alone, shareable metadata records should include info not used locally and exclude info only used locally. Collection level records can help to point users to other related items. As soon as a record is extracted from its local context it needs that data to travel with it.
4. Communication – info supplement to metadata can be used by an aggregator. It needs to let the aggregator know the intended audiences, record creation methods, controlled vocab used, content standards used, accrual practices (are things continually added), analytical or supplementary materials (other places that should be linked), and the provenance of materials (where did they come from).
5. Consistency – with consistent metadata creation processes, aggregators can apply the same indexing or enhancement logic to an entire group of records. For example, how elements are used, how vocab are used for a particular object
6. Conformance to standards sharing protocols (OAI-PMH), metadata structure standards, technical standards

High level workflow.
1. Plan by choosing standards – who are you sharing with, choose shared metadata formats (what formats can they handle) write metadata creation guidelines.
2. Create metadata – keeping in mind shareability.
3. Transform metadata – perform conceptual mapping, perform technical mapping, validate transformed metadata, tests shared metadata with protocol conformance tools.
4. Share – implement sharing protocal, communicate with aggregators.
5. Assess – who is using the metadata, how does it appear in the aggregator. You may need to go back to the transformation step at this point and massage the data.

Tools dictate parts of our workflow, tools should serve us not vice-versa. Start workflow design from well-defined goals. Ensuring shareability is the common theme underlying it all, generate multiple views from a single master.

It is helpful to define categories of material by resource type (i.e. text, documentary images which have different metadata then art images, art images, musical audio recordings or more specific data. Often the category is defined by managing institution – this can provide a barrier to end users but it does take into account institutional mission.

There are certain resuable parts of workflow:

  • decisions on metadata structure and standards
  • metadata creation tools
  • automated processing techniques
  • xslt stylesheets and other data management code
  • sip/aip/dip architecture – information package types
  • delivery systems

Generalization will allow you to minimize the effort of redoing something and focus more on the new stuff, fixing metadata is a necessary all the time.

Make the most of automation. Automate repetitive tasks as much as feasible but where it makes sense.

OAI data provider registry http://gita.grainger.uiuc.edu/registry/
Lightweight API (ex OpenLibrary -http://openlibrary.org/)
google siteMaps https://www.google.com/webmasters/tools/docs/en/protocol.html http://www.xml-sitemaps.com/
OpenURL http://en.wikipedia.org/wiki/OpenURL
SRU http://www.loc.gov/standards/sru/
OAI-ORE http://www.openarchives.org/ore/
linked data

One highly complex metadata structure (but low use) is Z39.50 – client-server protocol http://en.wikipedia.org/wiki/Z39.50

There are a variety of options for accessing data and we should look to the future of collaboration and allow users to add data.

Terminologies services http://www.oclc.org/research/projects/termservices/ can help point one set of data to other resources.

Necessary tools
record level validation, data type validation auto-complete spell check integration of metadata creation guidelines into software tools
ingegration of controlled vocabularies

The library can focus on working around system limitations found in digital asset management systems.

Shareable metadata is a shift in thinking, metadata is no longer a destination but a building block, metadata must operate in an increasingly decontextualized environment.

The True Demise of Web 2.0

Essentially what gave birth to the second generation of web development was a suite of new tools and applications which made it possible and expedient for a class of web enthusiasts to project their personalities, skills, and ideas into the global, virtual sphere. Among these tools were the creation of the “weblog” or “blog,” and Blogger in ’99 which helped jumpstart the movement, Wikipedia in 2001, then MySpace, and Flicker. As the userbase of these services grew, so did the amount of web “territory” devoted to personal and creative expression. In retrospect, web 1.0 seemed like a huge white-washed monstrosity of sterile information and monochromatic data, or what the endless stretch of subway and city walls would be without the brilliant, vibrant colors of neothematic graffiti. Yet I must say, although there are only a few years between now and that glorious dawn of the second internet millenium, I believe what brought Web 2.0 to life, is steathily bringing it to its knees.

Don’t get me wrong, I’m not saying that the internet is becoming less popular or any less enjoyable, after all, there are nearly a billion and a half internet users across the world. (That’s undoubtedly an underestimate.) What I’m saying is that the conceptual expansion and growth of Web 2.0 is becoming stagnant. The idea that every avid internet fan can and should delve into web development and programming is not producing results that favor expansion of internet usability. A quick glance at Sourceforge, the influx of new “services” and web communities, should give onto the conclusion that I have reached. Web developers are using their expertise to make an endless array of “tools” and “widgets” that generally boil down to be some old idea recycled, reskinned, and given a new name and URL.

What if the masters of the Renaissance had bent all of their creativity to making new kinds of paintbrushes, canvases, and easels instead of producing art? We would scarce have much to look back and marvel upon from that age. Or if Gutenberg’s printing press had spawned instead of a mass of printing of scholastic texts, a host of printers who were instead devoted to making different types of printing presses and gadgetry. Basically, what I’m getting at is that web developers need to devote more attention to actually developing the virtual world. This could mean either using tools already in existence, or making new tools to achieve a creative and innovative end. We must stop making tools for the sake of making tools. Web 2.0 fails when for every 50 new web user-developers it inspires, 40 of them are devoting their time to making new weather gadgets, calculators, CMSes, and social networking sites, while only 10 of them are using services and applications already in place while trying to keep up with the infinite variations being regurgitated by their peers.

DVD Studio Pro + Lynda.com

Having recently completed a lynda.com training series on DVD Studio Pro 4 – I think it is due time to blog about the experience – as well as the numerous things that I have picked up from the series.

To begin – the method employed by lynda.com is that of video tutorials done by experts in a particular program, which I think is highly effective, especially if you tend to be a visually oriented person. Further, seeing the different operations in a screen capture allowed me to have a memory of what the operation looked like, rather than it just being explained. The lessons are broken up into larger sections that have a heading, under which videos ranging from 1 minute to 15 minutes (depending on the topic) are listed. There is usually also a few “basic” headings, which one would be able to view and get a elementary grasp of the program at hand, paired with more “advanced” headings that allow one to gain more skills and learn more difficult processes beyond the basics.

So, on to DVD SP. The program is included in the Final Cut Pro Suite, and is thus most readily suited to be used in direct correlation with FCP. Similar to FCP, the interface of the program is not quite intuitive, that is, unlike iDVD and iMovie, one would probably not be able to just open DVD SP and end up with a decent looking DVD an hour later. However, like FCP, DVD SP allows one to make some extremely complex, beautiful, and professional end products in a way that is not always possible in iDVD. There are many tools in DVD SP that really allow you to take control of the DVD creation process, and once one learns how to use the program, there are many elements which are actually easier than iDVD. For example, creating one’s own theme in iDVD is certainly possible, but is fraught with many aspects that can make the process quite frustrating. In iDVD, the order of button selection is tied to the order in which assets are brought into the project, whereas the user has complete control over the buttons in DVD SP.

More about the technical aspects. To begin, DVD SP is primarily a “wiring” program, that being a program which takes assets from other places and “wires” them together into an end product (that being a DVD). So, DVD SP is not for creating or designing menus – which can sometimes add an extra step onto your workflow (albeit one which can result in a better end product). DVD SP, like iDVD, has templates already created for menus, buttons, chapter indices, etc. – but also allows the somewhat simple import of a photoshop document that can serve as a menu. It is a somewhat simple task to create a menu in Photoshop, even including cool looking buttons and text, and bring it into DVD SP. DVD SP will recognize buttons and text that you create in a Photoshop document (as long as it is on a different layer) and will allow you to have it be able to be selected.

Rather than make an exhaustive commentary on everything I have learned (you can go to lynda.com for that!) – I found the tutorial to be helpful in generally creating DVDs, in iDVD or DVD SP – I will outline a couple of general points under headings, that should allow a future reader to come out with a better end product where they making a DVD!

Menus:

1. Always use a minimum of 24 point text in buttons, titles, etc. – any smaller and the difference in computer screen resolution and TV screen can make the text difficult to read.
2. When creating buttons, try your hardest to create buttons that have a symbol appear next to the words, or a line underneath them, or a box around them, etc. when selected. Try to avoid having the text itself be highlighted – DVD programs do not allow aliased objects and your text will look choppy and mouse-eaten when highlighted.
3. Always be sure your buttons are tied to an asset – test this first in the simulator (DVD SP has a tool that will search your entire DVD and tell you if you have any unconnected buttons).

Finalizing DVDs:

1. TEST, TEST, TEST. There is nothing more frustrating than spending two hours compressing a DVD, only to find out that your misspelled “manipulating.”

2. To test your DVD, first look at it is either in the graphical tab in DVD SP, or look at the DVD map in iDVD – make sure everything is tied together and will play the way you intend (in DVD SP you can use the tool to make sure there are no loose connections here, iDVD typically ties things up automatically for you).

3. Next, use the simulator to check the actual playback – press each button, play the assets to the end to be sure they return to the main menu. Use both the mouse and the remote control to be sure both computer and set-top users will be able to view and access the DVD in the way that you desire.

4. Now, either “build” your DVD (in DVD SP) or choose “File” –> “Save as VIDEO_TS Folder” (in iDVD). This will take some time, but will allow you to open the resulting VIDEO_TS folder in the DVD player on your computer without burning any discs yet. Never choose the “build and format” (in DVD SP) or “burn” (in iDVD) as neither gives you a chance to do this process, and they will also require the long process of encoding to go again should you desire to make more copies at a later time. Again, click your buttons, watch the assets, and be sure both remote control and mouse controls work.

5. Now using Toast – which I found out is the preferred burning program for the pros, choose “Burn VIDEO_TS Folder” and burn your DVD. Some tips about burning:
- Slow down your burn speed (2x-4x), it will take longer, but will be less likely to corrupt the DVD.
- Do not try to burn a dual-layer disc – these fail 90% of the time.
- Use Toast Titanium to burn, it verifies and often will fix small problems with burned DVDs.

6. Check your DVD in a set-top player and computer, and be careful when labeling the disc.

General:

1. Computers and TV sets often differ in the shape of pixels, so if you do not want a stretched-out looking photo in a background, it must be cropped in iPhoto or Photoshop to 720x534x72 (odd I know, but it works).

2. Try to avoid JPEG files as backgrounds, they save space, but can pixelate on a TV easily. Try to stick with TIFF or PNG files.

3. When undertaking a DVD project that you know will push the space limitation of a DVD (4.3 G for a burned disc = about 2 hours of video) – be sure to “bit budget” – that is, be sure not to create menus or other additional material which will result in more than 4.3 G of information (you can compress your video further to fit more on a single-layer DVD, but the quality rapidly degrades as you do so).

4. If your DVD is complex, or involves a lot of menus, it is a good idea to make a sketch of how you want the DVD to flow before you start working on it (I do this even on simpler DVDs, just to have a clear idea of where I want to go).

That is about it! A lot I know, but hopefully it will prove helpful to someone somewhere. The materials above are conglomerations of what I learned and my own practice and experience.

ITG-TimeCapsule

I set up the ITG-TimeCapsule on Friday, so I decided to post the screen capture demos here that I made with Jing. I explain how to join the wireless network, connect to the printer (Dewback), and configure TimeMachine to back-up to the TimeCapsule. The screen captures are as follows:

How to join the ITG-TimeCapsule:

itg-timecapsule

How to configure the ITG printer:

itg-timecapsule-printer

How to configure TimeMachine to back-up to the TimeCapsule:

itg-timecapsule-timemachine

Jing(le) All The Way

I found this really neat utility called Jing (http://www.jingproject.com/) that allows users to either capture screen shots or record a video (of your screen) and then upload the file to a server, including Flickr (the files can also be saved locally). I think this would be a great tool to use for creating FAQs since a user can do everything on the fly and it is super easy to use. When making movie files, they are saved in .swf format. Also, audio can be added to the movie files for narration. An added bonus is that the client is available for both Mac and PC and it’s free!

Here I have posted a screen shot (with annotations) to Flickr: http://www.flickr.com/photos/27607674@N02/2591589352/

The movie (I don’t have a mic on my home computer so it is silent) is posted at ScreenCast:
http://www.screencast.com/t/yZVDdwob91

Note: If the movie does not play properly, it may be due to the latest version of Flash Player:
http://blog.techsmith.com/screencast_blog/2008/02/flash-video-difficulties.html

The Jing site has more information, of course, but one of the demos is of the narrator sending a friend a video tutorial on how to use the Jing software via IM. Just think of the possibilities.

Metadata Hoedown

First of all, I think Ryan’s observations below are absolutely correct.

I blogged a little bit about this issue a year ago, when trying to figure out how to intelligently parse wmv (asf) files for descriptor data. (“Meta-Headache, or dealing with metadata in various media wrappers”)

If you’re on a Mac, Spotlight actually sports a fairly sophisitcated metadata index function, and you can use the super-cool “mdls” command to isolate and print descriptor data for pretty much any file – even if the metadata is hidden away in some obscure header. Of course, as Ryan mentions, we would need a skilled programmer to put together a script to make this function useful for mass exportation of descriptor data in an intelligible format.

Metadata Hootenanny is a decent app. But, for our purposes, a new program called “MetaX” might be even better. Not only can it automatically generate metadata based on earlier entries and templates (or even from IMDB and other online repositories), it can export your descriptor data so that you can “share” it with others. It also allows you to do loads of other cool stuff, like setting poster frames or dividing your movie into chapters. More experimentation is needed…

If ease-of-use is a factor, another program that looks extremely promising is RoadMovie. It will batch convert any given video files to virtually any QT-compatible format, allows you to enter descriptive metadata, and can even automatically upload the final product to a remote server via FTP, WebDav, etc. Everything is done from the same app, and it’s incredibly easy to set-up and automate the process. It’s also fully compatible with my old friend, the multiple-award-winning Elgato Turbo.264 hardware encoder. In fact, their batch processing API seems heavily indebted to Elgato’s minimalist design. I think even the most technophobic of faculty members would be able to master this program.

Metadata and Movies

Per the request of a faculty member, we have been looking into ways in which one might be able to embed metadata into a given movie, and view that same metadata without having to have an extra program, or opening Quicktime Pro. The benefit of such a process would be to be able to view metadata as a plain text file, and further, to perform batch edits of the information, similar to what is regularly done with digital image files. The verdict is…no dice unless you are a developer with significant skills. It turns out that the metadata that one embeds using QT Pro’s interface of “Show Movie Properties” in stored in what developers refer to as an “opaque container” – this being a specific key value associated with Quicktime Pro – thus requiring anyone desiring to view such metadata in a plain text or other non-native QT format (i.e. us) to create some type of program that would find each of these opaque containers, the key values associated with them, and then to export the said values out to the desired format. All in all far beyond my abilities – so any takers?

This link has more on the way that the metadata works.

http://developer.apple.com/documentation/QuickTime/Conceptual/QT7UpdateGuide/Chapter02/chapter_2_section_9.html

But all was not for nothing. I found a very nice open-source program called Metadata Hootenanny that allows you to look at a number of files at once and to edit metadata fields in a simpler interface than QT Pro. The program is a tad buggy at times, but the bugs only affect the appearance, which can be annoying, but does not present any issues with the way the program is supposed to work. The data you embed in the files is then exported and is able to be read in QT Pro. So, the program does nothing for the above issue, but is nonetheless helpful when entering data for large numbers of files (so that you do not have to open each one by one, and go through the pull down menus for say… 25 files).

Here is the link:

http://www.applesolutions.com/bantha/MH.html

Pages

I have recently been working on some help documentation for the new version of WordPress (to which I realize this site is not yet updated) – and have posted the pages to the ITG blog. You can find them right at the top of the blog sidebar. Now, being predominantly a video/audio specialist – I have little to no experience with HTML or any type of web publishing that doesn’t involve compressing videos or audio for upload. This is where all the net publishing gurus come in. Look at the pages I have created, and please either catch me at ITG during the mornings or message me via the wiki to help me out in my endeavor to create both helpful as well as visually appealing WordPress documentation. I have done the tedious work of screen captures, Photoshop editing, and writing out the steps. Now, I just need a little help from my friends (in the words of John, Paul, George, Ringo, and later…Joe Cocker).