The work process began before the Conference. It began by preparing mobile equipment for the trip -- digital camera, laptop, CompactFlash reader, supporting software on CDs, manuals, phone numbers, cables, spare FlashCards, batteries, chargers. . . . The usual suspects! It also involved looking through the agenda and my photo archives from past years (digital photos on CDs) to identify what photos I might already have, and what stuff I needed to get this time.The goal at this point was simply to have all the equipment I needed to capture material for a post-Conference website. However, the same elements, captured in the same resolution, would have been useable for print products or for this CD-ROM.
On Site at Alexis Park:
I always travel the day before the Conference, and with airline delays being what that are, I try to travel as early as possible, going west, to avoid those "delayed" flights. I arrived in Las Vegas in time shoot photos of the room setup process at the Alexis Park. Once I could see that Stuart's wireless network was functional, I bought an 802.11 adapter card for my laptop, and got it running. I found a large road case in a rear corner of the ballroom, and set up "shop". I now had a networked laptop, an adapter to dump photos from my camera to the laptop, and AC power for both the laptop and camera battery chargers -- essentially, a remote production studio.I was at the Conference, as was everyone else, to absorb new knowledge, but I was also there to collect the visual elements I might need for any post-Conference documentation. With a simple structure, this was not hard to do. I simply planned to
- find an easy-to-reach location near the podium, and shoot 3-4 digital photos of every presenter, sometime during their presentation. This turned out to be at the front of the center isle, just behind a TV monitor.
- try to get at least one wide-shot of the room, with the projection displays, to provide a content reference, if I forgot who the presenter was. (** the digital camera provides a sequential serial number to each frame, so I could always reconstruct the chronology of the photos, and link it to the agenda)
- get shots of the room showing attendees
- when sessions split, with IT going to another room, catch the same style presenter and crowd-shots in that room.
This only took a couple of minutes per presentation, and, since it didn't interrupt my ability to listen to what was being said, I was able to keep up with most of the presentation content. I did miss most of the IT presentations, as my primary focus was the Engineering sessions.My laptop developed a . . . reliability issue. It would randomly reboot, apparently from overheating. It made it difficult to use the network facilities, and impossible to find certain directories that Stuart had built. It would be Sunday afternoon before we could connect, and collect, the PowerPoint presentations that were stored on the network, and identify those that had not been saved from the presenters' laptops.
Collecting photos from the social events was the next element. With a 128meg CompactFlash card, I can shoot about 300 photos without stopping . . . well, maybe for batteries, but I always have three-sets of the NiMH style, and, so far, I've not run out of power. Each night I dump all photos to the laptop (6-gbyte hard drive), erase the FlashCard, and charge all batteries.
Everything leaves Las Vegas on my laptop, and is dumped to one of my home machines, the first day back. It remains on my laptop, as a backup, until I've had time to burn a CD-R of all the raw information.
About a week after we returned from Las Vegas, I got a package from Ralph, containing the PhotoCDs and audio cassettes. I put everything away for another three-weeks, to concentrate on a major WHRO production project.
Initial Analysis and Construction:
I began -- each project begins -- with a mental "wide-angle view" of the event. From a distance, I may have seemed to be asleep, but I was not!! I needed to find a theme, a mental image of the event, that I could build a physical image around. I look through all the photos, look thru the PowerPoint presentations, and listen to some of the cassettes, as I copy them into another computer for editing. I'm looking for a pattern. I'm looking for the essence of the Conference. I'm looking for some creative spark. . . .This one happened to be a combination of the audio from David Liroff's Rocky Road To Digital presentation playing into an adjacent computer, while I looked at a series of photos shot north of Las Vegas, Sunday afternoon, between the end of the Technology Conference and the Sony Dinner.
From the theme evolves one or more graphic interpretations. From a selected interpretation evolves a visual template for a webpage. From the template, and an analysis of the Conference content we'd been able to collect, evolved section -- or subject - templates. From the section templates I build each webpage, adding the text and images to support that element.
There may be several iterations of this discovery process, at least down to creating a couple of basic webpages, until I find the interpretation that seems the most appropriate for the project. This time there were only two iterations of the theme; this was the best interpretation.
The rest of the process is harder to describe, logically, because it tends to be an organic evolution of individual pages, that link to one another. It is, again, an iterative process, where a page is created expressing a specific element of the conference, and from it springs the need to express or explain some related thought on a different page. Twenty ot thirty pages later, a web of information exists that may sketch the whole subject, and which usually suggest the need for additional pages, or hyperlinks within existing existing pages, to make navigating between subjects and streams-of-thought more . . . . natural . . . intuitive.
Assembly and Tweaking:
Now parallel to this page-generation process, I've been going through all the Conference materials brought back from Las Vegas. I have edited and converted all the audio files, and sent them off -- also on CD-R -- to PBS for review and approval. I have identified which presentations we do not have, and contacted the authors, for whatever content they wish to contribute -- usually it is a copy of the original PowerPoint presentation; sometimes it is a PDF version. I have converted the PowerPoint presentations I do have to Acrobat PDF format, and looked that the resulting files. I've culled through all the photos that I shot, and those provided by others, throwing away (erasing) the duplicates. (** not erasing the "original" images, but copies, in a specific directory structure, that represents the work-in-progress files) See the subject-specific descriptions below!! I've started a running inventory of the available material, the file sizes and formats, so that we can develop a priority-list of what should and should not go on the CD-ROM. I've copied every original file -- in large groups --off to CD-R so that if all my computers exploded at one time, there would still be copies of every original element, somewhere in my office.
Now begins the assembly process. Photos add illustration to the page-text; tables add structure to repetitive, or like-data forms. Indexes and tables-of-contents provide one-stop locator resources. How much space do we have left ?? What else can we add ?? If we're out of space, what else can we do to make the sound-files or the photos smaller, to "make" more space for something else.
Parallel to these activities, I'm brainstorming packaging ideas -- label and jewelbox elements. Label ideas go to Ralph for review, and back into the rotation for improvements. We started out with a shot of the Stratosphere as the primary element this year, but the sky-color clashed with the Conference logo, and the detail in the overall image made it difficult to read text, no matter the text-color. Plan "B" was a variation of one of the pre-conference design elements -- a posterized palm tree. The Conference logo and title were dropped onto a plain white background, and a grove of colorful (isn't everything in Las Vegas) palms placed on the opposite side, for balance. User text was dropped into the lower right, and it was time to print. These happen to be Memorex press-on labels, printed on an HP inkjet printer. I've got some CDs that are 3-years old, and these paper labels seem to hold up well, though others have had some drying and lifting of the paper, over time.
Now, whole products are generated -- meaning I burn a CD-R of everything I think is supposed to be on the disk, then I open the disk on another computer, and see if it works as I intended. Labels and packaging are tested. Even the shipping container is used to send a version to Ralph, to check how everything works "together".
This becomes an iterative process; build-a-ROM, find a mistake -- spelling sometimes, but more often an issue where a file name has magically developed a "capital" letter some where in its name (usually when copied from one machine to another), which is no longer compatible with the HTML name, so a link fails -- fix it, build-another-ROM.
If one version seems to be correct, I burn 4-5 additional copies, and pass them out to beta testers -- err, friends, who try to crash the disk on their machines. We went through three beta rounds, on both Wintel and Mac platforms, and believe that, technically, its a sound disk -- there may still be spelling errors !!!
The process never seems to end; I'm still adding links and pages -- this page, specifically -- when the disk is essentially ready for duplication. Within minutes of completing this page, I will burn the master disk, and from it replicate several hundred CD-R copies.
Were there more time, there would be more pages . . . . .
..
PowerPoint Presentations:
Microsoft Powerpoint seems to be the defacto standard for Conference presentations. Even in the non-Windows world, Star Office produces a presentation file that is compatible with PowerPoint. If the author and/or presenter provided his presentation in the PowerPoint format, it is included on this disk. The primary advantage of providing the presentation in its original format is that it often includes the authors original script or notes, particularly important if the sound file for that presentation is not available. Secondly, it is an easily accessed on-screen display format, and its possible to both view the PowerPoint and have an audio file playing in the background.
My first concern here, was the overall size of the collected presentation. The 640-megabytes available on a CD-ROM seems like an almost infinite space, until you start adding up 6-to-11-megabyte PowerPoint files, and unknown quantite of other materials. We were fortunate that, while a number of the files were graphic-intensive, multi-megabyte files, a number of others were primarily text, some under a megabyte.
A few presenters were concerned about the potential for mis-use of their original material (complex graphics, corporate logos, etc) in the easily-modified PowerPoint format, and chose not to provide material in that form. In each case, they did provide an Adobe Acrobat (PDF) version for distribution. We respect their choice to retain control their intellectual property, and are happy to have their presentation available, in any form !! In the final construction, we ended up with about 46-megabytes of PowerPoint presentation files, a about 7-percent of the available space.
Adobe Acrobat (PDF) Conversions:
Where Microsoft PowerPoint represents the defacto presentation standard, Adobe Acrobat is the default standard for universal document printing. There are likely more people who have access to Acrobat Reader for on-screen viewing of the presentations, than have Microsoft Office and PowerPoint. Providing presentations in PDF format also greatly simplifies the printing process, for those that simply must have dead-tree versions.PBS has commonly provided presentation handouts at the Conference in a three-images-per-page booklet form. We've provided PDF versions in that form, as well as the original, full-screen, single-slide-per-page form. The former is useful for most presentations, but the latter really helps on those presentations with complex graphs or charts !!!
A problem that developed in the conversion process may be related to Acrobat, but is more likely related to the evolution of PowerPoint, and the control features that have been removed over the last couple of years. The symptom is very long conversion times or documents that crash and never complete their conversion. There seems to be a association between how a PowerPoint presentation is assembled, and the problems associated with the conversion to PDF. Presentations that use multiple graphic elements, or layers of elements cause the most problems. It appears that Acrobat tries to render and save each element individually, as opposed to seeing it as part of the whole page. Older versions of PowerPoint had more export and printing options that eliminated similar conversion problems in past years, but I don't have Office95 any more, and the new versions lack the "controls". . . . The Linux presentation by J.R.Wessels was one that I could not convert, for hand-out form, and the slide-form version took over 20-minutes on the P-3/500mHz platform to produce a PDF file !!
There are a few presentations that are only available in this format, as the author chose not to provide the original PowerPoint presentation. In final form, the PDF files filled about 102-megabytes, or about 16-percent of the available space.
The Audio Presentations:
The contractor who provided the sound reinforcement system for the Technology Conference, also supplied a cassette recorder; not particularly planned, but a simple, no-cost add-on to the primary sound system for the Apollo ballroom. They - Ralph, Stu, Keith, and the other guys who really ran the conference technology, recorded the sessions, as in some past years, with no specific application or use in mind; they simply knew that if something "inspired" happened in one of the sessions, and they didn't have a recording . . . . In several of the sessions, particularly those that involved spontaneous panel discussions, this was indeed the case. No powerpoint graphic or printed handout could have captured the essence of the event as well as the audio recording. At least a few audio files would need to be included in any Conference collection, to convey that information. We began the CD-ROM process with about 18-hours of cassette tape recordings.
Working .WAVs:
From cassette-to-CD-audio became an interesting clash in technologies. First, the tape speed was wrong. I tried several playback decks, assuming, since there are so few cassette decks still working in my environment, that the problem was old or mis-adjusted equipment on my end. I finally found a 10-year-old Nakamichi MR-2, with variable speed rolled full-slow, began to approximate the voices I knew -- Ralph Schuetz's voice became my "test-signal". Next was the issue of content blocks -- the cassettes used for recording ran about 45 minutes on a side (closer to 50, with the speed-problems), and often ran out during a presentation, meaning a good bit of time would be needed to "join" the related sections in to cohesive "sessions". Because of the need to work from a linear original (a cassette tape), I estimate that the conversion and editing process probably consumed about 5-hours, for every hour of finished audio -- much more time than I expected. Once converted to .WAV format, its possible to perform some activities in non-linear and faster-than-real-time modes, but the speed, noise and equalization issues mean that there are a number of additional processing steps required for an analog tape original, than would have been required of a disk-recorded original.
I made a fateful decision to transfer the audio from analog, to the more drive-space efficient, 22kHz mono .WAV format. After converting 4-5 tapes, whole-side-to-single-file, on a 20gb drive, I began to work on editing those files into individual "sessions". Fast Edit is still one of the easiest WAV editors I've used, and it renders really fast on a 500meg platform!! I quickly noticed hiss and equalization issues that would be potential problems, as I later moved to MP3 compression. I just as quickly learned that my ProTools package wouldn't let me work on 22kHz-sampled files. . . . so I had to search for a "legacy" application that might. An old version of Dart Pro provided the tools I needed, and the flexibility to deal with 22kHz files.
The World of MP3:
I'm sure its obvious to all that there would be no way to include audio files in this package, if .WAV files were the only option. Even in mono format, two hours of audio will fill a conventional CD. We needed a compression method that could provide high-compression rates and reasonable quality, and use commonly available playback components. Streaming formats were considered, but when compared to MP3, there were no specific advantages for a CD-ROM based product, so we went with MP3.
It became an iterative process of editing, normalizing, equalizing, then tweaking final gains a second time, to bring the disparate recordings into some common volume range. Sometimes I'd convert a file to MP3, hear things that I didn't like, then go back to modify the WAV file again.
The conversion from .WAV to .MP3 format was accomplished using Media Jukebox, from WinGear, which uses the "GoGo-no-Coda" GPL encoder kernel.
Once all the files were edited, converted and supplied to PBS for review, we still needed to get the presenters' permission to use each on the CD-ROM. When an author or presenter declined, and a few did, we deleted that file, and re-calculated the available disk space for other content. In the final construction, we used 437-megabytes, or 68-percent of the available ROM-real-estate to provide about 16-hours of Conference audio.
The Photographs:
This is where I came in, and where, for me, this whole project began.I bought a Kodak 260 digital camera two years ago and discovered, after 25-years experience with 35mm SLR cameras, how much easier it was to shoot "electronically". I learned I could easily document equipment problems or construction issues, and email them to others involved -- improving communications and speeding problem resolution. I learned that I could turn them into web-pages, so that others could access them when they needed, not simply when I sent an email.
Those experiences evolved into the 2000 PBS Technology Conference website, which generated enough interest to evolve the 2000 Traffic and Operations Conference site, and the 2000 Fall Engineering "Online" Conference site, to support a single-camera, PBS satellite-fed video program. The 2001 Technology Conference site, and this CD-ROM, are the result of these "learning" experiences.
The Landscape:
I've been shooting photos in and around Las Vegas, during CES and NAB events, since 1994; first 35mm, and since 1999, with a Kodak digital camera. I have thousands of frames of stock material from past PBS Technology Conferences, but the goal is to still get-the-shot, for each and every new event and presentation. I really enjoy photography, so each shot is more of a personal challenge than . . . "work".My digital camera was still a workhorse-component for this project. I shot about 800 frames during the week in Las Vegas; about half of those during the three-days we were at the Alexis Park. David Ferraro, Director of Production at WHRO, shot (and donated) about 200 frames (2000x1500 pixels) using a Nikon 990 digital camera. Dave Voightritter, Network Engineer at WHRO, shot (and donated) about 100 frames on his digital camera, before his battery began to die . . . and he discovered that his charger (unique to that camera) was still in Virginia!!
For the Madame Tussaud's party, Ralph bought a dozen-or-so of the disposable 35mm cameras, and distributed them to random party guests. Those cameras were processed to produce both film-prints and Kodak Photo-CD components. The Photo-CDs were used for this project.
The Work:
All of the original digital images were at least 1500x1000 pixels in size. All were low-compression, typically 6:1-to-8:1, .JPG format, with file sizes between 300-kilobytes and 1.2-megabytes -- small by some standards, but uncomfortably large in the constraints of a single CD-ROM. I tried various compression ratios on several represenative images and found that even at 50:1 compression, there were minor visible artifacts, but much smaller files !!
The final image issue was one of rendering speed from the CD-ROM. Using HTML, it is simple to describe the display of an image in a specific size on the page, no matter the size of the original image. It is not so simple to have a page full of these images render quickly. . . its a major issue for web-distribution of images, and I found it a suprisingly difficult issue for the CD-ROM. In the end, it seemed easier (at least for the end-user) if I made image-proxies -- specifically-resized images to be the thumbnails on the full-page presentations -- than to accept the slow loading times of the larger images. It would mean two versions of every image, one large and one small, but it would . . . "work" better. In the final construction, there are 786 photo-files -- 383 separate photos. The take up about 49-megabytes of space, or about 7-percent of the available CD-ROM surface.
