| Close |
| Comment |
The Scientific American Frontiers ETV
Experience
Local Data Enhancements
Greg Daigle - July 2001
This overview highlights a spring 2001 trial for inserting local data
enhancements into national digital television broadcasts. The challenge to the
local production team was to create interactions based upon information from
local sources and insert those interactions seamlessly between nationally
produced enhancements accompanying the broadcast program. This overview is for
television producers who will be working with interactive producers to create
enhanced television experiences through digital television in accordance with
the Advanced Television Enhancement Forum (ATVEF) specification.
Background
ATVEF provides two means for the delivery of interactive content for an
associated television program. In one variation known as Transport A, content
"triggers" arriving in the broadcast instruct the receiving set-top
box to pull assets (animations, graphics, text, interactive elements, etc.)
from the Web via an Internet connection. This trial instead utilized ATVEF
Transport B to send the entire set of assets over the broadcast stream. No Web
connection was required.
The subject of this trial was the PBS program "Scientific American
Frontiers", produced by Chedd-Angier. It took place during the
broadcast of four episodes: "The Bionic Body", "Chimps R
Us", "Flying Free" and "Fat and Happy". The
Scientific American Frontiers (SAF) enhancement effort involved two production
levels: the national enhancement team from Chedd-Angier and seven local
enhancement teams at various PBS stations.
The Minneapolis/St. Paul (TPT -Twin Cities Public Television) team included
a Technical Director, an Executive Producer, and a development team from
interactive agency SixtyFootSpider. The development team consisted of a
Producer (team leader, writer), a Programmer (scripting, testing) and a
Designer (visual, interactive asset creation). Quality assurance and testing
was performed by a Webmaster.
Figure 1
Establishing the User Experience
The national enhancement team had responsibility for establishing an
overall "look and feel" for the user experience. This included a
template for the screen layout and the underlying HTML coding. The template
took into consideration factors such as NTSC-safe colors and reduced screen
resolution. Its design took into account an aesthetic selection of colors
established to coordinate with each of the episodes (e.g. palette of
"jungle" greens for the primate episode and "sky" blues
for the episode on flight, etc.).
During viewing of the enhancements, the broadcast image was reduced to
roughly two-thirds screen size and relocated to the upper right of the screen
corner (see Figure 1). The remainder of the screen was reserved for the
display of enhancement assets and interactive elements. This maintained the
integrity of the viewing experience and kept assets and interactions from
obscuring the broadcast image.
Figure 2
The flexible use of the template was negotiated between the local team and
the national team. This gave the local team the ability to portray unique
local source information. For example, in the "Flying Free"
episode a range of animations from the Minnesota Museum of Science depicting
airflow were selectable by shape and angle. Also, a unique multiple choice
navigation scheme was developed for the "Chimps R Us" episode,
allowing viewers to participate in a quiz developed in concert with the
Minnesota-based Jane Goodall Institute's Center for Primate Studies (see
Figure 2).
Location of TPT's signature (lower right in the text area) was also
negotiated early in the process, as was the end credit for SixtyFootSpider.
Project Constraints
Several technical constraints made the development of enhancements
different than typical Web page development. The remote control device
provided with the set-top box hardware limited the users to
"tabbing" through selection options on the screen by depressing one
of four navigation directional arrows on a "thumb-stick". A blue
rectangular highlight on the screen indicated which element was selected.
Depressing the thumb-stick would "enter" the selection. This is
very different from a PC's point-and-click interface provided by a mouse
or touch-pad and was a key consideration in simplifying the onscreen
navigations (see Figure 3).
The available memory on the receiving platform (a Triveni Digital set-top
box) established a 1 megabyte limit for the assets to be displayed. Reception
of broadcasts on a PC-based datacasting receiver would have fewer memory
constraints but the hardware is less common than set-top boxes. Wavexpress, a
provider of PC-based datacasting receiver hardware, also participated in this
national trial but was not a platform available in TPT's trial reception
area.
Other multimedia elements such as sounds were not included in the
enhancements. Imbedded navigational and voiceover sound files were not
considered as they might conflict with the show's soundtrack.
Local Production
The local development team was charged with finding and creating local
enhancement experiences that would coordinate with the concurrent broadcast
experience. The local team was given final scripts. Scripts indicated in
their text the approximate insertion point of the enhancement trigger (no
timecode reference was available). A rough-cut of the episode was not
available for viewing in advance of developing the enhancements.
It was the local Producer's responsibility to develop the local
enhancement concept for each episode. Upon approval from TPT the Producer then
contacted potential source organizations for information and reviewed source
materials. Text and captions were developed in coordination with a content
editor. The Programmer was advised of interactive requirements and gave input
and recommendations to the Producer. Any new graphical schemes and assets were
developed and prepped by a Designer. When all assets were ready the Programmer
authored scripts in HTML and Javascript. The Webmaster reviewed scripts for
best practices and made QA changes.
Outcomes
Securing permissions from local organizations for use of material in a
single enhancement rather than multiple enhancements made the production time
longer. Building those relationships, developing understanding and trust with
local organizations took a large chunk of time relative to the outcome. Lack
of familiarity with enhanced television also required some time before the
organizations were willing to sign-off on use of materials.
The average production time to produce each enhancement was 2 to 3
production-days (including quality assurance testing). The production team
agreed that the level of difficulty was similar to that of a small Web site.
Most production slowdowns during the trial were not unlike issues
encountered during Web site development. They include the gathering of assets
(photographs, animations) and iterative reviews of editorial content with local
sources, TPT and Chedd-Angier. The same graphical procedures and coding
practices applicable to Web site design were easily adapted to this media.
Early in the development process the team was cautioned that an accurate
test of functionality and screen layout of text and design elements required
the use of a set-top box emulator installed on a PC. The emulator possessed
less functionality than a PC browser and lacked simple built-in navigational
controls such as "back" and "forward". This deviation
from typical Web development should be taken into account when establishing
production timetables.
Recommendations For Future Trials
The local enhancement team recommends some improvements to the process for
the next trial:
- A more detailed description of where enhancements are triggered in the
script (i.e. by timecode) and where the enhancements are "flushed"
from the set-top box memory.
- Establishing naming conventions and procedures early in the process to
track most recent versions of graphics, text and HTML code.
- Supply local enhancement teams with a consolidated "design style
guide" establishing all of the guidelines for fonts, graphics, color,
layout and navigation.
- A prioritization of style guide rules so that when a team needs to
"break the rules" they will know which rules have flex and which
are inviolate.
- As the local enhancement team was developing for a specific platform they
were not faced with the added difficulty of designing for multiple set-top box
platforms. The ideal of Create Once Play Everywhere (COPE) is still just that
... an ideal; just as it is in the Web world. Coordination with the
leading interactive television authoring tool developers would be a valuable
step toward that goal.
- As sound has become an important part of the Web, so too will sound play
some role in enhanced content. Usage may be minimal at first (navigation
sounds and confirmations of interactive selections) but some thought as to
their co-existence with the broadcast soundtrack is warranted.
Figure 3.
