Sin t

1
Fundació Barcelona Media Universitat Pompeu Fabra

How good Mobile TV needs to be?
2
Convergence is on the agenda to serve users on
the move
TV and telco friends or foes?
3
Specially in mobile content less is more
4
Helga Schmidt in BerlinThe "broadcast driven"
scenario

• The scenario illustrates a day in the life a of
  young urban woman who uses a pocketable TV set
  with an integrated handset, as a basic device for
  infotainment and to facilitate her urban
  lifestyle. The services are offered over a
  broadcast portal, mainly free-to-air access, on a
  public service basis.
• The services
• Reception of standard TV
• all-round regional news and information service
• re-purposing existing TV, radio, online, teletext
  content
• personalisation options, alerts
• links to relevant web pages
• pushed and pulled video and/or audio content
• interactive voting and forums
• saving broadcast content to terminal
•  The terminal is a light-weight pocketable TV set

5
John Doe in BritannyThe "telco driven" scenario

• The scenario illustrates the beginning of a
  working day of a white-collar commuter who
  organises his day when travelling to his job by
  train. In the morning and during his journey JD
  uses a dedicated state-of-the-art handset, to
  enjoy mobile TV channels and video sport events
  he has been alerted of. Infotainment paid
  services are the main components of John Doe
  scenario.
• Services
• Personalised Electronic Service Guide
• Pushed (group) alerts,
• including enriched alerts e.g. on-event
  one-click video access
• Mobile TV
• Mobile portal and P2P services access
• Global/group downloads
• Terminal is a TV enhanced handset

6
Maria Da Silva in Brazil"service driven" scenario

• The scenario illustrates a day in the life of a
  low-income single woman with a seven-year-old
  son. She is working and studying to improve her
  living conditions. During the day Maria uses
  alternatively a portable TV set and a mobile
  handset to access INSTINCT services, at home, in
  her neighborhood, or when commuting. Using a
  local connection she transfers dedicated contents
  received on her TV set to her handset, in order
  to use it later during her spare time, when she
  is on the move.
• Main services are
• Distance-education (e-classes) including
  selecting content and bookmarking items
• Access to portals, and selecting of content
• Some of it can be transferred to cell phone to be
  watched later
• Reception of alerts
• Participation in surveys and vote
• Sending / receiving E-mails
• E-commerce (E-banking, ordering of tickets, )
• Terminal(s) is portable TV set linked with
• and a mobile handset

7
Instinct Phases
8
Are these standards suitable for mobile TV
acceptability evaluation?

• The effect of using a particular sample with
  specific characteristics as we consider being the
  case for the user group for INSTINCT services.
• The duration of the videos and audio samples is
  10 second, it is not clear that 10 seconds video
  sequence is long enough to experience the types
  of degradations common to multimedia
  communication
• The effect of the context of usage of the system
• The standards do not contemplate the
  circumstances in which the quality of the video
  is quite poor
• These standards recommend a 5-point scale to
  gather user responses vs. a continue not
  intrusive response
• Need a more sophisticated statistical analysis to
  find threshold for acceptance

9
Psychophysics approach

• Introduced for the first time by McCarthy, Sasse
  and Miras (2004) ? they concluded that the rule
  high motion high frame rate does not apply to
  small screens for sport content.
• Their claim was that a new method to elicit
  continuous ratings of quality with minimal effort
  on the users part was needed.
• This method is based on gradually increasing and
  decreasing video quality within a single clip to
  identify the threshold level at which quality
  becomes acceptable or unacceptable to these
  users.
• -    - It is easy for users to understand
• -   - It is less disruptive to the user than
  other continues rating techniques (like slides)
• -   - It can be used with variable video quality
• - It is more relevant to service providers
  (since it will allow to test the variability that
  would be found in real networks)

10
Experimental design

• The clips users watched were 210 seconds in
  length and the quality was increased or decreased
  in discrete steps every 30 seconds. Users were
  not aware of this quality structure, the authors
  simply told them they would be watching films
  that varied in quality
• The variables were
• - Compression bit rate
• - high-compression bit rate MPEG4
• - low-compression bit rate MPEG2
• -  Video Content sport, news and video clips
• - Video Quality
• -MPEG4 sequences values 0.45-0.8-1.15-1.45 Mbps
• -MPEG2 sequences 1.45-1.15-0.8-0.45 Mbps
• This binary acceptability ratings (e.g. 1
  acceptable, 0 unacceptable) were transformed to
  a ratio measure by calculating the proportion of
  time during each 30 second period that quality
  was rated as acceptable

11
Experimental design
12
User sample (32 participants)
13
Analysis

• We performed a two repeated measures ANOVA with
  the following variables
• ANOVA among high-compression bit rate conditions
• Gradient of quality, with two values, increasing
  and decreasing quality.
• Contents, with three values, Clip A (music), Clip
  B (football), and Clip C (news)
• Quality, with 4 levels 0.45-0.8-1.15-1.45 Mbps
• ANOVA among low-compression bit rate conditions
• Gradient of quality, with two values, increasing
  and decreasing quality.
• Contents, with three values, Clip A (music), Clip
  B (football), and Clip C (news)
• Quality, with 4 levels 0.45-0.8-1.15-1.45 Mbps

14
Summary of the data

• With mobile TV
• We found that it is not possible to have a one
  size fits all for the quality of video in which
  content should be presented to users for mobile
  TV services, because of dependencies on the
  content.
• Special attention should be paid to the
  transmission of news content, and specifically to
  the transmission of the text part of the video
  were the main problem with this video seems to
  be.
• We found that users do not like the content when
  it was just re-purpose for a small screen.

15
INSTINCT Participants
16
Back-up slides
17
Escenario ilustrado
John switches to TV programme and enjoys the news
with multimedia enrichment an URL linking to
additional Web pages related to the ongoing news.
18
The convergence experience INSTINCT project
19
Demo at IBC cross services
20
Users in the driving seat

• At present, content creation is focus on the
  technologies that deliver the content
• Problems with this approach
• Do not understand users experience with the
  content
• Accessibility problems
• Cognitive accessibility
• Mismatching between content and context
• Better quality does not always correspond with
  user acceptance
• Example of mobile scalable video content study

Alternative User Centered Content Creation