1
00:00:07,360 --> 00:00:10,720
EASIT: Easy Access
for Social Inclusion Training

2
00:00:12,360 --> 00:00:14,440
This is unit 1, Media Accessibility;

3
00:00:14,680 --> 00:00:17,520
element 5,
Media accessibility services,

4
00:00:17,800 --> 00:00:20,080
video lecture "Audio subtitling".

5
00:00:20,400 --> 00:00:23,720
I am Anna Matamala, from
Universitat Autònoma de Barcelona.

6
00:00:24,200 --> 00:00:27,800
In this short lecture
I will be talking about audio subtitles.

7
00:00:28,240 --> 00:00:30,800
They are also called spoken subtitles.

8
00:00:31,400 --> 00:00:35,760
Audio subtitles are subtitles
converted into spoken words,

9
00:00:36,000 --> 00:00:38,120
converted into an audio format.

10
00:00:38,440 --> 00:00:40,120
Why would this be necessary?

11
00:00:40,480 --> 00:00:43,440
Well, imagine you are
watching a movie in Chinese

12
00:00:43,680 --> 00:00:45,400
subtitled into Catalan.

13
00:00:45,760 --> 00:00:49,040
You are a Catalan speaker
and don't understand Chinese.

14
00:00:49,360 --> 00:00:50,920
Subtitles can help you.

15
00:00:51,280 --> 00:00:54,880
But what if you are a slow reader
or a person with sight loss?

16
00:00:55,240 --> 00:00:58,800
In this case, you need the subtitles
to be read out loud.

17
00:00:59,600 --> 00:01:02,560
Audio subtitles can be
an independent access service

18
00:01:02,680 --> 00:01:05,560
or they can be integrated
in an audio description.

19
00:01:05,960 --> 00:01:09,160
If you are a slow reader,
you can already see the images:

20
00:01:09,440 --> 00:01:11,360
you only need audio subtitling.

21
00:01:11,720 --> 00:01:16,120
If you are a person with sight loss,
you will need both the audio subtitles

22
00:01:16,360 --> 00:01:20,280
and a description of the images,
that is, an audio description.

23
00:01:21,440 --> 00:01:24,440
Audio subtitling is
a lesser-known access service:

24
00:01:24,800 --> 00:01:27,640
there is less research
and there are fewer countries

25
00:01:27,800 --> 00:01:32,400
where audio subtitling is offered,
especially as an independent service.

26
00:01:32,760 --> 00:01:37,160
According to a survey
launched by EBU in 2016,

27
00:01:37,680 --> 00:01:41,080
eight television broadcasters
in Europe out of 36

28
00:01:41,560 --> 00:01:45,120
were offering audio subtitles
as an independent service:

29
00:01:45,440 --> 00:01:48,200
Finland, Sweden, Norway, Denmark,

30
00:01:48,480 --> 00:01:51,360
Estonia, Belgium,
the Netherlands and Spain,

31
00:01:51,600 --> 00:01:54,520
in this latter case
in Catalan television.

32
00:01:55,080 --> 00:01:57,040
The presence of audio subtitles

33
00:01:57,160 --> 00:01:59,800
integrated in audio descriptions
is wider,

34
00:02:00,240 --> 00:02:05,120
and you can find them in films
which are fully or partially subtitled.

35
00:02:06,280 --> 00:02:10,360
In a 2019 book on audiovisual
translation and accessibility,

36
00:02:10,720 --> 00:02:15,080
I classified audio subtitles
according to different parameters.

37
00:02:15,840 --> 00:02:19,720
Depending on the time in which
they are created and delivered,

38
00:02:20,080 --> 00:02:23,760
one can talk
about recorded or live subtitling,

39
00:02:24,280 --> 00:02:25,760
with or without planning.

40
00:02:26,600 --> 00:02:29,560
Depending on the creation process
and the voice,

41
00:02:29,960 --> 00:02:33,560
there are audio subtitles voiced
through a text-to-speech system

42
00:02:33,840 --> 00:02:36,120
and human-voiced subtitles.

43
00:02:36,680 --> 00:02:38,400
When a human voice is used,

44
00:02:38,720 --> 00:02:41,800
audio subtitles
can be read in an acted way

45
00:02:42,280 --> 00:02:44,320
or in a less emphatic way.

46
00:02:45,000 --> 00:02:47,440
These prosodic features
(the intonation)

47
00:02:47,800 --> 00:02:51,880
together with the synchronization
of the audio subtitles with the original

48
00:02:52,280 --> 00:02:56,720
give place to two effects
identified by Braun and Orero:

49
00:02:57,640 --> 00:02:58,920
a voice-over effect,

50
00:02:59,120 --> 00:03:03,360
in which the audio subtitles
overlap with the source content,

51
00:03:03,800 --> 00:03:07,000
and a dubbing effect,
in which the audio subtitles

52
00:03:07,120 --> 00:03:09,360
replace the original dialogues.

53
00:03:10,240 --> 00:03:12,320
Sometimes a single voice is used

54
00:03:12,480 --> 00:03:15,400
to read the audio subtitles
and the audio description.

55
00:03:15,600 --> 00:03:17,440
So one single voice for both.

56
00:03:17,840 --> 00:03:20,120
Sometimes different voices are used.

57
00:03:20,520 --> 00:03:22,760
Here you can find different situations.

58
00:03:23,080 --> 00:03:27,680
For example, a single male voice
could be used for all male characters

59
00:03:28,240 --> 00:03:31,240
and one female voice
for all female characters.

60
00:03:31,600 --> 00:03:34,560
Or you could have one voice
for the audio description

61
00:03:34,840 --> 00:03:37,800
and only one voice
for all the audio subtitles.

62
00:03:38,960 --> 00:03:42,280
Another categorization
is related to synchronization.

63
00:03:42,720 --> 00:03:44,600
Audio subtitles are synchronous

64
00:03:44,960 --> 00:03:48,440
when they are read at the same time
they are shown on screen.

65
00:03:48,920 --> 00:03:53,240
They are asynchronous
when they are read before or after.

66
00:03:54,440 --> 00:03:57,680
When audio subtitles are integrated
with an audio description,

67
00:03:57,920 --> 00:04:02,400
it is important that users can
identify the audio subtitles as such.

68
00:04:02,720 --> 00:04:05,000
This can be achieved
by different means:

69
00:04:05,320 --> 00:04:08,720
using different voices,
changing the intonation,

70
00:04:09,000 --> 00:04:12,880
or adding a label
before the audio subtitles are voiced.

71
00:04:13,120 --> 00:04:15,400
For instance, "Subtitles read:"

72
00:04:16,880 --> 00:04:20,760
It is also very important
that users know who is speaking.

73
00:04:21,040 --> 00:04:22,880
Again, this can be achieved by:

74
00:04:23,360 --> 00:04:25,960
using a different voice
for each character;

75
00:04:26,320 --> 00:04:30,320
adding an indication of who
is speaking before the subtitle is read

76
00:04:30,840 --> 00:04:32,120
(for instance, "Mary:");

77
00:04:32,520 --> 00:04:37,040
using reported speech,
or using a voice-over effect,

78
00:04:37,160 --> 00:04:38,800
in which, as I mentioned before,

79
00:04:39,040 --> 00:04:41,040
the original voices can be heard

80
00:04:41,360 --> 00:04:44,240
and hopefully
the characters can be identified.

81
00:04:45,240 --> 00:04:48,960
Synchronisation of audio subtitles
with the original dialogues

82
00:04:49,080 --> 00:04:51,240
is one of the most critical problems.

83
00:04:51,640 --> 00:04:57,320
Imagine you need to read out
a number like 1,234.

84
00:04:58,240 --> 00:05:00,040
It takes quite a long time,

85
00:05:00,240 --> 00:05:03,920
longer than reading it
on a written subtitle on screen.

86
00:05:04,480 --> 00:05:10,040
In this case, either you speak
a bit faster or you edit the subtitle.

87
00:05:10,840 --> 00:05:13,920
In fact, audio subtitles
can be read literally

88
00:05:14,040 --> 00:05:17,200
as they appear on screen
or they can be edited

89
00:05:17,560 --> 00:05:20,800
to make them shorter
or to add some oral features.

90
00:05:21,200 --> 00:05:24,600
You need to take into account
that an audio subtitle

91
00:05:24,960 --> 00:05:28,200
corresponds to spoken words
in a language A

92
00:05:28,840 --> 00:05:32,520
that are translated into written words
in a language B

93
00:05:33,040 --> 00:05:36,440
that are then transferred
into spoken words

94
00:05:36,560 --> 00:05:38,360
in the same language B.

95
00:05:39,200 --> 00:05:42,280
You can find recommendations
about audio subtitles

96
00:05:42,400 --> 00:05:44,600
in some audio description standards,

97
00:05:44,840 --> 00:05:48,320
but I would like to draw your attention
to an ISO standard

98
00:05:48,520 --> 00:05:50,560
which provides guidance

99
00:05:50,760 --> 00:05:53,160
on the audio presentation
of text in videos,

100
00:05:53,440 --> 00:05:57,360
including captions,
subtitles and other on-screen text.

101
00:05:57,720 --> 00:06:02,560
It is a technical specification that
not only deals with audio subtitles

102
00:06:02,800 --> 00:06:05,600
but with
other types of on-screen texts

103
00:06:05,720 --> 00:06:08,640
that need to be translated
into audio texts.

104
00:06:09,880 --> 00:06:12,440
This video lecture
has been prepared by Anna Matamala,

105
00:06:12,560 --> 00:06:14,200
from
Universitat Autònoma de Barcelona.

106
00:06:14,440 --> 00:06:17,600
You can reach me
at anna.matamala@uab.cat.

107
00:06:18,760 --> 00:06:21,040
The project EASIT
has received funding

108
00:06:21,160 --> 00:06:22,360
from the European Commission

109
00:06:22,560 --> 00:06:24,920
under the Erasmus+
Strategic Partnerships

110
00:06:25,040 --> 00:06:26,600
for Higher Education programme,

111
00:06:26,800 --> 00:06:33,880
grant agreement
2018-1-ES01-KA203-05275.

112
00:06:34,400 --> 00:06:37,600
The European Commission support
for the production of this publication

113
00:06:37,720 --> 00:06:40,080
does not constitute
an endorsement of the contents,

114
00:06:40,400 --> 00:06:42,640
which reflect
the views only of the authors,

115
00:06:42,960 --> 00:06:45,040
and the Commission
cannot be held responsible

116
00:06:45,200 --> 00:06:48,800
for any use which may be made
of the information contained therein.

117
00:06:49,440 --> 00:06:52,120
This work is licensed
under a Creative Commons

118
00:06:52,240 --> 00:06:55,800
Attribution-ShareAlike
4.0 International License.

119
00:06:56,480 --> 00:06:58,240
Partners of the EASIT project:

120
00:06:58,680 --> 00:07:00,880
Universtitat Autònoma de Barcelona,

121
00:07:01,160 --> 00:07:03,320
Università degli Studi di Trieste,

122
00:07:03,640 --> 00:07:05,160
Universidade de Vigo,

123
00:07:05,480 --> 00:07:07,520
Stiftung Universität Hildesheim,

124
00:07:07,720 --> 00:07:09,000
SDI München,

125
00:07:09,320 --> 00:07:10,560
Dyslexiförbundet,

126
00:07:10,840 --> 00:07:12,480
Radiotelevisija Slovenija,

127
00:07:12,680 --> 00:07:13,920
Zavod Risa.

128
00:07:21,000 --> 00:07:24,360
EASIT: Easy Access
for Social Inclusion Training