1
00:00:00,400 --> 00:00:02,680
This is an ADLAB PRO video.

2
00:00:02,760 --> 00:00:06,080
ADLAB PRO is a project funded
by the European Union.

3
00:00:06,160 --> 00:00:10,000
This video is made available to you
on a Creative Commons licence.

4
00:00:10,160 --> 00:00:13,520
Hello. I’m Anna Matamala, from UAB,
and in this ADLAB PRO video

5
00:00:13,600 --> 00:00:16,360
I will explain
what audio subtitling is.

6
00:00:16,440 --> 00:00:20,360
This is Unit 1 "Audio subtitling"
in Module 5 "Additional services".

7
00:00:20,520 --> 00:00:23,440
Imagine you are a Norwegian speaker
and go to the movies.

8
00:00:23,520 --> 00:00:26,200
The film is in Japanese
with subtitles in Norwegian.

9
00:00:26,320 --> 00:00:29,680
You cannot understand Japanese
and you cannot read the subtitles:

10
00:00:29,760 --> 00:00:33,440
you are blind, you have dyslexia
or the subtitling speed is too fast.

11
00:00:33,520 --> 00:00:35,800
Audio subtitling is the solution.

12
00:00:35,880 --> 00:00:38,560
Subtitles are one type
of text on screen.

13
00:00:38,680 --> 00:00:41,680
Other types are logos,
film credits,

14
00:00:41,760 --> 00:00:45,280
superimposed titles,
popping-up text messages, etc.,

15
00:00:45,360 --> 00:00:47,400
but we won't be dealing
with those.

16
00:00:47,520 --> 00:00:50,120
In some countries
subtitling is the standard way

17
00:00:50,200 --> 00:00:53,200
of translating audiovisual content.
In other countries,

18
00:00:53,280 --> 00:00:56,160
what we call dubbing countries
or voice-over countries,

19
00:00:56,240 --> 00:00:58,560
subtitles are also used,
but a bit less.

20
00:00:58,720 --> 00:01:02,320
Audio subtitles is the rendering
of written subtitles into audio.

21
00:01:02,440 --> 00:01:07,240
It’s the subtitles read aloud. Terms
such as “audio subtitles/captions”,

22
00:01:07,320 --> 00:01:12,080
“spoken subtitles/captions”,
or even “audio text” are also used.

23
00:01:12,160 --> 00:01:15,960
Here I will use “audio subtitling”
and “audio subtitles”

24
00:01:16,040 --> 00:01:20,040
and the abbreviation “AST” together
with “AD” for audio description.

25
00:01:20,240 --> 00:01:23,240
I will discuss
the relationship of AST with AD,

26
00:01:23,320 --> 00:01:26,280
its main features
in terms of creation and delivery,

27
00:01:26,360 --> 00:01:30,040
and the role of users.
AST can be an independent service

28
00:01:30,120 --> 00:01:33,200
or can be integrated with the AD.
One example: when you are watching

29
00:01:33,280 --> 00:01:36,360
live broadcasts or live streamed TV
on online TV channels

30
00:01:36,440 --> 00:01:38,800
on the Swedish
public service broadcaster

31
00:01:38,880 --> 00:01:41,920
you can get automatic
text-to-speech audio subtitles

32
00:01:42,000 --> 00:01:45,160
through your digital terrestrial
network. Another example:

33
00:01:45,240 --> 00:01:48,800
the multilingual film "Inglourious
Basterds", by Quentin Tarantino,

34
00:01:48,880 --> 00:01:50,840
is mainly in English
but has dialogues

35
00:01:50,920 --> 00:01:53,680
in French and Italian
subtitled into English.

36
00:01:53,760 --> 00:01:58,320
The AD describes the visuals but
it also incorporates an oral version

37
00:01:58,400 --> 00:02:02,000
of the written subtitles.
It is an AD with AST.

38
00:02:02,160 --> 00:02:05,240
In terms of delivery,
ASTs can be read aloud

39
00:02:05,320 --> 00:02:09,240
by a human voice or can be read
by a text-to-speech system.

40
00:02:09,320 --> 00:02:13,240
The voice should be clear and should
be easily understood by users.

41
00:02:13,320 --> 00:02:17,440
ASTs can be live or recorded.
Sometimes, when some preparation

42
00:02:17,520 --> 00:02:22,040
is possible for a live delivery,
we talk about semi-live ASTs.

43
00:02:22,240 --> 00:02:25,240
ASTs address many users:
for example,

44
00:02:25,320 --> 00:02:27,760
persons who cannot access
the visuals, who cannot

45
00:02:27,840 --> 00:02:31,800
or do not want to read subtitles,
persons with reading difficulties,

46
00:02:31,880 --> 00:02:34,480
persons with sight loss,
but also children,

47
00:02:34,560 --> 00:02:37,720
the elderly and language learners.
It is important that users

48
00:02:37,800 --> 00:02:40,640
are informed
about the availability of AST

49
00:02:40,720 --> 00:02:43,320
by including
an accessible notification.

50
00:02:43,440 --> 00:02:47,520
Users should also be involved in
the creation and evaluation of ASTs.

51
00:02:47,600 --> 00:02:51,880
The ISO standard on AST recommends
that users should be able to select

52
00:02:51,960 --> 00:02:54,520
different types
of access services independently

53
00:02:54,640 --> 00:02:57,920
or in combination with AD.
When delivering ASTs,

54
00:02:58,000 --> 00:03:00,400
synchronisation
and timing are critical.

55
00:03:00,480 --> 00:03:03,040
This can be difficult
when the text is very dense

56
00:03:03,120 --> 00:03:05,720
or when there are numbers
and abbreviations.

57
00:03:05,800 --> 00:03:09,600
In this case, you may need
to increase the narration speed

58
00:03:09,680 --> 00:03:13,040
or change the text of the subtitles,
make it shorter, rephrase it.

59
00:03:13,160 --> 00:03:15,480
There are two main types
of delivery effects:

60
00:03:15,560 --> 00:03:18,040
when the ASTs
in the target language are heard

61
00:03:18,120 --> 00:03:21,760
on top of the original dialogue,
a “voice-over effect” is achieved:

62
00:03:21,840 --> 00:03:23,960
you can still hear
the original dialogues.

63
00:03:24,040 --> 00:03:27,640
When the original is not heard
at all, we have a “dubbing effect”.

64
00:03:27,720 --> 00:03:30,760
This often involves some acting
by the voice talents.

65
00:03:30,920 --> 00:03:33,440
When audio subtitles
are delivered together with an AD,

66
00:03:33,520 --> 00:03:37,240
it is important to discern them.
Various strategies can be used:

67
00:03:37,320 --> 00:03:40,160
different voices: for example,
a female voice for the AD

68
00:03:40,240 --> 00:03:43,520
and a male voice for the AST,
prosodic features, that is,

69
00:03:43,640 --> 00:03:46,720
changing the intonation,
or descriptive words,

70
00:03:46,800 --> 00:03:49,040
such as, for example,
“Subtitles read:”.

71
00:03:49,160 --> 00:03:51,960
Similar strategies can be used
when you want to identify

72
00:03:52,040 --> 00:03:55,400
the character speaking: different
voices, prosodic features...

73
00:03:55,480 --> 00:03:59,400
or you can use the name of the character
or use the reported speech.

74
00:03:59,600 --> 00:04:03,560
If you want to know more,
I recommend four resources:

75
00:04:03,640 --> 00:04:06,880
The ADLAB guidelines.
Good advice on how to create AST

76
00:04:06,960 --> 00:04:09,080
with a step-by-step
decision process.

77
00:04:09,160 --> 00:04:12,480
The HBB4ALL guidelines,
with a specific document on AST.

78
00:04:12,600 --> 00:04:16,120
The article “Audio description with
audio subtitling”, Braun and Orero.

79
00:04:16,200 --> 00:04:18,440
The ISO technical specification

80
00:04:18,520 --> 00:04:23,080
20071-25:2017. A must-read!

81
00:04:26,280 --> 00:04:28,440
The preparation
of this presentation

82
00:04:28,520 --> 00:04:32,120
was supported by ADLAB PRO.
Audio Description:

83
00:04:32,200 --> 00:04:35,840
A Laboratory for the Development
of a New Professional Profile

84
00:04:35,960 --> 00:04:40,240
financed by the European Union
under the Erasmus+ Programme,

85
00:04:40,320 --> 00:04:43,600
Key Action 2,
Strategic Partnerships.

86
00:04:43,680 --> 00:04:44,880
Project number:

87
00:04:44,960 --> 00:04:53,320
2016-1-IT02-KA203-024311.

88
00:04:53,880 --> 00:04:57,200
The information and views
set out in this presentation

89
00:04:57,280 --> 00:05:00,320
are those of the authors
and do not necessarily reflect

90
00:05:00,400 --> 00:05:03,480
the official opinion
of the European Union.

91
00:05:03,560 --> 00:05:06,920
Neither the European Union
institutions and bodies

92
00:05:07,000 --> 00:05:09,520
nor any person acting
on their behalf

93
00:05:09,600 --> 00:05:11,440
may be held responsible
for the use

94
00:05:11,520 --> 00:05:14,160
which may be made of
the information contained therein.

95
00:05:14,360 --> 00:05:16,360
This was an ADLAB PRO video.

96
00:05:16,440 --> 00:05:20,080
ADLAB PRO partners are:
University of Trieste,

97
00:05:20,160 --> 00:05:23,160
Universitat Autònoma
de Barcelona,

98
00:05:23,240 --> 00:05:27,240
University of Antwerp,
Adam Mickiewicz University,

99
00:05:27,320 --> 00:05:30,320
RTV Slovenia,
Utopian Voices,

100
00:05:30,400 --> 00:05:34,600
Soundfocus and Royal National
Institute of Blind People.

101
00:05:34,680 --> 00:05:37,640
ADLAB PRO was funded
by the European Union.

102
00:05:37,720 --> 00:05:41,480
This video was made available to you
on a Creative Commons licence.