1 00:00:00,964 --> 00:00:07,189 (ELECTRONIC MUSIC) 2 00:00:17,970 --> 00:00:21,918 Hello. My name is Carme Mangiron. I am from Universitat Autònoma. 3 00:00:22,078 --> 00:00:24,473 I have curly, brown hair, brown eyes, 4 00:00:24,633 --> 00:00:28,861 and I am wearing a black blouse with white flowers and red petals. 5 00:00:29,021 --> 00:00:32,911 I am standing in front of a while wall in my Barcelona apartment. 6 00:00:33,071 --> 00:00:36,374 Together with Anna Matamala, I am one of the leaders of 7 00:00:36,519 --> 00:00:38,875 the Researching Audio Description Project. 8 00:00:39,242 --> 00:00:43,136 It is an interdisciplinary project that lasts three years 9 00:00:43,296 --> 00:00:46,090 and is funded by the Spanish Ministry of Science, 10 00:00:46,250 --> 00:00:48,333 Innovation and Universities. 11 00:00:49,076 --> 00:00:50,842 The objective of the project 12 00:00:51,002 --> 00:00:54,085 is to bring forward the research on audio description 13 00:00:54,320 --> 00:00:57,929 by focusing on four underexplored areas. 14 00:00:58,089 --> 00:01:03,372 The first one is the delivery of AD, and specially focusing on prosody. 15 00:01:03,728 --> 00:01:07,243 The second one is about AD translation, 16 00:01:07,403 --> 00:01:10,581 and we'll be looking at less-studied languages, 17 00:01:10,741 --> 00:01:12,455 such as Spanish and Chinese. 18 00:01:13,418 --> 00:01:15,146 Then we have two case studies. 19 00:01:15,306 --> 00:01:17,796 The first case study is video games, 20 00:01:17,956 --> 00:01:23,042 and we are going to be looking at how AD could improve game accessibility 21 00:01:23,202 --> 00:01:25,234 for blind and low vision players. 22 00:01:25,394 --> 00:01:26,587 We'll talk with players, 23 00:01:26,747 --> 00:01:32,488 and try to find out what their needs, expectations and preferences are. 24 00:01:32,648 --> 00:01:34,081 Finally, 25 00:01:34,241 --> 00:01:37,057 the second case scenario is about opera AD, 26 00:01:37,217 --> 00:01:39,119 and again, we'll talk to users 27 00:01:39,279 --> 00:01:41,707 to find out their needs and preferences. 28 00:01:42,778 --> 00:01:45,855 We hope that the Researching Audio Description Project 29 00:01:46,015 --> 00:01:48,742 will contribute to improving accessibility 30 00:01:48,902 --> 00:01:52,251 for blind and low vision users. 31 00:01:52,411 --> 00:01:53,411 Thank you. 32 00:02:01,169 --> 00:02:03,931 Good morning. I am María Machuca, 33 00:02:04,091 --> 00:02:08,437 I have brown hair and I am wearing glasses 34 00:02:08,597 --> 00:02:14,179 and at this moment I am at home in my office. 35 00:02:14,339 --> 00:02:18,707 I am responsible of the prosody part 36 00:02:18,867 --> 00:02:22,976 in the Researching Audio Description Project, 37 00:02:23,136 --> 00:02:25,217 whose coordinators are 38 00:02:25,377 --> 00:02:28,673 Carme Mangiron and Anna Matamala 39 00:02:28,888 --> 00:02:34,436 and I want to show you how we annotate the audio signal 40 00:02:34,596 --> 00:02:38,455 and how we extract the acoustic parameters. 41 00:02:38,615 --> 00:02:42,723 The films were transcribed with ELAN. 42 00:02:42,883 --> 00:02:48,328 ELAN is an annotation tool for video and audio recordings. 43 00:02:48,488 --> 00:02:52,901 Then when the recordings are annotated 44 00:02:53,061 --> 00:02:57,784 we extract the transcriptions and we open at the same time 45 00:02:57,944 --> 00:03:01,207 the signal and the transcription in Praat software. 46 00:03:01,367 --> 00:03:03,416 This is the Praat software. 47 00:03:03,576 --> 00:03:07,279 It allows us to divide audio description units 48 00:03:07,439 --> 00:03:08,774 in intonation groups. 49 00:03:08,934 --> 00:03:10,735 Here you are a sample. 50 00:03:15,951 --> 00:03:18,228 At the bottom, you can observe 51 00:03:18,388 --> 00:03:21,643 how the audio description units can be divided 52 00:03:21,803 --> 00:03:23,504 into different intonation groups. 53 00:03:23,664 --> 00:03:29,116 We can label the prosodic features that separate the prosodic boundary, 54 00:03:29,276 --> 00:03:32,040 for example, pitch, in this case, 55 00:03:32,200 --> 00:03:35,069 or pause, in this case. 56 00:03:35,229 --> 00:03:39,570 The blue line indicates the evolution of the speaker's pitch 57 00:03:39,730 --> 00:03:42,974 when he pronounces this audio description unit. 58 00:03:43,134 --> 00:03:46,929 Here the prosodic boundary is the pitch, 59 00:03:47,089 --> 00:03:50,246 so we label it as pitch. 60 00:03:54,827 --> 00:03:57,123 In this group of intonation, 61 00:03:58,457 --> 00:04:03,853 the pitch also drops and then rises, 62 00:04:04,013 --> 00:04:07,568 so we label it as pitch too, 63 00:04:13,190 --> 00:04:16,653 but here, in this other group, 64 00:04:16,813 --> 00:04:17,879 there is nothing, 65 00:04:18,039 --> 00:04:21,296 the speaker has realized a pause, 66 00:04:21,456 --> 00:04:26,721 here it’s a pause, so we label it as a pause. 67 00:04:30,235 --> 00:04:33,183 When we have finished this process, 68 00:04:33,343 --> 00:04:36,202 we know the prosodic features that divide 69 00:04:36,362 --> 00:04:38,925 the intonation groups for the speaker 70 00:04:39,085 --> 00:04:42,065 and, also, what duration, 71 00:04:42,225 --> 00:04:44,646 pitch and intensity values 72 00:04:44,806 --> 00:04:47,218 we have for each intonation group. 73 00:04:47,378 --> 00:04:48,913 In this way, 74 00:04:49,073 --> 00:04:52,824 we describe the voice of this speaker.