RSS feed

Audio description for live performances and audience participation

Elena Di Giovanni, University of Macerata, Italy


Access services aiming to make live performances accessible to persons with sensory impairments are more and more a priority and a widespread practice. If providing accessibility to theatre plays, operas and other forms of live entertainment requires a deep knowledge −and detailed consideration− of the diverse needs of these segments of the audience, they also open up avenues for audience expansion. Relying on a two-year experiment in making opera accessible for people with visual impairments, and on the feedback they provided before, during and after accessible performances (in 2015 and 2016), this article reflects on audience participation as a tool for empowerment, increased awareness, sharing, universality. It offers a detailed discussion of the methodology and results obtained from observation protocols, questionnaires and interviews with accessibility providers and receivers. With a theoretical framework informed by audiovisual translation studies, reception and audience studies, the article also focuses on the positive, reversed trend whereby accessibility for special audiences becomes an asset for all. It concludes with overall comments regarding the findings of the experiment.


Accessibility, audio description, theatre, live performances, audience participation, reception.


There is an established discourse in which 'audience' simply refers to the readers of, viewers of, listeners to, one or other media channel [...]. Beyond commonsense usage, there is much room for differences of meaning, misunderstandings, and theoretical conflicts. The problems surrounding the concept stem mainly from the fact that a single and simple word is being applied to an increasingly diverse and complex reality (McQuail 1997: 1).

1. Introduction

When trying to define the concept of ‘audience,’ from whatever perspective, the most appropriate adjective seems to be diverse. Even in apparently homogeneous settings, from clear-cut theoretical grounds, in well-defined contexts, audiences are hard to define essentially because they are dynamic, shifting as the media they consume. Within media studies, attempts at grasping the essence of audience(s) have been made considering their local vs global nature (Athique 2016), their ethnographically classifiable features (Schroeder et al. 2003), the effects media produce on them (Mc Quail 2010) and many more, all attempting to define their diversity from one standpoint.

For Abercrombie and Longhurst, the notion of audience is tied to that of performance: as the two scholars maintain, a performance involves a relationship between performer(s) and audience(s) “in which a liminal space, however slight, is opened up” (1998: 40). In this view audiences are, therefore, actively involved in a performance; they share a space with it and contribute to its process of signification. This concept of audience participation in assigning meaning and influencing choices will be central in this chapter and explored with reference to a series of experiments carried out in 2015 and 2016 in relation to access services to live opera performances.

Participation in relation to end users immediately brings to mind action research, which, as Gaventa and Cornwall explain, is collaborative by nature: “action research is participative research, and all participative research must be action” (2001: 74). In the rather recent but prolific and multifarious development of action research applications, participatory approaches have always been central; for Gaventa and Cornwall, for instance, participation is at the core of a triangle whose points are knowledge, consciousness and action. Participatory research approaches promote understanding and knowledge, they enhance awareness and foster action. With this framework in mind, and driven by the wish to consider audiences for live opera audio description as actively involved in the meaning-making process, as well as in the enhancement and further dissemination of audio description (AD) itself, my accessibility team embarked on a multifarious, two-year experiment at Macerata Opera Festival, in Italy1. Our2 aim was to frame and analyse the production, consumption and reception of AD and other access services, to gain a better knowledge of what needs to be improved and to help raise awareness both within and beyond the opera house walls. We hope to achieve all of the above by getting the audience involved, and by implicitly empowering it.

In this chapter, action and participation are central concepts, along with audience and its empowerment. Audience is here assumed as per Abercrombie and Longhurst's definition, as diffused: “in contemporary society, everyone becomes an audience all the time” (1998: 68). As a matter of fact, we consume media texts and entertainment as part of our daily routine, living as we do in a media-drenched society. Also, audiences are not clearly separable from media and entertainment providers, especially with regards to live events, at which both space and time are shared as well as processes of signification.

In the next sections, we shall first reflect on the nature of audio description and the shifting, expanding nature of its audience. We shall also see how an expansion of AD audience can enhance awareness and empowerment. Subsequently, we will focus on live entertainment in conjunction with the research so far carried out, mainly but not exclusively from an audiovisual translation studies perspective. Finally, we will report on all the major steps in the two-year experiment at Macerata Opera Festival, Italy.

2. Audio description and its audiences

As Anna Jankowska writes, “blind and partially sighted persons are the main beneficiaries of audio description” (2015: 50), where ‘main’ leaves the door open to other end users. Although one of the most important reasons behind the flourishing of audience-centered empirical research in AD is precisely the identifiable features of its end users, i.e. the blind and partially sighted (B&PS), audio description is an access service which can potentially benefit “blind and partially sighted as well as many others, especially within the realm of live performances. As for its primary beneficiaries, Louise Fryer points out that even the B&PS, whom AD is conceived for, are “as diverse as any sighted audience” (2016: 42), thus rightly laying emphasis on the need to become deeply acquainted with the primary end users in all their abilities and needs, and to always consider all the possible layers audiences are composed of. This is indeed an avenue for further research in AD but also other access services (i.e. the testing and documenting of differentiated comprehension and preferences based on the types and degree of visual impairment).

In any case, be it primary or secondary, audience diversity has to always remain central in the provision of access services; a valuable set of strategies to becoming acquainted with such diversity consists precisely in reaching out to audiences, testing comprehension, asking for opinions and, even more powerfully, engaging them in the very creation and revision of their services.

To date, empirical studies on AD from the perspective of the audience have been copious and of diverse nature. Over slightly more than ten years, even if only considering research originating within AVT, many scholars have been contributing to framing comprehension and reception from many angles, occasionally relying also on feedback provided by non-visually impaired individuals (see Kruger 2012; Vilaro et al. 2012). All in all, research carried out so far has focused on the B&PS, whereas recent years have witnessed the onset of a reversed trend, especially in relation to live events and their accessibility. Scholars such as Neves (2012; 2016), for instance, have explored AD as part of a wider network of access services aiming at inclusion, and tested its appreciation with a multitude of users, in a participatory, empowering way.

Within this approach, both AD and its audience come to be seen as part of a potentially universal and unified entertainment experience. Twelve years after the appearance of the ‘for-all’ approach to media accessibility3, we are now witnessing the establishment of a reversed ‘for-all’ trend, whereby services are created with all end users in mind, and those traditionally connected with the sensory impaired are expanded to all potential audiences. Not overlooking the risks of excessive hybridity and pervasiveness, this approach to research, advocacy and practice is here praised and largely shared, as we shall see in the following sections.

3. Live performances and access for the blind

As defined by the Merriam Webster's, a live performance involves “a presentation, (such as a play or concert) in which both the performers and audience are physically present” (Merriam Webster Dictionary). As stated earlier, live performances, be they related to theatre, opera or other art forms, are uniquely characterised by simultaneity, a common time and space shared by both performers and audiences.

Over the past ten years, soon after media accessibility powerfully entered the realm of audiovisual translation research in Europe, several scholars explored the topic of access services to live entertainment from different angles and with different aims in mind. To date, most of the studies available are related to theatre and opera accessibility, although in recent years museums and the enjoyment of art more in general have come increasingly to the fore (Szarkowska et al. 2016; Neves 2016), thus rendering audiovisual translation research more sophisticated and interdisciplinary (Di Giovanni forthcoming a).

If hybridity and strong interdisciplinarity can pose problems for researchers and their evaluation, indeed they have the merit of strengthening approaches to media accessibility and, even more importantly, of enhancing the practice behind the research. As a matter of fact, if translation studies is often considered as an applied discipline, this is all the more true of audiovisual translation and especially media accessibility. For the latter, the connection between the practice and any theoretical or methodological reflection is always central, making media accessibility research best defined as action research.

In relation to theatre and opera, most research published to date has focused on surtitling and audio description. Surtitles have been the object of scholarly investigation for over ten years, with several contributions by Marta Mateo (2007a, 2007b) and Lucile Desblache (2007) proving ground-breaking and paving the way for further research. In the following years, Judi Palmer (2013) offered her “surtitler’s perspective” along with Jacqueline Page’s (2013) “translator’s perspective”, in two chapters for the volume Music, Text and Translation edited by Helen Julia Minors.

In 2015, Estella Oncins added to the methodological and empirical reflection on surtitling. Captions, both for the deaf and hard of hearing and for the hearing audience, were the object of a 2015 project carried out by StageText, a UK-based provider of captions for theatres, and the University of Roehampton in London: dropping the term ‘surtitles’ to usher in ‘captions,’ the project aimed at evaluating comprehension and appreciation of intra- and inter-lingual titles provided on mobile devices as a complement to, or replacement for, regular surtitles projected on screens. The project, carried out by Fryer et al., yielded remarkable results both in terms of appreciation of the captions delivered through mobile devices (for hearing and hearing-impaired viewers) and of attention distribution between the stage and the device.

If titles for live events have generated quite a conspicuous amount of applied research, so has audio description, with systematic contributions since 2007. That year Greg York, a professional audio describer at the Royal Opera House in London published a descriptive − but also implicitly prescriptive − article on audio introductions for operas and ballets, quoted repeatedly since its publication both for being one of the very first to appear and for providing practical advice for the making and improving of audio description (AD) and audio introduction (AI). The same year, Puigdomènech, Matamala and Orero published another practice-based article aiming to illustrate and support the application of a protocol for the preparation of an audio description for the opera. Several other articles with a practical slant followed: Cabeza Caceres (2010) reported on his experience as audio describer at the Liceu opera house in Barcelona, whereas the year before Udo and Fels published two articles based on experiences in theatres (2009a) and with school students (2009b), although from a non-AVT perspective. The greatest merit of these two articles lies in their support of unconventional audio description, breaking away from objective rules and aiming to encompass the intentions of the creators of a performance. More crucially, they aimed to inscribe AD within universal design theory and practice. The latter point is particularly relevant for our study: Udo and Fels (2009a) observe that it is only in the description of live audio performances and events that universal design principles can be applied, thus making the entertaining experience a truly inclusive one.

In 2013, Fels and Whitfield, an engineer and a philosopher respectively, published an article along the lines defined by Udo and Fels some years earlier (see above). Keeping universal design at the core of their empirical study, Fels and Whitfield advocate for an alternative approach to the creation of AD “involving actors, scriptwriters, musicians and directors” (2013: 2). As they report, although the performance creative team had no previous experience of inclusive design and AD, they successfully managed to give life to AD which was then tested with the B&PS, especially in terms of comprehension and overall enjoyment. The results were positive: the end users widely appreciated this alternative approach, at all levels.

Since 2014, audio description for Macerata Opera Festival has been enriched with the direct voices of the creators of the performances: directors, costume and set designers are all involved in the making of AD and given prominence mainly in its introductory section. This is one of the reasons that led us to setting up the first experiment, in 2015: our participatory AD had to be tested for comprehension and appreciation with its primary end users and, after them, with other potential beneficiaries, i.e. the wider opera audience. To close the circle of participatory AD making and testing, a decision was made, in 2016, to elicit feedback from some of the strategic and technical figures behind its provision.

4. Evaluating participation and reception: Macerata Opera Festival

For Abercrombie and Longhurst, living as we do in a media-drenched society, performance is part of our everyday life, media of mass communications and entertainment providing us with “an important resource for everyday performance” (1998: 75). Thus, the notion of performance itself becomes diffused, like that of audience: boundaries are blurred and roles are shared in an audience-performance relation that is primarily characterised by engagement, participation, and interaction.

Central to our analysis of audience participation in AD for live opera is the notion of reception, i.e. how live performances are understood, appreciated and remembered through and with AD. For the purpose of this and other studies (see Di Giovanni forthcoming b), reception is assumed to be the way/s in which individuals and groups interact with media content, and the manner in which a text is interpreted, appreciated, remembered (Staiger 2005). What we will describe below is, therefore, a participatory experiment in making, revising, understanding and appreciating −or criticising − audio description for live operas, with the ultimate aim to inscribe AD and other access services within a universal design framework, highlighting the importance of inclusion of all agents involved, from creators to primary and secondary receivers.

With these concepts in mind, let us embark on the description of a two-year experiment carried out during the summers of 2015 and 2016. Since 2008, Macerata Opera Festival (MOF) has been offering Italian audio description for at least two of its operas, with a constant expansion of its accessibility programme over the following years. In 2015, the festival offered three audio described operas in Italian (Rigoletto, Cavalleria Rusticana, La Bohème), two thematic touch tours (settings and props, costumes), tactile materials and downloadable audio introductions in Italian. In our quest for feedback, especially with reference to the type/amount of information and description frequency, we asked blind patrons to fill out an accessible questionnaire which was emailed to them the morning after each audio described performance, with instructions to return it within 24 hours. Out of the 13 questionnaires that were returned, nine provided valuable data for analysis.

The following year, access services for the blind at Macerata Opera Festival were further expanded, to include an extra touch tour with musical instruments and package holidays including audio described opera performances, especially designed for the B&PS.

In the wake of the previous year's experiment in feedback collection, a decision was made to expand the project in 2016, aiming to monitor audience participation and reception from additional perspectives and to consider audience in a diffused, universal-design way. Therefore, we involved the B&PS in the creation, revision and reception evaluation for AD and other services, and expanded our questionnaires to the sighted viewers accompanying the B&PS, as well as some of the theatre employees involved in the provision of the services. The reasons behind the experiment expansion were manifold: first of all, data for downloads of recorded audio introductions, in 2015, yielded interesting figures: between July 27 and August 15 (two days after the end of the festival), 200 downloads were recorded, whereas 35 additional downloads occurred from 16 August to 1 October, well after the conclusion of the opera season. Given the numbers and the download localisation, especially during the festival, we have reasons to believe that many sighted individuals accessed this service (a few of them have provided feedback either via email or directly). Moreover, in the previous years, several non-blind individuals accompanying the B&PS had expressed interest in listening to the AD and some of them had afterwards informally reported their satisfaction. As for festival managers and operators, our decision to administer a questionnaire to them came upon realising that their knowledge of AD and its value was developed during their first experience of it at MOF, which saw them both as co-creators and receivers of this service.

4.1 MOF in 2015: questionnaires on audio description

Audio description for live performances requires a lengthy creation, due precisely to the live nature of the performance to be described, but also to the parallel genesis of both production and AD. In the case of operas, audio description is often scripted during rehearsals, thus having to cope with variations in singing, acting, sets and props positions. For these reasons, a great deal of revising is involved, often leading to many layers of change from the first draft to the last one. In contrast to AD for cinema and television, AD for live performances, and opera in particular, relies on detailed introductory sections, provided before the beginning of the performance and during the breaks, before each new act begins. These introductory sections allow describers to provide information about the history of the opera, its plot, its original creators, the production on stage and all its aspects, from set and costume design to the vision of the opera by its director.

Audio description at MOF has always been subject to change and improvement. The first years saw a progressive increase in the number of short descriptions during the performance, upon suggestions provided by the B&PS patrons. In 2014, a decision was made to experiment with participatory AD by giving voice to the creators of the productions on stage, namely the director, the set and costume designer. The following year, the accessibility team decided to ask the primary audience for AD to formally express their opinion about the changes implemented and the overall reception. To this end, a questionnaire was designed and distributed to the B&PS by email the morning after the audio described performances. The questionnaire was made accessible using Google format. It comprised 22 questions, of which three were demographic questions aimed to elicit information about age, type of visual impairment, and gender. Of the remaining 19 questions, four were closed and 15 were open.

Out of the nine respondents who provided usable data, seven were female and two male, their average age being 56 and ranging from 28 to 73. All participants were completely blind. 55.5% of them were blind from birth and the remaining 44.5% turned blind at a later stage.

The first set of content-related questions (seven) aimed to elicit information about comprehension and preferences in relation to the audio described performance the respondents had attended. The question “what AD elements do you remember best?” scored a large majority of mentions for the settings. Settings description were best remembered by seven out of nine respondents, two of them added that they remembered details such as the furniture structure and its arrangement in the flat set up for La Bohème. When asked: “What elements did you find particularly useful in the description?”, respondents once again referred mainly to the settings: seven out of nine highlighted the importance of describing settings, also referred to as “the context for the action” or “the places.” Moreover, four out of nine respondents focused on movements as essential in descriptions, referring to the characters' movements on stage as well as to settings relocation during the performance. Question 8 aimed at identifying possible redundancies in the AD by asking respondents what elements they found excessive/redundant: 100% of ‘no redundant elements’ was scored with this question, expressed with slightly different formulas. These replies provided important feedback in relation to the increase in the number and length of in-act descriptions from the previous years.

The next four questions focused on AD delivery and started by asking respondents what they thought of the delivery system used. Seven out of nine declared that they had appreciated it, whereas of the remaining three, one person stated she would like to be able to regulate the volume and another pointed to a slight overlap with the music in the opening scenes. When asked what system for AD delivery they deemed most appropriate for opera or theatre performances, most either replied “I don't know” or did not provide any reply, whereas the youngest participant, aged 28, pointed out that it would be useful to have a braille-printed libretto along with the live AD. Contrary to our expectations, none of the participants mentioned any other technology. This may be due to the lack of other experiences for some participants, but also to the overall appreciation of the system used at MOF. When asked explicitly what they thought of the use of smartphones and other portable devices for AD delivery, to our surprise the only enthusiastic reply came from the 28 year-old respondent, whereas all others either rejected the idea or expressed concern for the correct functioning of the service, the risk of batteries running low, etc. The next set of questions referred to the download option that had just been implemented on the MOF website. All participants expressed full appreciation for this additional opportunity and, while one of them stated it could be useful for B&PS persons only, the remaining eight declared that it could and should be for all, three of them adding that information about plot, settings, costumes and directors' intentions should be shared with all viewers.

One of the very last questions, aiming specifically to elicit response on the perceived presence of additional voices in the AD, was: “Do you think the AD only reflects the audio describer's voice or any other?” Two out of nine participants did not provide any reply, whereas the remaining seven referred to “the director” (3), “the director and costume designer” (2), “the creators of the opera” (1), and “all” (1).

4.2 MOF in 2016: a wider participation and reception experiment

In 2016, three operas featured in the MOF programme: Otello, Il Trovatore and Norma. All three were audio described, for an overall three evenings made accessible to the B&PS with touch tours before the audio described performances. Bookings for both performances and touch tours registered an increase from the previous year and a steady request for touch tour participation from non-blind audiences. On the whole, approximately 230 blind and non-blind individuals participated in the accessible programme.

Two months before the beginning of rehearsals and the festival itself, a large audience participation and reception experiment was designed. Its stages were as follows:

  • AD script revision and final editing with two blind persons: Observation and oral questionnaires.
  • Touch tours: observation protocol for the B&PS participants.
  • Audio described performances: reception questionnaire for the B&PS and the non-blind accompanying them.
  • Audio description awareness and appreciation: questionnaire with the service providers at MOF.

The experiment design was complex, with different strategies, tools and measures applied to what we conceived as a comprehensive study of the participation and reception of as many members of our diffused audience as possible. In its diversity, the experiment was unified by being strictly qualitative in nature. The reasons for an overall qualitative approach were: 1) Our wish to involve several groups of participants but in small samples; 2) The need to interact with participants at varying degrees (i.e. directly through oral questionnaires, indirectly - but simultaneously - through observation, and a posteriori through written questionnaires); 3) Our wish to carry out a detailed, subject-per-subject, analysis. As Jensen puts it, qualitative research allows for the study of “media in contexts of social action” (2012: 179), seeing media text creation and consumption as socially-embedded actions. Qualitative research methods often opt for an anthropological approach to research, media ethnography being steeped in qualitative traditions (see for instance Pink 2006), whereby the researcher him/herself is immersed in the experience at the core of the study, as it was this case.

The sub-sections below provide an account of each segment of the experiment and an overview of the results.

4.2.1 AD scripting and revision: observation and oral questionnaires

In the first section of our experiment, we aimed to reflect on the creative process of AD writing with two B&PS persons. As stated earlier, opera AD for new productions is scripted during rehearsals, with the support of the opera creators both to elicit their reading of the work on stage and to clarify aspects related to the settings, costumes, movements, with a view to better describe them. For MOF, ADs have for years been structured with long introductions of six to eight minutes which are provided before the beginning of the performance. One or more long sections (three to five minutes) is provided between the acts, whereas only short descriptions are given during the actual performances, with a duration varying approximately from two to ten seconds for each description. The introductory section offers details about the history of the opera, its plot, the overall production context, the set and costume design for the first act and the main movements which will be seen on stage. In providing these descriptions, we add the voices of the stage director, the costume and set designer by using direct or indirect quotes.

The two persons who were involved in the AD revision were a partially sighted, 45 year-old man and a completely blind 57 year-old woman. Both of them had attended audio described performances in the previous years. The revision session took place on July 25, 2016, from 3 to 6.30 p.m. Two audio describers were present, with the AD scripts for two different operas. A third member of the accessibility team was present. That person followed the session through the lens of an observation protocol designed in cooperation with a social psychologist from the University of Macerata. The session was also audio-recorded.

The session started with a general discussion of the structure of the AD. Both B&PS participants recommended changing the order of the blocks of information given in the introduction. They suggested moving the plot from the first to the last section of the introduction, closer to the beginning of the performance.

Besides general remarks of this kind, comments provided by the two B&PS persons addressed linguistic-semantic and structural issues. As for the first, we shall here mention only two main aspects which were discussed, namely the use of anaphoric and cataphoric references and the identification of semantic redundancies or gaps. Remarks on the use of anaphoric and cataphoric references to characters were not provided spontaneously by the two B&PS participants, but initially elicited by one of the two describers by means of examples from the AD scripts. The two participants were asked to state whether they found the examples clear or unclear, and in all but one instance they indicated that comprehension was ensured, even when anaphoric/cataphoric references appeared over two different descriptions with music or singing in between. On the semantic level, the two participants pointed to occasional gaps in the provision of information about costumes. As a matter of fact, costumes are described in detail in the introductory section of the AD, but only for the main characters. For all other characters only brief descriptions are provided during the performance, always considering time constraints and the actual relevance of the description.

On the structural level, besides initial remarks on the order of information in introduction, the B&PS participants spontaneously and repeatedly expressed appreciation for sentences adding a touch of interpretation to objective description. In particular, they appreciated the occasional references to face and body movements, but also, more specifically, the use of manner adverbs to classify actions. The session closed on advice from the two B&PS participants as to the delivery mode for the AD. On the whole, this session was very helpful for describers and inspired further action along these lines.

4.2.2 Touch tours: observation of B&PS participants

The second step in our audience participation project involved observing the B&PS during touch tours. In fact, this was the only opportunity to directly observe the behaviour of our primary audience while consuming one of the access services provided by MOF. We decided to concentrate on the first of the three touch tours only, which was certainly a new experience for all participants (some of them also attended the following two). The whole tour was video recorded. Having a fairly large group at our disposal and the opportunity to test real fruition, we designed a systematic observation protocol with the support of the social psychologist. We identified two behavioural measures, as defined in behavioural research and exemplified by Bordens and Abbott (2008: 139). In particular, we selected frequency and latency: frequency refers to the number of occurrences of a type of behaviour/action in a specific setting, whereas latency refers to the time required to perform a task. In order to evaluate frequency and latency, we defined the following tasks:

  1. a) Touching the tactile materials and 3-D model provided
  2. b) Asking for clarification on (a)
  3. c) Touching the theatre walls
  4. d) Asking for clarification on (c)
  5. e) Touching the main sets
  6. f) Asking for clarification on (e)
  7. g) Touching the props
  8. h) Asking for clarification on (g)
  9. i) Asking for additional support.

To evaluate latency, we hypothesised durations for each main task (a, c, e, g), considering the group size (16 B&PS in the 18 to 63 age range, plus twelve accompanying persons), the number of trained operators running the touch tours (3), the theatre size (approx. 120 metres long and 60 meters wide), the length of the stage (90 metres). Hypothesised latencies were: a) five to seven minutes, c) six to eight minutes, e) twelve to 14 minutes, g) five to seven minutes, allowing two-minute intervals for each task.

Below is a table with figures for frequency and latency, as calculated directly during the touch tours and further verified through the recording.

Main task




Overall: 16 pax, 100%
Tactile tables: 16
3D model: 14

Average: 8.1 minutes
(five to ten-minute range)


Overall: 13 pax, 81,25%
Theatre walls: 13
Tiers: 7

Average: 7.2 minutes
(six to eight-minute range)


Overall: 16 pax, 100%

Average: 14.2 minutes
(13 to 15-minute range)


Overall: 12 pax, 75%

Average: 4.5 minutes
(four to six-minute range)

Table 1. Frequently and latency for main tasks in touch tours.

The first task (tactile materials and 3-model) was performed by the participants in the main hall leading inside the open-air arena before commencing the tour. Over the years, the touch tour organisation has been fine-tuned in collaboration with the B&PS and with experts on museum accessibility for the visually impaired.4 One recurring suggestion was to provide tactile tables for the overall structure of the arena, the stage area and the settings of the production on stage, for the participants to touch before starting the tour. To this end, tactile tables are produced every year, and a 2-metre long 3D model of the Sferisterio arena is also made available to be touched. This task was performed by all blind participants, gathered together around a large table with the tactile material. Average latency shows that the time spent on this task is higher than hypothesised (on average, +8%), the reasons being that: 1) each participant had the opportunity to touch 4-5 tables and ask for clarification, 2) this task opened the touch tour and gave essential information for the following phases. Sub-task (b), related to (a), scored 75% frequency, latency being included in the average value for the main task.

Once inside the arena, with the main tour leader illustrating the history and building stages, task (c) (touching the theatre walls) started. Due to the size and structure of the arena, with semi-circular walls running all around the stage for over 120 meters and with tiers slightly above the circular walls, this task required several minutes, not so much for continuous touching but for walking through and reaching the stage. When walking, participants had the opportunity to touch the walls and the seats, and to move to the tier-level by means of four steps. Not all participants decided to walk strictly along the walls (the oldest preferred to remain in the central path), and even less (seven persons, i.e. 43.75% of the total) took the steps to explore the tiers. Latency, as can be seen in the table, was in line with our hypothesis, the two-minute range coinciding perfectly. For task (d), deriving from (c), frequency was 31.25% (five participants) and latency corresponds approximately to 25% of the overall latency for the main task.

Upon reaching the stage from its left edge, the group was led onto it via 4 steps. Task (e), involving touching the settings, started here. As the stage is 90 meters long, settings usually fail to cover the whole space and tend to concentrate on the central area (approximately 50 meters). Due to the size of the stage area and the settings themselves (covering 50 meters in length and 25 in height), average latency recorded for this task is slightly higher than foreseen: 14.2 minutes, whereas the initial hypothesis ranged from 12 to 14 minutes. Frequency reached 100%: all B&PS participants touched the settings, in virtually all of its parts and more than once (see figure 1 below for an example). As a sub-task, (f) concerned all participants (frequency 100%) and covered approximately 40% of the time spent on this task.

Moving towards the opposite end of the stage, participants approached the area in which most of the props were gathered, to perform task g) (touching the props). Frequency for this task was 75% (twelve persons), with the remaining 25% leaving the stage to move back to the starting point and main entrance. As for latency, 4.5 minutes were needed, on average, to complete the task, i.e. 25% less time than hypothesised. The main reason for this discrepancy is that when performing this task very little clarification was required (task h) covering 20% of the overall latency value). To conclude with task (i), 43.75% of the participants (seven out of 16) asked for additional support, mainly about the tactile tables and their relationship with the actual settings, the costumes to match the settings and props, and extra information about the festival organisation. On the whole, all participants maintained an exploratory approach throughout the tour and were satisfied at the end.

4.2.3 Audio description comprehension and reception: questionnaires with the B&PS

Among the many possible tools for enquiry into the reception of AD, questionnaires are the most frequently used. If not administered orally, as has occasionally been done (Di Giovanni 2014), questionnaires have to be made accessible, so that B&PS respondents can easily fill it in. For stage 3 in our large-scale experiment, we designed two post-AD questionnaires: one accessible questionnaire for the B&PS, and one standard-format questionnaire for those accompanying persons who had requested headsets to follow the opera with AD. Audio description was delivered live at the Sferisterio arena: an operator manually launched pre-recorded audio clips and delivered them to the B&PS's cabled seats with headsets. Each headset is connected to a double plug, so that the accompanying person or whoever is sitting next to the visually impaired person can listen to the audio description. When booking their seats, the B&PS are informed that they should be inside the theatre at least 20 minutes before the beginning of the performance, so as not to miss the introductory section. Free, single headsets were distributed by the festival staff and are simply left on the seats at the end of the performance.

In order to test AD comprehension and overall reception of the service by its primary audience, i.e. the B&PS, an accessible questionnaire was sent to those patrons who had attended at least two audio described performances in 2016. The questionnaire was sent by email the morning after the second and third audio described show, with a request to return it within 24 hours.

On the whole, ten duly filled in questionnaires were returned. Our B&PS respondents were seven female and three male, their age ranging from 19 to 73 years (average age: 52). Nine out of ten were completely blind, and out of these nine, seven were born blind. The questionnaire was divided into five sections: The first aimed to gather demographic information (age, gender, name, type of visual impairment), the second contained specific questions about the AD of Otello, the third about Il Trovatore, the fourth about Norma, and the fifth asked for general feedback on the appreciation of AD, i.e. the respondents’ opinions. Each opera-specific section comprised four questions, two open and two closed. For the sake of brevity, we will only report on replies about two questions for each opera.

For Otello, the first open question was, which of the elements that were mentioned in the AD introduction listeners remembered most. All ten respondents referred to the settings, four of them mentioning the costumes along with the settings. A few respondents added details, by stating for instance that the settings descriptions were “even reporting on the height of the elements on stage.” A respondent reported the description of the golden, winged lion that is part of the main setting, and another mentioned the extras' movements through the settings. The first closed question was about the structure of the AD and the presence of the director's voice: “How useful is the voice of director Paco Azorin?” All ten participants rated it as extremely important.

In the section about Il Trovatore, the second open question was about the symbolic meaning of two central elements of the settings, i.e. the 25-metre long tables which run through the stage, one on the left and one on the right. Seven out of ten respondents rightfully reported that they symbolise the past and the future, the present lying in between. Of the remaining three, two only mentioned the past and the future, whereas one replied “I don't remember.” On the whole, although this information was given before the beginning of the performance, 90% of the respondents remembered it correctly. The second closed question enquired about the amount of information provided in the AD, asking whether respondents thought it was sufficient or insufficient. Seven out of ten participants declared it was sufficient, whereas three wrote that they would have liked more information about the costumes, the movements, and even the settings. We would regard these three replies as praise, rather than as a critique of the AD structure: in the first open question all respondents declared that they remembered particularly well the descriptions of costumes, settings and their symbolic values. Therefore, apart from movements, their request for more may be taken as high praise rather than as a critique of gaps.

With reference to Norma, the first open question which enquired about the best-remembered elements provided in the introduction led, again, to a majority of references to the settings, recalled for their symbolic meaning which, according to four participants, wouldn't be grasped without the description. The settings for Norma were indeed highly abstract and symbolic, and the voice of the director and set designer, appreciated by all respondents (five in the 1-to-5 scale for nine out of ten respondents), provided the key to understanding them. In a future experiment, questions of this type should be directed at non-blind participants as well in order to check whether they appreciate the introduction to a similar extent. The second open question for Norma addressed changes in lights and colours during the performance, not an easy question to ask the B&PS. Apart from two participants who wrote “I don't remember,” all other respondents described variations in colors and light intensity: the game with lights, for this production, is highly meaningful and indeed prominent throughout the performance.

4.2.4 Audio description comprehension and reception: questionnaires with accompanying persons

Of all accompanying persons for the audio described performance of Otello, three asked spontaneously if they could have headsets to follow the AD. Four more asked after seeing their neighbours in the seating area reserved for the B&PS. After the performance, we managed to distribute questionnaires only to five accompanying persons who had followed the AD. Four of them returned them duly filled within 30 minutes from the end of the performance (three women and one man, from 32 to 73 years of age, for an average age of 54). Although the sample of respondents is clearly small, their replies provided important insights for further reflection as they confirmed that listeners who are not B&PS also find the AD useful. The questionnaire was deliberately concise, with three demographic questions (name, age, name of B&PS person accompanied), followed by four closed and two open questions. The first closed question asked if participants had followed an audio described opera performance for the first time, to which all 4 of them replied positively. The second question asked whether they had found the AD useful. All four responded that they have. Then, when asked to specify whether they deemed AD useful either for the B&PS or for the general audience, three out of four replied that it is useful for everybody. The fourth and final closed question asked to rate the coherence of descriptions with what was seen on stage (costumes, movements, settings, etc.). The average score was 4, with one 3 (given by the oldest of the four respondents), two 4 and a 5.

More significant replies were provided for the two open questions that followed. The first asked: “Is there any element that you would have liked to be described in more details?” to which three out of four respondents replied negatively, stating that descriptions were sufficient and satisfactory. The remaining respondent stated that the protagonists' movements on stage could have been further described. The second open question asked respondents to add any additional comment they deemed useful. One respondent suggested that not even a single note should be covered by the in-act descriptions, whereas the second recommended further information about the need to be inside the theatre well in advance, so as not to miss the most important part of the AD, i.e. the introduction. All in all, replies and comments proved useful and thought-provoking: a reply to the first and second open question respectively seems to be in contrast with one another, first requesting further descriptions for characters' movements, then requiring that no single musical note, let alone the voices, is covered when providing AD. These remarks point to the variety of opinions on AD and its structure, as well as to the challenges of catering to all viewers. On the whole, all respondents submitted their questionnaire confirming their enthusiasm about the AD and the touch tour and thanked us for being allowed to express their opinions.

4.2.5 Discovering, making, understanding audio description: A questionnaire behind the scenes

As the last stage in our experiment, we decided to go one step further and engage some of the supporters in the making of access services at MOF. Immediately after the touch tours and audio described performances, a short questionnaire was sent to 4 key figures for the provision of AD and touch tours at MOF. These were: The general manager of the opera festival (male, 45), the stage manager (male, 42), the staff manager (female, 31) and the main AD operator (female, 30). They received the questionnaire by email and were asked to return it within 24 hours. All 4 persons have a key role in making AD happen at Macerata Opera Festival. The General Manager approves the accessibility programme and its related budget. The stage manager takes care of all technical aspects pertaining to AD creation and delivery, as well as the organisation of touch tours. The Staff Manager trains and supervises all staff working during touch tours and audio described performances. And the AD operator works on AD preparation, recording, and live delivery. All four were involved with access services for the B&PS in relation to live performances at Macerata Opera Festival for the first time. Except for the AD operator, they had no prior knowledge of such services.

The questionnaire had a short demographic section. This was followed by five questions, four open and one closed (the last one). The first enquired about previous knowledge of access services for live events. As for the remaining four questions, replies are provided in the tables below and then discussed. All replies have been translated from Italian into English by the author of this article.

What is the role of AD and of the accessibility programme within MOF?


“Music is a universal language, which has to be universal for all. This is what we aim for, with our access programme. Accessibility is key for us in all senses. We feel like we are also audiences.”


“It is a great opportunity to make space for, and engage, many people who would not otherwise engage with opera.”


“It is increasingly important. More and more people get to know the services, there is more and more interest by all audiences, beyond the blind.”


“AD and live access are more and more central, they address and appeal to all, from children to ordinary viewers. We should aim at providing service for everyone.”

Table 2. Questionnaire for theatre operators: question 2.

Replies to question 2, reported above, highlight and expand a number of key issues for our experiment and for the overall study here presented: access services should be, and are de facto for all, music is a universal language and it has to reach out to universal audiences. Moreover, as can be seen in replies 2 and 3, the audience active engagement is reported, as a sign of growth and participation.

What would you do to change and improve the service, especially AD?


“We are aiming to improve the delivery of AD, as much as possible. We will keep experimenting over the years, as we have done so far in collaboration with the blind. We are thinking of personalised access, too, from individual devices.”


“We could expand the service and offer it every night, for every performance. I think this could be feasible, so that audiences can freely choose their own performances. We have to check costs, but I think it is feasible.”


“Over the years, many improvements have been made, with the advice of end users. My staff and I gather spontaneous feedback in the theatre, during and after performances, and always report to the service organisers. Talking to users is always so interesting!”


“We should aim to make AD available for all, by advertising free downloads of introductory sections, by providing it on many devices and in many formats.”

Table 3. Questionnaire for theatre operators: question 3.

When asked how services could be improved, with special reference to AD, 3 participants out of 4 (excluding the Staff Manager) pointed to possible enhancements of the service to better reach out to wider audiences. The Staff Manager, on the other hand, reflected on improvements carried out over the years, and on feedback spontaneously provided by participants. The live nature of opera performance encourages active participation, even in feedback provision. Such active participation becomes empowerment when feedback is received, understood and integrated into access services.

Do you know of other techniques and strategies used for accessibility to live events?




“Not for the sensory impaired. In my 24 years on a stage as a technician, I’ve seen many attempts at accessibility, most of them failed. I wonder if this was due to lack of engagement.”


“I know surtitles, we also provide them. In my opinion, the best service is touch tours. People love to actively engage, and seeing them in the tour is amazing for us.”


“I think AD is the most useful technique, and should be advertised with everyone. It really is access for all.”

Table 4. Questionnaire for theatre operators: question 4.

Among the replies to question 4, two seem to be particularly worth noting. The Stage Manager’s reply indirectly focuses on audience engagement as one of the main contributors for success in access service provision. Audience is here to be considered as diffused, including agents supporting the performance, i.e. the operators themselves. The Staff Manager also mentioned the participants’ active engagement, which brings along the engagement of service providers as well.

The fifth and final closed question asked to rate Macerata Opera Festival’s commitment to accessibility as either a) limited, b) fair, or c) excessive. All 4 participants replied that it was fair. Two out of four wrote that in most Italian theatres and opera houses, the commitment is definitely too limited.

5. Conclusion

Although qualitative and deliberately heterogeneous in nature, involving small groups of participants and the use of various techniques for feedback collection, this two-year experiment has engaged approximately 50 people, in an effort to encompass virtually all agents involved in the provision, consumption and reception of AD and related services. As a corollary to the detailed accounts provided in the previous section, it seems plausible to deduct that for both years results have been largely positive in terms of comprehension and appreciation, with room for improvement and an overall desire for more active participation. A diffused audience has thus highlighted the need, and wish, for ever-more (positively) diffused, participatory performances.

Moreover, as a further comment on the overall results, we can say that the access services offered at MOF and analysed for this experiment seem somewhat blurred: we have reason to think that when more access services are provided, their appreciation and the effect they produce are amalgamated, which is indeed a positive sign.

From the experimental researcher’s perspective, this two-year empirical endeavour has been steeped in action from beginning to end: it re-oriented practice in the provision of all services starting from AD, but it has also stirred a comprehensive methodological reflection which has been reported in this chapter and will be developed further in the future. Thus, far from being “armchair theorizing” (Heron and Reason 2001: 151), this experiment has started from action and is open ended, with a call for further action.

As feedback from this two-year empirical study is poured into a renewed and reviewed provision of access services for MOF, its influence is expanding to theatres and active researchers beyond the MOF experience, increasingly reaching out towards universal design theory and practice.

  • Abercrombie, Nicholas and Brian Longhurst (1998). Audiences. A Sociological Theory of Performance and Imagination. London/Thousand Oaks/New Delhi: Sage Publications.
  • Athique, Adrian (2016). Transnational Audiences. Media Reception on a Global Scale. Cambridge: Polity.
  • Bordens, Kenneth and Bruce Abbott (2008). Research Design and Methods: A Process Approach (7th edition). New York: Mc Graw Hill.
  • Cabeza i Càceres, Cristobal (2010). “Opera audio description at Barcelona’s Liceu Theatre.” Jorge Dìaz Cintas, Anna Matamala and Joselia Neves (eds) (2010).New Insights into Audiovisual Translation and Media Accessibility: Media for All. Amsterdam/New York: Rodopi, 227-237.
  • Desblache, Lucile (2007). “Music to my ears, but words to my eyes?” Aline Remael and Joselia Neves (eds) (2007). A Tool for Social Integration? Audiovisual Translation from Different Angles. Linguistica Antverpiensia 6. Antwerp: Hogeschool Antwerpen: 155-170.
  • Di Giovanni, Elena (2014). “Visual and narrative priorities of the blind and non-blind: Eye tracking and audio description.” Perspectives: Studies in Translatology 22(1), 136-153.
  • (forthcoming a). “Dubbing, perception and reception.” Elena Di Giovanni and Yves Gambier (eds) (forthcoming). Reception Studies and Audiovisual Translation. Amsterdam: John Benjamins.
  • (forthcoming b). “Audio description and reception-centred research.” Elena Di Giovanni and Yves Gambier (eds) (forthcoming). Reception Studies and Audiovisual Translation. Amsterdam: John Benjamins.
  • Fels, Deborah I. and Margot Whitfield (2013). “Inclusive design, audio description and diversity of theatre experiences.” The Design Journal 16(2), 219-238.
  • Fryer, Louise (2016). An Introduction to Audio description. London/New York: Routledge.
  • Gaventa, John and Andrea Cornwall (2001) “Power and knowledge.” Peter Reason and Hilary Bradbury (eds) (2001). Handbook of Action Research. L.A./London/New Delhi/Singapore: Sage Publications, 71-82.
  • Heron, John and Peter Reason (2001). “The practice of co-operative inquiry: Research ‘with’ rather than ‘on’ people.” Peter Reason and Hilary Bradbury (eds) (2001). Handbook of Action Research. L.A./London/New Delhi/Singapore: Sage Publications, 144-154.
  • Jankowska, Anna (2015). Translating Audio Description Scripts. Bern: Peter Lang.
  • Jensen, Klaus Bruhn (2012). A Handbook of Media and Communication Research. London/New York: Routledge.
  • Kruger, Jan Louis (2012). “Making meaning in AVT”. Perspectives: Studies in Translatology. 20(1), 67-86.
  • Mateo, Marta (2007a). “Reception, text and context in the study of opera surtitles.” Yves Gambier, Miriam Shlesinger and Radegundis Stolze (eds) (2007). Doubts and Directions in Translation studies. Amsterdam: John Benjamins, 169-182.
  • (2007b), “Surtitling today: New uses, attitudes and developments.” Aline Remael and Joselia Neves (eds) (2007). A Tool for Social Integration? Audiovisual Translation from Different Angles. Linguistica Antverpiensia 6. Antwerp: Hogeschool Antwerpen, 135-154.
  • Merriam Webster Dictionary, (consulted 25/10/2017)
  • McQuail, Denis (1997). Audience Analysis. Thousand Oaks/London/New Delhi: Sage Publications.
  • (2010). McQuail’s Mass Communication Theory (6th edition). L.A./London/New Delhi/Singapore/Washington D.C.: Sage Publications.
  • Neves, Josélia (2012). “Multi sensory approaches to (audio) describing the visual arts.” MonTI 4, 277-293.
  • (2016). “Enriched descriptive guides: A case for collaborative meaning-making in museums.” Cultus 9(2), 137-154.
  • Oncins, Estella (2015). “The tyranny of the tool: Surtitling live performances.” Perspectives: Studies in Translatology 23(1), 42-61.
  • Page, Jacqueline (2013). “Surtitling opera: A translator’s perspective.” Helen Julia Minors (ed.) (2013). Music, Text and Translation. London: Bloomsbury, 35-47.
  • Palmer, Judi (2013). “Surtitling opera: A surtitler’s perspective on making and breaking the rules.” Helen Julia Minors (ed.) (2013). Music, Text and Translation. London: Bloomsbury, 21-33.
  • Pink, Sarah (2006). The Future of Visual Anthropology. London/New York: Routledge.
  • Puigdomènech, Laura, Anna Matamala and Pilar Orero (2007). “The making of a protocol for opera audio description.” Luis Pegenaute et al. (eds) (2007).Actas del III congreso internacional de la asociación ibérica de estudios de traducción e interpretación. La traducción del futuro: mediación lingüístca y cultural en el siglo XXI. Barcelona: PPU, 381-392.
  • Schroeder, Kim et al. (2003). Researching Audiences. London: Arnold.
  • Staiger, Janet (2005). Media Reception Studies. NY: New York University Press.
  • Szarkowska, Agnieszka et al. (2016). “Open art: designing accessible content in a multimedia guide app for visitors with and without sensory impairments.” Anna Matamala and Pilar Orero (eds) (2016). Researching Audio Description. New Approaches. London: Palgrave Macmillan, 301-320.
  • Udo, John-Patrick and Deborah Fels (2009a). “Suit the action to the word, the word to the action:
An unconventional approach to describing Shakespeare’s Hamlet.” Ted Rogers School of Information Technology Management Publications and Research. Paper 16, (consulted 25/10/2017)
  • Udo, John-Patrick and Deborah Fels (2009b). “The development of a new theatrical tradition: Sighted students audio describe school play for a blind and low vision audience.” International Journal of Education and the Arts 10(20), 1-27.
  • (2010). “Universal design on stage: Live audio description for theatrical performances.” Perspectives: Studies in Translatology 18(3), 189-203.
  • Vilaro, Anna et al. (2012). “How sound is the Pear Tree story? Testing the effects of varying audio stimuli on visual attention distribution.” Perspectives: Studies in Translatology 20(1), 55-65.
  • York, Greg (2007). “Verdi made visible. Audio introduction for opera and ballet.”Jorge Díaz Cintas, Pilar Orero and Aline Remael (eds) (2007). Media for All. Accessibility in Audiovisual Translation. Amsterdam: Rodopi, 215-229.

Elena Di Giovanni is Associate Professor of English Translation at the University of Macerata, Italy. She has a Degree in Specialised Translation and a PhD in English and Audiovisual Translation. She has taught audiovisual translation and media accessibility for 20 years, throughout Italy as well as in Spain, the United Kingdom, France, Belgium, the U.S., Egypt and China. From 2008 to 2015, she has been Visiting Lecturer at the University of Roehampton, London, Masters' Degree in audiovisual translation. Since 2013, she lectures on cinema accessibility at the Venice International Film Festival, within the European Parliament-funded LUX Prize for cinema (28 Times Cinema). From 2014 to 2016, she was Guest Lecturer at Montclair State University, USA. Since November, 2016, she is president of ESIST, European association for the study of audiovisual translation ( She is coordinator of accessibility services at Macerata Opera Festival ( and Teatro Grande di Brescia (, with more than 100 events made accessible in the past few years. She has published extensively on translation and the media. Her publications are here: - content=publications.

Note 1:
This article stems from a nine-year experience as audio describer and overall accessibility coordinator for Macerata Opera Festival, an international opera festival taking place every summer, from mid-July to mid-August, at the Sferisterio open air theatre in Macerata, Italy and bringing over 30,000 people to the opera each year.
Return to this point in the text

Note 2:
The observation protocols, the questionnaires and interviews were carried out with the support of Chiara Pazzelli as part of her MA thesis research. Two other postgraduate students provided support with questionnaire administration and collection in 2016.
Return to this point in the text

Note 3:
The first conference of the successful 'Media for all' series was held in Barcelona, Spain, in 2005. In 2007, the EU-funded 'Digital Television for All' was launched, to be carried out over three years.
Return to this point in the text

Note 4:
Touch tours are organised in collaboration with a team from Museo Omero, the first national tactile museum in Italy, which also provides the tactile tables.
Return to this point in the text