Description :
Technology has become a crucial component in audio-visual localization. While translation and technology companies are working on and releasing exciting cutting-edge tools, it's also important to make the linguistic communities aware of the newly released innovations and things that are about to get out on to the market.

Machine translation is gaining more and more ground in this area as well. While only several years ago using MT for subtitles seemed like a sci-fi fantasy, with neural MT it became not only possible but extremely useful. However, like any technology it does have its limitations and that is why it is only used in technology-assisted human workflows.

In this workshop, we will discuss the advantages of applying MT post-editing for translation of subtitles, and the main technical challenges as well as linguistic challenges faced by subtitlers. We will start with a hands-on explanation of how neural MT works and the specific nuances it has when applied in subtitling.  This will help the audience understand the limitations and strengths of the technology and enable the participants to apply it thoughtfully only in situations when it makes sense.

The second part of the workshop will focus on the linguist aspects of post-editing for subtitles. These include MT errors typical for this content type, errors that are specifically critical or easy to miss, handling character limitation and reading speed when post-editing, among others. This part also includes practical tips on how to maximize post-editing productivity to take full advantage of the technology.

Finally, the third part of the workshop will include practical exercises in different languages based on a short video sample. The task will be performed in an easy-to-use subtitling environment. In this exercise, participants will perform MT post-editing following a discussion of the results. Has MT been useful for the content type? Is post-editing faster than plain translation? What were the main challenges?

Proposed Agenda:

• Introduction: A dive into MT - how neural MT works (30 mins).
• Linguistic aspects of post-editing: typical MT errors (20 mins).
• Post-editing subtitles: specific challenges of this content type (30 mins).
• 5-minute break
• Demo (15 mins).
• Practical exercise and discussion (40 mins).
• Concluding remarks (10 mins).

Target Audience(s):
Translator, subtitler, audio-visual translator

Target Audience Sector(s):
Translation and localisation services

Participant Level:
Any level

Prerequisite Knowledge:
Basic knowledge on subtitling translation

Expected Outcome(s)::
After the workshop, participants will have an advanced understanding of the technology, feel more confident in deciding when and how to apply MT for translation of subtitles, how to handle the challenges of post-editing and become experts in MT for subtitling.

Description:
Easy-to-understand language, ranging from Easy Language (or Easy-to-Read) to Plain Language, has traditionally put the emphasis on written printed texts. However, in the last years a growing interest in how easy-to-understand language could be used in audiovisual media has developed. This has been one of the central topics of the EASIT (Easy Access for Social Inclusion Training) project, an Erasmus + project funded by the European Commission in the 2018-2021 period.

This workshop will be built upon a selection of video presentations created as part of the EASIT project (pagines.uab.cat/easit). The ultimate aim of the workshop is that participants gain a better understanding of the following aspects:

• What is easy-to-understand language? What is the difference between Easy-to-Read (or Easy Language) and Plain Language?
• Who are the end users of easy-to-understand content?
• What are the basic rules of Easy-to-Understand Language?
• Can we create easy-to-understand subtitles? How?
• Can we create easy-to-understand audio descriptions? How?

Apart from the video presentations, we hope to engage in a lively discussion with workshop participants, who will be proposed some hands-on activities.

Target Audience(s):
All interested participants 

Target Audience Sector(s): 
N.A.

Participant Level:
Any level

Prerequisite knowledge:
N.A.

Expected Outcome(s):
After the workshop, participants will have a better understanding of easy-to-understand language and its application in subtitling and audio description

Description:
This workshop will provide an overview of global dubbing practices, processes, and technologies with a focus on cloud dubbing, and more specifically on the translation and adaptation phase. Emerging market demands in terms of professional roles and required skills will also be discussed.

The differences between traditional and cloud dubbing workflows will be outlined from a practitioner and trainer perspective. This comparison is intended to highlight the impact of cloud dubbing systems and their integrated assistive tools on 1) the translators’ role; 2) the product-oriented quality standards; 3) the workflow itself from a functional perspective. It also aims to emphasize the advantages of incorporating cloud dubbing technologies in a training context, while proposing related strategies.

This will be followed by a detailed hands-on illustration of an end-to-end cloud dubbing system.

The workshop will be non-language specific and will adopt a global approach. It is intended for researchers and trainers who teach dubbing, and who might not be familiar with the rapidly evolving industry practices. It may also provide interesting insights to representatives of localisation or software companies who are eager to familiarise themselves with specific details related to dubbing challenges, workflows, conventions and technologies currently supporting the dubbing industry globally. Practitioners who would like to expand their client portfolio across geographical borders may also find this workshop of interest.

Proposed Agenda:

• Overview. Evolving dubbing practices and technologies: cloud-based platforms and assistive tools to facilitate studio recording, text adaptation, and workflow management (30 mins).
• Comparison.Traditional versus Cloud Dubbing workflows from a translator perspective : outcome, impact, quality, functionality, emerging demands, required skills (30 mins).
• Training. Cloud Dubbing from a trainer perspective: localisation talent expansion, skills development. (15 mins).
• Illustration. Cloud dubbing hands-on walk through (ZOOdubs): scripting and adaptation (30 mins).
• Open discussion / Q&A (15mins).

Target Audience(s):
Trainer, lecturer, teaching fellow, localisation manager, head of productions, dubbing project manager, dubbing manager, software developer, product owner, dubbing practitioner, dubbing adaptor, dubbing translator, dubbing researcher, any AVT professionals

Target Audience Sector(s):
Academia, content localisation, software development, localisation training

Participant Level:
Beginner to advanced

Prerequisite knowledge:
Basic knowledge of the dubbing process and dubbing translation

Expected Outcome(s):
1) A thorough understanding of cloud dubbing versus traditional dubbing.
2) Detailed insights into the challenges, quality standards, processes, workflows and tools from an industry perspective. 

Description:
This workshop presented by Dr. Joel Snyder, a pioneer in the development of audio description (c. 1981), will be an interactive examination of advanced audio description techniques.  Effective audio describers strive to offer language that has a foundation in the “The Four Fundamentals of Audio Description” developed by Dr. Snyder:  Observation, Editing-Identifying Key Elements, Language, and Vocal Skills. The workshop will also explore / demonstrate ways in which the delivery of audio description has expanded beyond traditional techniques, i.e., the use of apps providing for the download of audio description tracks and its sync with original soundtracks of video and film; pre-recorded audio description at live events; integrated audio description that incorporates description from the beginning of production design; and the involvement of description consumers as consultants, as voice talents and as audio editors.

Proposed Agenda:

• The Four Fundamentals of Audio Description and their application (30 mins).
• Hands-on walk-through: Audio Description scripting (30 mins).
• Short break (5 mins).
• Demonstration of/interaction with new and advanced Audio Description techniques.  (55 mins).
• Open Discussion / Q&A  (30 mins).

Target Audience(s):
Audio describers, audiovisual translation students and practitioners

Target Audience Sector(s):
Audiovisual translation students; prospective audio describers; audio description consumers (people who are blind or have low vision)  

Participant Level:
Beginner to Intermediate

Prerequisite knowledge:
None

Expected Outcome(s):
Awareness of the fundamentals of audio description – all participants will receive a PDF copy of Dr. Snyder’s book The Visual Made Verbal : A Comprehensive Training Manual and Guide to the History and Applications of Audio Description. 

The workshop deals with two of the most widespread fast-writing techniques, namely respeaking and velotyping, used to produce live subtitles for TV programs, text-to-speech interpreting during conferences, and live reports in parliaments. It is divided into three parts:

The first part is theoretical and is devoted to respeaking and velotyping as techniques which enable professionals to produce an over-120 words per minute (wpm) text. In particular, workshop leaders will show you first-hand:

• How these two fast-typing techniques work
• What the applications of these two techniques are
• What a professional does in parliaments and conferences
• How to test one's knowledge at the ECQA certificate platform

The second part is practical and focuses on the use of voice or the Velotype keyboard to produce a real-time transcript of a fast-spoken text. In particular, participants will understand:

• How to listen and speak/type at the same time
• How to use voice or velotyping to write fast, even in real time.

Because the workshop is online, we cannot provide software or a Velotype machine as expected. Candidates can install a speech-to-text freeware to practice voice writing.

In the third part workshop leaders will introduce MARS, or the capacity of a fast writer to rapidly and accurately transcribe a speech. Thanks to an online tool allowing to measure one's MARS, participants will:

• Understand how this tool works and what it is aimed at
• Familiarise with it
• Practice a 15-minute test to measure one's MARS

 

Proposed Agenda

Part 1: 1 hour

• Welcome and introduction
• Basic concepts: real-time subtitling
• End-users needs
• How do velotyping and respeaking work?
• What does it mean to be a professional (concrete examples) and self-assessment test
• Group discussion
• Organisation of working groups: respeaking and velotyping. Attendants will be split in two smaller groups. The groups will try out both techniques during the second part of the workshop

Break: 15 mins

Part 2: 1.5 hours

• Group 1: respeaking exercises facilitated by Carlo Eugeni
• Group 2: velotyping demonstration facilitated by Wim Gerbecks

After 45 minutes

• Group 1: Velotyping demonstration facilitated by Wim Gerbecks
• Group 2: respeaking exercises facilitated by Carlo Eugeni

Workshop review and satisfaction survey: 15 minutes

Part 3: 0.5 hours

• Introduction to MARS
• Individual test

Target Audience(s):
Translators, interpreters, captioners, subtitlers, court and parliamentary reporters, transcribers, accessibility experts, curious people.

Target Audience Sector(s): 
The present workshop is intended for but not limited to professionals, teachers, and scholars who could find it useful to know more about two fast-writing techniques which not only increase their text production capability but also open to new emerging niches of the market, such as live interlingual subtitling, live intralingual captioning, and live reporting.

Participant Level: Beginner

Prerequisite knowledge: 
Proficiency skills in English

Expected Outcome(s):
By completion of the workshop, participants will be able to: 

• Explain basic concepts concerning accessibility, deafness, and the needs of persons with hearing difficulties.
• Describe the skills and competences which are necessary to work as a real-time intralingual professional by respeaking or velotyping.
• Create basic intralingual transcripts for basic-level texts at their own pace.
• Know how to train one's text-production skills.

To attend the respeaking hands-on session, we recommend you have:

• A headset (no earphones!)
• A great desire to challenge yourself!

Respeaking software
Feel free to contact us at rocio.bernabe@sdi-muenchen.de for further information.

Language of the workshop
The workshop is held in English. All exercises will be in English.