Voice Banking, Message Banking and Story Banking A Guide for preserving the voice for the use of augmentative communication technology This information may be used as a guide when discussing medical or health changes that affect an individual s ability to speak. Technology and anecdotal research have led to the development of several strategies that will help an individual retain access to their natural speaking voice (or an approximation of it) despite physical limitations that may prevent verbal communication. VOICE BANKING 1 : requires special software in order to record a large sampling of a person s speech that is then consolidated to create a synthetic voice that approximates their natural speaking voice. MESSAGE BANKING: is the process of using a digital recorder (or computer recording software), to record and store words, phrases, sentences, and personally meaningful expressions in one s natural speaking voice while it is still intact. STORY BANKING: is a process similar to message banking, however, the user focuses on recording favorite stories. These personal, legacy stories are then available for sharing with others later, when speech has become more difficult. The decision whether or not to voice/message/story bank and how to go about doing so is a personal one. It is possible that a person does a combination of all three methods. Any type of banking allows the user to secure their natural speaking voice in the event that they may require or choose to use augmentative communication. We call this an insurance plan for the voice. VOICE BANKING Things to know By using specialized software to bank the speaker s speech characteristics, the result is a SAPI 5 voice. A SAPI 5 voice is what is used in communication devices that provide voice output. In this case, the speaker is able to spell and create unique messages which are spoken in a voice that approximates one s natural speech, instead of it being a 1 Definitions adapted from: Costello, John, M., Message Banking Guide, Boston Children s Hospital (2012),
foreign computer-synthesized voice. For more information, visit the Modeltalker website, www.asel.udel.edu/speech/modeltalker.html. Modeltalker is a software program developed by the University of Delaware Speech Research Lab. It is currently in beta-testing and is offered for free. Recording your voice for speech synthesis: Doing the recordings in a way that gives you the best chance of making a high quality synthetic voice is difficult. Get all the information you need first, by doing the Modeltalker tutorial within the software (after installing it). If you decide to bank your voice using the Modeltalker software, you will need: A computer (lap-top or desk-top) A quiet location to perform the recordings (small room with a door, preferably carpeted, curtained or having no windows) 2 A USB headset with earphones and an adjustable microphone. The Logitech G330 USB Headset or Logitech H530 USB Headset (widely available and priced in the $35-$50 range) are possible choices. Time and energy- 10-20 hours depending on how effectively you are able to make your recordings Willingness to keep trying if unsuccessful at first attempt! Steps: 1) Register and create an account on the Modeltalker website: http://www.modeltalker.com/demo.php. 2) They will then email you the registration/account code that is used within the software. 3) Install the Modeltalker software and sign in using the registration code provided (username and password). 4) Record the 10 sample sentences a. Set up the environment: turn off unnecessary electrical devices or other noisy items, close door, hang curtains/blankets over the window, ask your spouse to avoid making noises b. Put headphones on, position the mic so that it s near the corner of your mouth c. Record each sentence (as per instructions given in tutorial) d. Upload them to the Modeltalker uploader. 5) The recordings will be reviewed by the Modeltalker technicians, and they will inform you by email regarding whether or not the recordings were acceptable. They will give you suggestions if you must make a second attempt. 2 Soft surfaces such as carpet and curtains absorb sound reverberations and help block outside noises. Some individuals report having to turn off all lights, the fridge, as well as the furnace to adequately diminish back-ground noise and general acoustic interferences. Even the sounds of birds or street noise can interfere with the recordings.
6) If accepted, the full inventory (1600 sentences) will be made available to you. 7) Record the full inventory of sentences. You may choose to spend 2 hours a day for a few days, and then be done with it; or, you may choose to spend half an hour a week for three months. It is up to you. However, the Modeltalker technicians suggest the recording process be done over several days only, a few hours at a time, and at the same time each day. People's voices are usually strongest in the morning. In order to obtain a good quality synthetic voice, **consistency is crucial**. So the sound of the voice should not waver much throughout the recordings. If you feel you are getting tired, stop and continue the next day. (Modeltalker coordinator, email with speech instructions, Oct 2013). Result: The resulting SAPI 5 voice is only an approximate version of your voice. It will have similar speech characteristics but it is still going to sound slightly robotic. Single words or short phrases often sound better than long, expressive sentences or stories, where the natural rise and fall of intonation will be lacking. It is a worthwhile endeavour, however, especially if you anticipate being a long-term augmentative communication user. One client reported completing the process in only a few hours, once an optimal recording environment had been established. Although this client never used the voice, she felt better having it just in case, and was happy with how it sounded. Other individuals were unsatisfied with the end result, and did not use the voice for augmentative communication. It is important to be aware that you may not like the sound of the synthesized voice, and should not have unrealistic expectations about how it will sound. The Modeltalker technicians are very receptive to your questions and feedback. They want to ensure that you are successful in creating a synthesized voice. You can call or email them with questions at 302-224-2475 or beta@modeltalker.com. MESSAGE BANKING 3 Things to know Digitally recorded messages give you access to a set of common phrases, spoken in your own voice. Recordings are often made within each of these categories (see template attached): 3 It is helpful to have the guidance of an SLP or SLP assistant when initiating the process of digital recording. At the beginning, there may be technical and/or conceptual questions that are resolved through just getting started.
1) Greetings, e.g. What you say in the morning, or upon waking up 2) Needs and comfort, e.g. the things you may ask for help with 3) Health/medical, e.g. describing pain 4) Food/drink, e.g. food preferences 5) Conversation/chat, e.g. questions you may want to ask your grand kids, your spouse etc 6) Entertainment needs or topics, e.g. getting assistance to turn on the TV or sign-in to Facebook 7) Expressing emotions, concern, or negation, e.g. telling someone you d rather talk another time 8) Specific personal messages, e.g. how you say goodnight to your daughter 9) Money/legal matters, e.g. asking if the bills have been paid Record messages by using: A digital recorder- recommended is the H1-Zoom 4, a handheld and battery powered device with an SD card, or Speaking Dynamically Pro recording software (used on a computer) and an external mic, or Audacity recording software (also used on a computer) and an external mic A handheld digital recorder makes it easy to add messages at various times of day, as they are thought of. The SD card also makes file transfer quick and straight-forward. Result: The messages are saved and catalogued as.wav files. This may be done by the SLP, Assistant or the individual themselves, if they are comfortable doing so. The.wav files are labeled according to their content, i.e. How are you.wav and can later be used in a variety of augmentative communication technologies by linking the audio file with a button. This will allow the user to retrieve a message and speak it in one s own voice but does not allow for the creation of novel messages by spelling. Individual words can be combined to create unique messages, however, the output will sound more staccato than natural speech, or when compared with a synthesized voice. Clients report that there is a significant difference between an artificial voice saying personal phrases like I love you versus hearing their own voice saying I love you. Messages that are more personal/emotional should take priority if an individual chooses to record only a small set of recordings. 4 After consultation with John M. Costello, Director, Augmentative Communication Program, Boston Children s Hospital
STORY BANKING Things to know Similarly to message banking, either a digital recorder or recording software is used to record stories (see list above, in Message Banking ). Users often choose to record both short and longer stories that fall into these categories: 1) The way things were growing up 2) Funny family memories 3) Foolish mistakes you made and mischief with friends 4) Historical/political experiences you ve witnessed 5) Major milestones- schooling, jobs, moves, marriage, children 6) Romance stories- how you met your spouse or partner, how you became engaged to be married, joys and troubles you faced in the relationship, inside jokes you share 7) Accidents/injuries or near accidents/injuries 8) War experiences 9) Specialized knowledge and skills, eg. how to fish, how to make a quilt, how to bake the best bread 10) Children s stories/anecdotes, rhymes and folk tales Result: The recordings can be loaded into a communication device or onto a computer then retrieved at desired times. Everybody has stories to tell- memorable experiences, specialized knowledge, and funny jokes they love to share. In many ways, it is storytelling that provides social closeness and builds relationships. Recorded stories serve as a precious keepsake that can be passed down to future generations. ~~~ While there are many ways to preserve your voice, it is not crucial to do so. Modern technologies allow for brilliant voice output in an increasing number of devices (tablets, specialized augmentative communication devices and computer software) and in a number of accents and languages. The most important thing is to know the options so that you may make an informed choice about your future communication.
For further questions regarding services in BC, please contact: CAYA Communication Assistance for Youth and Adults 700-655 West Kent Ave. N. Vancouver, BC, Canada V6P 6T7 Tel: (604) 326-3500 Fax: (604) 266-2463 contact@cayabc.org, ALS Centre, Vancouver Coastal Health GF Strong Rehabilitation Centre 4255 Laurel Street Vancouver, BC, Canada V5Z 2G9 Tel:(604) 737-6320 Fax: (604) 737-6234 Brigitte Poirier, SLP Written by: Michelle Bunney M.Sc RSLP Certified in Speech-Language Pathology by SAC CAYA Communication Assistance for Youth and Adults CAYA would like to acknowledge and thank John M. Costello, Director, Augmentative Communication Program, Boston Children s Hospital, for his mentorship and consultation as we develop this service.
TEMPLATE FOR MESSAGE BANKING: Add to this list and modify the example phrases to reflect your personal style of speaking. Check off the messages as they are recorded. Greetings Hi there! How are you doing? Have a good day! Good to see you! Needs and comfort Can you put a blanket on my legs? Can you open the window? Please turn on some music. Can I have a tissue? I d love a foot rub! Can you tilt my chair back, please? Health/medical I have a cramp in my. There is numbness in my. I ve been feeling tired. It is difficult to speak. I need some advil. I think I have a cold.
Food/drink I d like some tea/coffee with. Water, please. I don t like. I don t eat. Can you add some salt? I d love a snack. What s for dinner/lunch/dessert/breakfast? Conversation/Chat Hello, how are you? What s new? Thanks for coming to visit. Please wait. I have something to say. That s not what I meant. How s?
Entertainment needs or topics Can you grab me a newspaper? What s on the sports channel? Can you change channels for me? I d like to check my email. Do you want to look at some photos? I ll listen to my audiobook. Expressing emotions, concern, or negation I m worried. I m not very happy about that. That is exciting! I d rather not go today. Let s talk another time. Sorry, but I can t. Specific personal messages You re my sunshine. Hi, sugar! Aren t you a sight for sore eyes. Love you, pookie! I love you, darling.
Happy Birthday! Money/legal matters Have you seen the telephone bill yet? Can we get some cash out? I haven t paid for that yet. I m not sure if that is covered by extended medical. We should call to find out about that. Miscellaneous personal expressions e.g Oh dear me oh my!