For a quick introduction to ModelTalker and the process of voice banking, see here.
My client wants to do the recording in my office/clinic/lab, what do I need to do?
Can/should I test out the recording software before my client comes in?
Can my client(s) build voices using my user account?
My client has successfully completed the screening process in my office/clinic/lab. Can (s)he finish the process at home?
My client wants to do the recording in my office/clinic/lab, what do I need to do?
You first need to set up a suitable recording environment. Offices in general are not ideal for recording, because they contain a lot of open space and/or hard surfaces that can create reverberation or echo in the recording. If your office or lab has a sound-proof recording booth, that would be ideal. If not, you need to find a location that will be shielded from background noise like phones, radio, people talking, appliances running, HVAC, etc. You will also need a suitable microphone, specifically a wired USB headset mic. For users recording at home, we used to recommend the (now discontinued) Sennheiser PC 36, because the cost was reasonable compared to professional quality recording equipment. A comparable microphone will suffice, but if you plan to assist clients with voice banking on an ongoing basis, it might be worth considering higher quality options. An investment of between $100 and $200 may allow you to purchase a better quality head-mounted microphone and USB audio interface. See this page for our current microphone recommendations.
Can/should I test out the recording software before my client comes in?
Yes, you should definitely first try the voice banking process yourself so that you know what to expect when you begin to assist your client. Go to our Registration page to create a “Clinician” account and get started. You will be led through the steps of getting set up to record, completing the test inventory (10 sentences) and getting feedback from our staff. If you are setting up a recording station in your office/clinic, this will help you to verify that your recording environment is acceptable.
Once you’ve completed the test inventory, you should go on and complete as much of either the standard or Gen3 inventories as time allows (say a minimum of 150 sentences) to get a solid feel for the recording interface. If you use the web-based recording tool, you will get to hear preliminary versions of your voice built from the recordings you’ve made (check the “Voices” tab after at least 25 recordings). You can then use this voice as a demo for your clients, to give them a very rough sense of what a ModelTalker voice sounds like (note that the final voice will be generated using a newer technology that cannot be implemented within the recording tool at this time). If you want to go further and record a complete voice, presently we are waiving the $100 USD fee to build a synthetic voice for clinicians.
Can my client(s) build voices using my user account?
No, please have clients create their own accounts. Since clients will be storing their speech on our servers, for legal reasons we want to be certain that each client has personally read and agreed to the terms of the service. Further, they need to let us know if we can retain their speech for use in our research or use examples of their speech in public settings. Note that we can only keep a client’s speech recordings if we can use them for our research. However, keeping the recordings will also mean that we will have a copy of their speech we can use to restore their voice should they lose it, and as the technology improves, we will be able to provide updated voices for clients who request it. Finally, because of the way the software is designed, a given user account can only be associated with ONE synthetic voice.
If you plan to assist multiple clients with voice banking on an ongoing basis, we would appreciate it if you would contact us first to let us know. In addition, if your clients create usernames that signal to us that they are working with you, that will help us keep you in the loop when necessary. One way to do this is to prefix their usernames with your initials. For example, SLP Mary Smith has three clients, and their usernames are MS_client1, MS_client2, and MS_client3. We have now added the ability for clients to specify a CC email address when registering. If you have the client enter your email there, the system will automatically copy you on any of our system-generated emails.
My client has successfully completed the screening process (the 10 test sentences) in my office/clinic/lab. Can she/he finish the process at home?
If anything changes in the configuration (such as moving from your office to a home), the test set will need to be redone. It is fine–even encouraged–for you to go over the software and the test sentences with clients in your office/clinic. Face-to-face assistance is usually way more valuable than the email assistance we are able to offer. However, if your client wishes to record at home, they will need to do (or redo) the test sentences at home. Basically, with any change to the recording location or the hardware being used to record (both microphone and computer), we ask that clients do a fresh test set so that we can make sure everything is still OK.
If you do have clients record a test set in your office for practice, please be sure to tell your client to re-record all 10 sentences at home — and contact us if the recordings were submitted. We may not be able to tell that re-recording is needed and could inadvertently advance then to the standard sentence set before they have redone the test set.