Whose voice is it?

Having ownership of your voice is something we can easily take for granted. We might experience a cold, even have a bout of laryngitis but for the most part when we do speak we sound like ourselves. Our voice is part of our identity. So, when you use augmentative and alternative communication, unless you can record a quality voice of your own, you use one supplied from a choice of manufacturers. The question then becomes whose voice is it?

Recording your own voice

Lots of people are now recording their own voices, especially when they know that in the future, their own voices may not be reliable. I’ve tried recording my own, with the voice of someone whose voice I like. Sadly, to date, I’ve found the quality is not good enough for the type of speaking I regularly do. Recorded voices appear to work well for short utterances of a couple of sentences. Unfortunately for me, they seem to struggle with longer presentations.

When I spoke to the developer, they said the main ‘commercial’ voices spent months in the lab being tweaked and perfected. However, when we record a voice using standardised phrases and put it overnight through a programme, it will be good, but not (yet) have the same consistency of delivery.

Consistency of voice

I use the same voice on two different devices from two different suppliers and in two very different AAC software packages. Actually, it is on more than two devices, as I also have a backup device and use it on my phone in emergencies. For over 15 years, I’ve sounded the same wherever I go. I’ve just had to make sure all the settings are identical, and I’ve been good to go.

A few weeks ago, I switched on one morning, and my voice appeared gruff. It seemed to have developed a bad cold overnight. I did all the usual stuff: loaded, unloaded, updated, restored from backup, and then contacted the AAC software supplier. My voice worked in one package and not in the other.

Yet, the one with no discernible delivery change had developed a couple of pronunciation errors. I spoke to the device developer, who was able to mimic my issue. The solution reached was to create speech exceptions (phonetic spelling in the background).

Developer updates and improvements

The software I use daily for presentations was definitely unwell. While I was online, I saw others complaining, too. It seems that there was a ‘deep neural update’ that made the voices more ‘lifelike’. Well, I certainly don’t have a cold and sore throat making my voice hoarse and hard to understand. Then, there is the glitchy stutter when I’ve been presenting for a few minutes, definitely a robotic element firing in! Eventually, I got back to something that was similar to before but still not quite as good.

Meanwhile, other people were complaining they had lost the option to get their personalised voice. They had to download other apps and re-sign in. Generally, there appeared to be a lot of unhappiness in the AAC user community. This was made worse when it was suggested that users try a different voice. Whose voice is it? It seems like we, as users, were not in charge.

Being out of control

Next, I sat down to do a final run-through for 2 upcoming presentations. These had already been done previously, tweaked for flow and pronunciation, and now the final PowerPoints were just being checked again for completeness. Attention to detail is important for any presentation. Guess what? My voice is not mine again, even though it had been the day before! I resolved this by restoring from an old backup, but I am completely at the mercy of others as to if and when this changes.

Who takes responsibility for maintaining my voice?

To top it all off, this week, I completely ‘lost’ the software I use for spontaneous speech. The voice stopped speaking, so I tried to re-download the app, and it wouldn’t work. Initially, the provider thought it was because I had customised the software so much that I had a very large data set. There were several suggestions on what to do, and this makes me concerned about further customising my dataset (another story).

Who is responsible for maintaining my backup?

The recommendation was to restore an old backup. Luckily I tend to do backup when I do larger dataset or number of programming changes. Yet, when I started to try, I was getting messages that all my backups were corrupted. After initial disbelief by the provider’s support team, they began getting other alerts coming in from around the world that there was a bug in the software. America was still sleeping when my issues began, so it wasn’t logged as anything other than a user error initially.

The bug was rectified overnight. Then, it took a few hours for Apple to give the associated approval. Overall, I was without my main day-to-day device for just over 24 hours. It doesn’t sound like much, but it makes a huge difference to everyday life.

The need for robust systems

Luckily, I have a backup system I can use short-term for everyday communication. However, I’m reliant on both my devices and two different sets of software to carry out my day-to-day activities. I never take for granted that I can speak, but when things happen, it really brings home the reality of the importance of robust tech. I immediately began looking at my diary and planning how I would manage the various commitments without the right kit, frustrating when I already had a busy day planned out.