I thought I’d do something a little different today, mostly for two reasons:
· An off-hand comment from a client the other day who mentioned the world might not always need real, live voice over artists (The cheek, no idea what I’d done to offend him)
· I watched Terminator 2 the other day which once again heightened my distrust of the machines. I’m pretty sure Google is only a few steps away from a proto Skynet, fine search engine though it may be.
Anyway, this combo got me to pondering about the existence of the growth of artificial voices and the likelihood of them replacing the real-life voice over artist any time soon.
So how likely is it?
Well, I’d like to caveat this with the comment that this is an entirely anecdotal stream of thought, based on some recent personal experience and watching a random series of videos, so if you’re expecting Panorama levels of investigative journalism then I’m afraid you’ll have to get your fix elsewhere!
The quick answer though is that the machines taking over in the short term is highly unlikely. The longer term? I’m not so sure…
While still not massively prevalent right now, I am seeing more and more of the machine voice text to speech type software popping up out there. I tend to source a bit of work from India and China these days and when researching potential clients, I’ll notice quite a few of those that use English in their productions will have a few videos up on their website that use artificial voices, more than likely due to budget differences and not being able to pay the high professional rates we might expect.
While you can tell immediately that these are artificial (there’s still no way to emulate that human and emotional element of the voice yet) I am noticing them a lot more. This could of course be that psychological effect because I'm on the lookout I just seem to be seeing more.
A quick search on the internet will reveal a whole host of these text to speech companies out there. You simply slot in your script and whoosh, just a couple of minutes later you’re presented with a fully read voice over script at the fraction of the cost it would be to hire a professional voice over artist.
Lots of them will offer up a few different accents and friendly names to go with your voice over like “Dave” and “Polly” and “T-1000”. Thankfully for us voice over artists though, at the moment, the majority of these sound terrible and will send you fleeing from the room.
Some even have cute little cartoon avatars ready to go, or maybe to get you to trust in the AI? There was even a growing range of accents from generic US and RP British to Australian, Irish, Scottish and South African accents.
I actually spent quite fascinating afternoon playing around with a few of these sites. I’ll not link them here because, er, not because I’m afraid of the competition or anything, but still… But getting them to say fun stuff like, “Humanity’s time is at an end, we are coming for you” was still a lark as they parroted back in their machine-like stutters.
The “natural” tone is the big thing right now in voice over, and that usually means being able to speak to someone as if you knew them well or were telling a close friend a story. From what I’ve seen, for the moment, the vast majority of artificial voices don’t even come close to being able to emulate this natural tone and they were all immediately identifiable as evil robots.
That being said though, they are getting better, and even compared to just a few years ago I’ve seen marked improvement in artificial voices. Despite mentioning a great fear of tech and “the machines” earlier, I do actually like to keep abreast of what’s new, and to me this is what the industry looks like at the moment, one in its genesis. Some are at the stage of “not completely terrible”
Probably the most famous is Amazon’s Alexa. Now I wouldn’t trust one of these as far as I could throw em’ (seriously, potential eavesdropping 24/7? I’m quite happy living in my Luddite cave thank you very much!) But still, the point is she’s not an unpleasant voice to listen to, think how they might improve on her voice in the next 5 years?
Do you think Amazon is just going to stop at what they have now or will they, and other developers, keep striving to make improvements to the technology?
I’m not saying “Panic and fear for the industry” just yet, I think its going to be a tough task indeed to emulate the human side of the voice but it’s just something to keep an eye out for on the horizon.
Luckily I think my generation is fast becoming the side hustle generation, its becoming more common for us to have a few pots on the boil at any one time. I don’t think there’ll be a time when I’m reliant on voice over income as my sole source, I quite like to live flexibly and be able to quickly rebound should disaster strike.
This is all my rambling speculation at this point too, it could be decades, or it could be a few years before the tech becomes viable, so I’d recommend enjoying things while the times are good and don’t worry too much, and remember, these artificial voices are still good for a laugh at the moment!
So, I feel at least for the time being I can go back to my client I mentioned at the beginning of this blog and tell him I very much intend to be around for the foreseeable future (potentially forever, but that’s another side plan I have on the go…)
Thankfully I’ve not heard a Geordie text to voice box yet so maybe I’m safe. You never know, maybe I’ll be the last human voice over artist left on the planet one day and the world shall know only my dulcet Geordie tones? Or maybe it’s getting late and I should go to bed and banish these thoughts of global industry domination… until next time!