Just how good are Google’s captioning capabilities? Well, we’ve all logged onto YouTube and relied on automatic captions to understand the bare minimum, but even that automatic speech recognition technology – while it has improved over time – leaves a lot to be desired. For those wishing to invest in Google’s Android phones, there’s similar tools: Live Transcribe and Live Caption.
The invitation to try out these features on the Google Pixel 6 Pro came just a few days before the start of this year’s Deaf Awareness Week, and I received the phone slap bang in the middle of it. First impressions were strong, as I admired its huge screen which slightly wraps around the sides, and its camera capabilities. My long-term commitment, the iPhone, has the edge in that regard (Android comes with a slight blur on its viewfinder), but its Motion settings and 360 degree settings are unique to its hardware.
The downsides – of which there are, tragically, many – include an oversensitive fingerprint scanner which fails to recognise your index finger more times than it does, a disorientating Maps app which doesn’t completely take you to your desired location, and a power off setting which is far from convenient. You have to press down on the power and volume button at the same time, and maybe it’s my dyspraxia or declining motor abilities revealing itself here, but it’s impossible to apply the right pressure on both buttons from one finger or thumb alone. Instead, I ended up opting for a menu which lingers on the right side of my screen, next to Live Transcribe – one which ultimately requires more buttons to perform the simple act of turning off a phone than one would typically expect.
Speaking of Live Transcribe, let’s finally talk about the feature which led to Google kindly sending me the Pixel in the first place. The areas in which I could supposedly use the feature are endless, and over the course of a handful of days testing out the tech, many circumstances presented themselves. I was making several train journeys with inaccessible tannoy announcements, and when in the welcoming walls of a theatre foyer, it was hard to hear the call for the audience to take their seats when it was drowned out by the social chatter of others.
Disappointingly, Live Transcribe failed in both of these situations. As an aside, it would also fail to transcribe mobile calls, as described.
And know that I say this as an individual who acknowledges and brushes aside his Apple bias when writing this. I was hoping the tech – designed with the help of Gallaudet University – would do what my (and many other) hearing aids could not, and make out a single voice among the background noise. It failed, instead choosing to inform me there was wind in the background when listening out for the tube announcement, and that there may have been a crowd nattering in the foyer.
Well, duh.
I soon learned it was better for the conversations which are closer to the microphone – an obvious and fair technical requirement which I’ll grant – but for Deaf people like me with residual hearing, it’s often the case that our hearing is somewhat dependant on proximity to the source of the sound. I have no doubt there could be greater benefits for profoundly deaf individuals, however.
Where this technology excelled for me, though, was in the form of Live Captions, capable of transcribing media playing out on the phone. There’s no denying that for the longer, uncaptioned YouTube videos where automatic captions fail, Live Captions offer a sufficient last resort.
I should mention that I was offered the opportunity to speak to the Google employee responsible for the Live Transcribe feature – and I would have asked him about the difference between Live Transcribe and automatic captions, and whether they use the same technology – but I am yet to be given a time for an interview.
There is, however, one app I’ll look to use even when my SIM card is back in the reassuring confines of my iPhone, and that’s Recorder. Granted, rival applications such as Otter offer some respite for the journalists who hate transcriptions, but that only covers you on the free plan for 30 minutes. There are no limits to Recorder, and its accuracy is impressive.
How I wish I could use that adjective to describe my experience with the Pixel 6 Pro and Live Transcribe as a whole.
By Liam O’Dell. Liam is a mildly deaf freelance journalist and campaigner from Bedfordshire. He wears bilateral hearing aids and can be found talking about disability, theatre, politics and more on Twitter and on his website.
Hilary McColl
May 9, 2022
Thanks for this, Liam. It’s always interesting to get an honest review of how devices work in practice. I’d like to hear more about Recorder, please…?
Liam O’Dell
May 9, 2022
Thanks, Hilary. What would you like to know?
Hilary McColl
May 9, 2022
In what ways is Recorder better than Otter?
Liam O'Dell
May 10, 2022
Sure, so I’ve certainly seen a higher level of accuracy on Recorder than with Otter, not least in the sense that Recorder does a much better job of sensing when a sentence has finished, while Otter will transcribe a huge chunk of text without any punctuation. I like to think it has a better understanding of context while transcribing, too (i.e. it will take a better guess at what a word might be based on context, than what Otter would do).
Gavin Songer
May 10, 2022
Hi Liam, where can we download Recorder from? Is it an Android or iOS app (or both?). Would like to download it myself to compare it with other speech-to-text apps that I’ve used. Thank you.
Hartmut
May 9, 2022
Please explain the differences between Live Transcript and Live Caption. And why technically makes it different between Apple based and Android based telephones. Would it also be different if played on a Window-based desktop computer with a larger screen that permits better displays of the output of the two systems.
Liam O’Dell
May 9, 2022
Hey Hartmut, I’ll answer each of these questions in turn. Live Transcription is mostly for transcribing environments and conversations in real life, ‘outside’ the phone (so conversations with friends, for example). Live Caption captions media playing ‘on’ the phone – so YouTube videos, videos on Facebook and Twitter etc.
It’s different to iPhones in this respect because there is no Apple-created app on iPhones which offer transcription – you have to rely on third-party apps like Otter. iPhones also don’t have any feature like Live Caption to transcribe media playing on the phone.
As for Windows PCs, I know there is a Chrome extension which can transcribe sound playing on a browser, but I don’t know a lot about that, and it wasn’t something I looked into for this phone review.
Gartmut
May 9, 2022
Why MUST the environmental sounds be transcribed for us Deafies? They have very little communicative function to transmit to us relevant information for comprehension of the spoken word. Why should we care, how it will be more informative for us? We can do without most of the junk that we must read that are uninformative like [telephone rings] upon seeing someone start to speak on the phone, or [door opens] then seeing someone entering the room. If you describe Live Transcription that way, well kick it out of the window and caption only what is spoken, When something falls, thus making a thump sound, why should we care. Do you care if a tree falls in a forest two miles away? (see the philosophical debate of Bishop Berkeley) You don’t hear it, so be it as if the tree did not fall. The blind do not care much about colors. Why must he be shouted that I have blond hair and blue eyes?
I am afraid you have a different conception of what transcription means and its usefulness as pieces of information. “outside the phone” and “playing on the phone” does not explain your conception of ‘transcription’ and ‘live captions’. I asked WHAT ARE DIFFERENCES (in detail) between them, contentwise and technically, what is being brought forth into the lines on the screen we must read.
Liam O'Dell
May 10, 2022
Hi Hartmut, I completely agree, and if you take another look at my review, you’ll see I criticise Live Transcribe for picking up on environmental sounds more than the actual speech and dialogue I want to hear. I believe I have already answered your question about the technical and material differences between Live Transcribe and Live Caption.
Cathy
May 9, 2022
I completely agree with the Live Transcribe assessment. Unless the speaker is up close and there is literally no background noise, it fails. I use it when I dont have my hearing aids in but otherwise it serves no purpose that the usual combination of aids, visual clues and context dont already provide.
Genuinely interested to find out more about how Live Captions and Record could benefit me though. I’ve used an Android phone for years without any knowledge of either of them!
Liam O'Dell
May 9, 2022
Glad I’m not the only one with a tricky experience with Live Transcribe!
Brenda
May 9, 2022
Hi
What is the recorder app? Hoping it is a transcribing app? Can’t immediately tell which one it is on the App store to download- Thankyou
Naomi
May 9, 2022
Thanks for asking this question because im looking on the app store and can’t see the recorder App either.
Liam O'Dell
May 9, 2022
It’s an Android app, so you won’t see it on the App Store, I’m afraid.
Meryl Evans, CPACC
May 9, 2022
Liam, you are spot on with your assessment. This has been the case with all apps that automatically transcribe audio. While imperfect, let’s celebrate progress. This is huge because Google is working with Gallaudet. Not many companies involve the people who would use the technology. Progress over perfection. Google will, no doubt, keep iterating and improving.
Liam O'Dell
May 9, 2022
You’re absolutely right, Meryl. Hopefully, over time, we’ll see greater progress from Google in this area, and when they improve, we must commend them for it.
Claire
May 9, 2022
Thanks for this review. I’ve also found Live Transcribe to be really poor. Captioning of any video content on my phone using the Live Captions in Chrome is usually okayish. I really hope both of these will continue to improve.
I immediately tried to find the Recorder app but it looks like it’s only available on Pixel phones and not all Androids, which is a huge shame.
I didn’t realise I could get 600 free mins per month on otter, though, so have downloaded that, thank you!