Translation in your ear

TIMEKETTLE: PROVIDES SIMULTANEOUS TWO-WAY VERSIONS

It supports 40 languages and covers 85% of world’s population.

Ever since science fiction author Douglas Adams came up with the Babel fish, a creature that could instantaneously translate between any language inside one’s ears, the concept has been the holy grail of translation technology.

Now it has leapt out of the pages of The Hitchhiker’s Guide to the Galaxy and into our ears. The first of these devices to reach our shores, TimeKettle, promises the world. Or at least, understanding the world’s languages.

It’s WT2 Plus AI realtime translator earbuds are described as “the real-time translating earphones for natural bilingual conversations”.

The system uses two earphones – one for each participant in the conversation – in combination with an app on a smartphone.

TimeKettle claims the system “enables foreign communication that is natural, fluent and preserves both eye contact and body language, allowing you to immerse yourself in the conversation, express yourself more completely and build more meaningful connections with the people you meet”.

The company says that, within a good networking environment, it takes one to three seconds to complete voice recognition and translation, drawing on support from Google, Microsoft and iFlytek, and providing translation accuracy up to 95%. In actual use, both time and accuracy proved rather optimistic.

Its biggest strength is that it supports 40 languages and 93 accents, covering 85% of the world’s population, and provides simultaneous two-way translation.

An offline voice package supports offline translation in seven languages, for when there is no network. However, this requires the use of tokens called, cleverly, Fish. Not so cleverly, the five Fish required for the offline package cost around R175.

It offers 3 modes, namely:

▶ Simul, for simultaneous, allowing one person to talk continuously in a quiet environment and the other person to hear translation continuously.

▶ Touch, designed for noisy environments: one touches earbuds to speak, and again to translate.

▶ Speaker, in which only one person wears an earphone, designed for quick questions or brief translations from various individuals. The user wears one earphone and holds their smartphone in front of the other person.

One can then hear the translation aloud or view the translated text on the app.

H ow does it work?

It uses a combination of speech recognition, text-to-speech technology, translation software and artificial intelligence to convert your speech into text, translate that text into another language, and then turn the resultant text into speech in the language of the listener.

H ow much is it?

Available online at $179 (R2 602) from Timekettle.com and $199 from Amazon.com. Both ship to South Africa, and the former adds a standard $20 shipping cost. Customs may levy a further charge.

W hy should you care?

If you have any interest in travel or doing business with people who speak only a foreign language, translation technology will eventually reach the point where it becomes expected that people will be able to understand each other via the gadgets they carry or wear.

While the TimeKettle is not yet that device, it shows what is possible, and is a clear pointer to a future where the concept of the Babel fish becomes a reality in our ears.

W hat are the biggest negatives?

In the era of the pandemic, the very idea of handing someone one of your own devices and asking them to stick it in their ears is bizarre, to say the least.

▶ The flow of translation is stopstart, and can be deeply frustrating. Typically, one needs to talk briefly, and then pause for halfa-minute or so while the translation catches up and the other person understand what you are saying.

▶ The quality of translation is low, equivalent to what one sees when copying chunks of text into Google Translate and other translation apps or sites. It still requires some interpretation to understand the translation, and that is going to be difficult in real-time.

▶ If you have to use an app anyway, and have to be as intrusive as having to ask someone to wear one of your earpods, you may as well stick to the translation apps. It’s certainly more hygienic.

▶ It does not support listening to music or phone calls. It has only one purpose, namely translation.

W hat are the biggest positives?

Great unboxing experience, with premium components and packaging.

▶ When it works, it is a spectacular display of a science fiction vision becoming reality.

▶ If we see it as an early proof of concept, then the future of translation technology is bright.