Slashdot Mirror

← Back to Stories (view on slashdot.org)

CNBC: Google's New 'Pixel Buds' Suck (yahoo.com)

Posted by EditorDavid on from the hammering-on-headphones dept.

Google's new Pixel Buds "are really bad" and "not worth buying," according to CNBC's technology products editor: The stand-out feature of Google Pixel Buds is that they're supposed to be able to translate spoken languages in near real-time. In my real-world tests, however, that wasn't the case at all. I took the Pixel Buds out on the streets of Manhattan, speaking to a Hungarian waiter in Little Italy, multiple vendors in Chinatown and more. If you press the right earbud and say "help me speak Chinese," for example, the buds will launch Google Translate, you can speak what you'd like to ask someone in another language, and a voice will read out the translated speech through your smartphone's speakers. Then, when someone replies, you'll hear that response through the Pixel Buds.

The microphone on the Pixel Buds is really bad, so it barely picked up my voice queries that I wanted to translate. I stood on the side of the road in Chinatown repeating myself at least 10 times trying to get the phone to pick up my speech in order to begin translation. It barely worked, even if I took the buds out and spoke directly into the microphone on the right earbud, and often only translated half of what I was trying to ask. In a quiet place, I was able to allow someone to respond to me, after which I'd hear the English translation through the headphones. That was neat, but it barely ever actually worked that way. To mitigate this, I found it was just easier to manually open the Google translate app, speak into my phone's microphone, and then let someone else also speak right into my phone. This executed the translation nearly perfectly, and meant that I didn't need the Pixel Buds at all.
The article ends by answering the question, Should you buy them? "Nope. There's nothing I recommend about the Pixel Buds.

"They're cheap-feeling and uncomfortable, and you're better off using the Google Translate app on a phone instead of trying to fumble with the headphones while trying to translate a conversation. The idea is neat, but it just doesn't work well enough to recommend to anyone on any level."

8 of 100 comments (clear)

  1. Questionable test by mridoni · · Score: 4, Insightful

    So, you take a brand new technology, that is expected to have some rough edges, you test it in the worst possibile environment (noisy and crowded streets with a lot of traffic) and you're surprised of the result? Moreover you used: a) a language (Chinese) that due to its nature is really difficult to recognize efficiently. b) a language (Italian) as spoken by Italian-Americans of several generations, so with a strong accent, regional influences and maybe a few grammatical errors in the mix (I'm Italian-Italian myself so I know what could be expected).

    I'm not saying that Google buds are great, maybe they really do suck, but this sounds more like a rant than a well-informed test. Then, of course, can debate whether Apple's approach (bringing a technology on the market when it's mature, instead of jumping first on the bandwagon) works better or it's just a strategy to make your competitors fail in a series of inevitable pitfalls.

    1. Re:Questionable test by dmomo · · Score: 3, Insightful

      Right. I have zero doubt or cynicism as to the sincerity of this post. But all of the complaints are merely bug reports, more or less. That is to say, the goals and approach of the technology is sound, and once the rough edges are worked out, it will be a game changer. The current state of the audio filtering technology is not up snuff. So be it. These real-world tests will only make it stronger. The poster uses strong language (sucks), but the frustrations are helpful and will only improve the tech.

    2. Re:Questionable test by willy_me · · Score: 5, Insightful

      The current state of the audio filtering technology is not up snuff.

      But all of the complaints are merely bug reports, more or less.

      Do not assume this is simply a software problem. To perform proper audio filtering, multiple microphones are required. Dedicated logic would then typically perform the filtering - a CPU would require too much power. If the Pixel Buds do not do this correctly now then it is highly unlikely a future software update will correct the problems. This is not a bug report but a fundamental hardware design flaw.

      It is possible future hardware will perform better. But the current hardware is what it is - a poor purchase according to the article. It might be great for you but if not, do not assume the ear-buds you purchase today will improve tomorrow.

    3. Re:Questionable test by mccalli · · Score: 3, Insightful

      They're asking whether you should buy them or not. That they might work in lab conditions is great, but using (not testing) in a realistic setting to do realistic tasks they're clearly short.

      Here's MacRumor's take on them. They pretty much come to the same conclusion - skip the Pixel Buds and if you want this kind of thing, then pick the AirPods. Now yes, it's MacRumors but they've recommended 3rd party above Apple in the past.

      For myself I have a set of Jaybird BlueBuds X, the first version pre-Logitech, and use them with some Comply Foam ear tips. Sounds great, and for me whilst I've not actually owned any 'truly wireless' I feel I'd prefer the cord which I links the ear pieces together, purely so I can keep track of of them and not lose anything. Other people's preference my vary of course.

    4. Re:Questionable test by jareth-0205 · · Score: 5, Insightful

      Right. I have zero doubt or cynicism as to the sincerity of this post. But all of the complaints are merely bug reports, more or less. That is to say, the goals and approach of the technology is sound, and once the rough edges are worked out, it will be a game changer. The current state of the audio filtering technology is not up snuff. So be it. These real-world tests will only make it stronger. The poster uses strong language (sucks), but the frustrations are helpful and will only improve the tech.

      Do we really care if it's a fundamental problem or a bug, as consumers? I care about what it does now, if I'm going to pay for it now... and I would hope that reviews do the same. It may get improved, but there's no guarantee, and indeed Google's habit of dropping products means I really need to be happy with it now. If it gets improved in the future then I'll reconsider, but for now these seem to be dud products.

      And I say this while posting from a new Pixelbook... I've been generally pretty positive about Googles slew of new products, but this one seems not ready.

    5. Re:Questionable test by SvnLyrBrto · · Score: 3, Insightful

      Those are real-world conditions in which real-time translation would actually be... you know... useful. The thing is nothing more than a bad gimmick if it only works in a quiet old-timey lecture hall on properly enunciated, grammatically perfect church latin.

      If *I* go out, in my city, the top non-english languages I'm going to hear are Chinese, Tagalog, and Spanish; all spoken in a wide variety of accents and dialects owing to what region within China, the Philippines, or Latin America the speaker (or his/her parents) originated, how long and where they've been in the US, their socio-economic class, and so on. And yes, I'm going to hear them on noisy streets and in restaurants and shops and bars.

      If a device that bills itself as a realtime Star Trek style universal translator can't cope with those languages and conditions; it's useless for that purpose. I may as well just try sticking a fish in my ear.

      --
      Imagine all the people...
  2. Breakthrough medical device coming... by Applehu+Akbar · · Score: 4, Insightful

    The current state of the audio filtering technology is not up snuff. So be it. These real-world tests will only make it stronger.

    And when audio filtering does get good, the world will finally have a hearing aid that works.

  3. Re: Google by ChatHuant · · Score: 3, Insightful

    I have never heard of Google spying on anyone.

    I'm afraid this only proves you're out of touch. Neither Google nor Slashdot are telling me they'll track me when visiting Slashdot; yet Slashdot notifies Google of my activity. Neither Google, nor my credit card company, nor the supermarket where I'm buying something tell me my purchase, even though not on-line will still end up in Google's databases. When Google cars drove in neighborhoods and "accidentally on purpose" intercepted WiFi activity for years, they never told me (or anyone) anything - until an audit requested by Germany's data protection agency caught them. So describing Google's activity as spying is perfectly correct.