The Soundhound software can recognise even a hummed tune, although they haven’t heard this man’s humming
An interesting feature of this is a heat-map, showing where you actually type on the virtual keyboard
Pretend you’re a kid again by building a Heath Robinson effort with Bithack
Constructing panoramas has become a lot easier with smart phones
One or two applications that are going to solve earworms, butter fingers, and idle hands.
Free or £2.99
For anyone who has ever tried to remember a tune – even being able to sing or hum it, but still not able to put a name to it – SoundHound could be a godsend. A music recogniser for Apple iOS, Android, Symbian and Windows Phone, not only can it recognise most music from just a few seconds of playing, it can recognise a hummed or sung tune. According to its developers, this'is because it works on sound,'without conversion to text.
As well as identifying songs, SoundHound can be speech-activated; speak the name of a band, say, and it looks up all sorts of information for you such as album art, links to relevant Wikipedia pages and videos on YouTube, tour dates, and lyrics if available – the service claims to have them for around a million tracks. The company also packages this capability as a separate app called Hound.
A new feature called LiveLyrics also scrolls through the lyrics in synch with the music. Unfortunately, this is only available in North America for now.
If you have a paid-for premium subscription to Spotify, or are willing to sign up for one on the spot, you can jump there from SoundHound and play the identified song straight away.
Of course it is not perfect. It is mainly rock and pop-oriented, and it failed on most classical tracks we tried, as well as on some Asian pop or alternative music. However, on other songs it was excellent, even recognising cover versions correctly – and its ability to recognise humming is simply stunning.
Free trial, then £1.88
There's a stack of alternative keyboards for touchscreen devices, and for Android in particular. Some change the key sizes, some let you swipe from letter to letter without lifting your finger off the screen – and some, such as Swiftkey X from London-based developer TouchType, aim to do a much better job of predicting what you will type next.
Billed as "the world's first social AI keyboard" and having only come out of beta testing late last year, Swiftkey is a highly effective predictive engine and keyboard for Android phones and tablets. It learns from your typing and, if you permit it, can also learn from how you type in other apps such as Facebook, Twitter and Gmail.
This makes it remarkably effective at personalised prediction. If I type 'shopping' it knows that the most probable next word for me is list, say. It also differentiates between capitalised and lower-case words, so for the next word after 'happy' it offers 'birthday' and 'Christmas', but 'I'm happy' returns 'for', 'to' and 'with'. Tap on the word you want and it is entered for you.
Swiftkey understands language structures and can even recall whole sequences of words, allowing you to enter a common phrase with just a few presses, and it handles punctuation intelligently, removing and adding spaces as needed. You can set it to predict in different ways, depending on whether you normally type fast and rely on your predictive text engine to correct errors for you – this is where all those messages which swap 'good' and 'home' or say 'nun' for 'mum' come from – or whether you typically choose one of the offered predictions before completing the word.
For tablets, you can use the phone version but there is also a tablet version with bigger keys in portrait mode and which splits in two when in landscape mode to put half the keys at each end of the 'board', rather like the ergonomic PC keyboards which were all the rage some years ago. The only downside is that this is almost twice the price of the phone version.
It supports some 35 languages, with up to three active at once, and multiple keyboard layouts, and it keeps stats on its performance. It can tell you how many words it has corrected or how many predictions you have chosen, and how much more efficient your typing has been as a result. An interesting part of this is a heat-map, showing where you actually type on the virtual keyboard.
Free trial or £1.80
"Build a complex machine to perform a simple task" runs the blurb for this lovingly rendered Android app of engineering and physics. Both a game and a virtual mechanical building kit, Apparatus lets you solve puzzles by hammering together bits of wood and adding wheels, weights, pulleys and motors – plus in the latter stages, controls so you can adjust or steer the device as it runs.
The free 'lite' version includes just a few game levels, but is a good way to learn the basics. The full version adds more challenges and provides a 'free build' option as you complete each of the game levels. This enables you to practise your building skills and try solving the puzzle in other ways – or simply amuse yourself by building weird contraptions. You can even modify the challenge to make it better or simply harder.
The full version also provides a physics sandbox where you can build even more complex constructions and test your engineering skills. It lets you build new game levels or challenges, and then publish your constructions.
All this is smoothly rendered as a sort of tilted tabletop, which you can move around and zoom in and out as you would expect. It can be remarkably addictive, so it is just as well that there is a save option for the various levels as you progress. The one thing lacking is much in the way of tips or advice, which would be especially useful in the early stages of learning the game, but it is pretty simple for anyone with an inquiring mind to pick up so that is no great problem.
Are 360° panoramic photos artistic or merely practical? In November last year, the Slovenian information commissioner Nata'a Pirc Musar declared that they were the latter, because in her view their main purpose is to show the space where they were taken, rather than the ambience or circumstances. She added that photographers therefore need to get consent from anyone photographed that way, or else blur their faces for anonymity.
Some might find this attitude quixotic, not least because panoramas no longer require "specialised photographic equipment", as Musar seems to think. Indeed, like so many other photo-related tasks they are now perfectly feasible on a smartphone, and where once you would have needed something a tad more powerful to actually stitch the images together, even that can be done in software on the handset.
One such tool is Pano, an app for iOS and Android that can automatically stitch together up to 16 shots, applying colour-correction, blending and so on as required. It is a very impressive trick – you simply turn on the spot, taking photos as you go. It is very simple to use: the app guides you in making sure each sufficiently overlaps the previous one, and after a minute of thinking it presents you with with all the photos assembled into one panoramic strip.
There are a few caveats; the finished resolution will generally not match that of the camera. That is because Pano first scales your photos – the amount depends on the phone – and then must discard any excess at top and bottom if the images do not align perfectly. In addition, unlike a few other panoramic apps, it only does horizontal images, not hemisperical.
A couple more apply to any panoramic app. For example, email programs such as the iPhone's may shrink an emailed panorama if they limit the width of photo attachments, and if people move around between overlapping photos, you may see ghosts. Oh, and think twice before trying it in Slovenia.
|To start a discussion topic about this article, please log in or register.|
"The benefits of footing the bill to put a British astronaut in space amount to more than just a restorative for national pride"
- NHS doctors sharing confidential data via unsecure devices
- India rejects Dassault jet in favour of obsolete domestic design
- Israeli team secures first launch agreement in Moon race
- Uber accuses rival of stealing data on its drivers
- Volkswagen emissions software also used in Europe
- Renewables hold the key to tackling climate change