Aerial warfare and the iPhone 4

When it comes to the actual technology inside the new iPhone, Apple has been heavy on the adjectives ("Magic", "Amazing", "Astounding", et cetera) and light on the details. This is smart marketing, but frustrating for those of us who like to know how things work - and how magic, amazing, astounding those bits will be in practice. The implication that we're supposed to take away is that Apple's tech is unique and years ahead: the suspicion, however, is that because the company uses the same suppliers as everyone else to do the same job as everyone else, the product is much the same as everyone else's, at least beneath the hood. And, given how hard it is to find out the practical details of what's going on, those suspicions are hard to shake.

With the new iPhone, there are three major areas where Apple has sprinkled the adjectival fairy dust: the A4 processor, the Retina display, and the antennas built into the metal sides of the device.

The A4 processor attracted a lot of speculation when it first appeared in the iPad, with some being so borne aloft by the glitz of the launch that they guessed it was enormously different to what had gone before - even to the point of it having an entirely new, Apple-invented architecture. Apple has not been quick to dispel such rumours. Yet now that the chip's been stripped down and mapped, it turns out that it's much the same design as the chip in the iPhone 3GS, albeit shrunk to 45nm. It is an Apple custom design, but built in a similar way and from similar design methodologies as other system-on-chip (SoC) parts from other companies. And, given that the iPhone and pals don't actually do anything magical in hardware that you can't do with parts from other companies, that all makes sense.

The Retina display is a bit of a punt. It's true that there's nothing quite like it on the market, but that's got nothing to do with Apple technology - built by LG Display and using In Plane Switching (IPS), it too is a take on a technology more generally available on the market to others. What Apple has done is take a gamble that this particular version is ready for cost-effective mass production - and, presumably, cut a deal with LG Display to soak up the production line's output for long enough to delay the competition. But there are production problems, and unlike the older iPhones (which used two different but interchangeable display technologies to alleviate supply issues), Apple has nowhere else to go.

The most interesting technological claim that Apple made for the new iPhone, though, is that it has solved one of the more intricate problems facing all mobile phone designers - what to do about the antennas. If your antenna doesn't work well, it doesn't really matter how good the rest of the phone is - without the ability to efficiently couple radio signals into and out of the device, you'll burn power, lose range and speed, and get a reputation for dropped calls. The wavelengths used for cellular comms mean that an efficient antenna should be around 10cm long, which just won't fit into a modern device, so compromise is all.

Antennas have to be designed to match the wavelengths they work at, and the sharper they match the more efficiently they work. At the simplest level, they're like an organ pipe: the precise length and shape of the pipe defines the note it plays when air is blown through it, and if you try to get a different note out of it you'll get little or no sound. You can have an efficient antenna that's physically much shorter than it should be, but that limits its bandwidth - it's tuned very sharply to the working frequency, and small disturbances can put it wildly out of tune. (Apologies to radio engineers, who'll now be sucking air in through their teeth and going "It's much more complex than that".)

In general, mobile phone antennas are buried within the device electronics where there's some freedom to tweak the layout and configuration, but where they're very close to the electronic noise and shielding effects of the other components. Apple has chosen instead to use the metal sides of the iPhone, which sidesteps these issues - there's enough room to have an electrically long antenna, and it's closer to the air.

Which would be fine - others have done this before, right back to walkie-talkie designs of the 1960s - except for one design component. The user. Users have hands and hands are made of flesh, which to a radio wave looks like a big bag of conductive saline solution. As submariners know, if you surround an antenna with a large lump of salty water, it stops working: it's shielded from the outside again, and its electrical length will change dramatically. It'll go out of tune - it's like surrounding that organ pipe with lead, and expecting sweet music. At those 60s walkie-talkie frequencies, the wavelengths were so long compared to the size of the human that this effect was muted. Not so today.

Now, it is possible to have automatic matching circuits that compensate for this effect but at the frequencies used in the iPhone, these would need to be clever indeed -- most likely using MEMS -- and I've seen no hints in the literature that any such technology is close to commercialisation. If Apple has this, then it will be genuine innovation that it simply must have patented. I'm agog to read such a patent.

But if Apple doesn't, then the laws of physics mean it'll have indifferent-to-poor radio performance, certainly less good than its rivals.

And even Steve Jobs cannot change the laws of physics. Even with adjectives.