Apple’s Antenna woes stemming from the poor design of the iPhone 4 are well known. But what has just gotten under my skin is this PC World article Antenna Expert: Apple is Right, iPhone 4 Signal Woes Overblown. The article is laced with the next best thing to professional lying, from a group of self professed Radio Experts AntennaSys, Inc The article blatantly claims;
“This was a non-scientific test, but it was done by two engineers who deal with RF devices for a living,”
and then this is where the spin/falsehood is;
“We succeeded in taking a five-bar display and reducing it to one bar by doing that,” Webb says. “But the call remained solid and never dropped.”
In plain sight they note First that they DID NOT DO A SCIENTIFIC TEST that yes, the signal drops, but the call is Ok. As a geek, and one who has been dealing with wireless communications for 10’s of years. This is insulting, and a professional attempt to baffle with bullshit
Speech is divided into 20 (ms) samples, each of which is encoded as 260 bits, giving a total bit rate of 13kbps (kilobits per second). This is the so-called full-rate speech coding
These 260 bit packets are transmitted in serial order with redundancy, so if a packet does not get received, it get’s retransmitted. The data is transmitted into a Time and Frequency Division Multiple Access digital channel. These slots in which the packets of voice are shared are not contiguously serial for each user. So when packets are transmitted they are separated in the channel. So the packet stream must be reassembled at the receiver side to reassemble the stream of packets back into the audio to be intelligible. the packets are chosen to be small enough so that should there be a packet timeout during transmitting/retransmitting the human ear will not detect it.
The professional lie above begins as such, in the use of unscientific terms as ‘the call remained solid’ in reality it never is, they are packets divided in the stream. They may sound solid, but that is due to the nature of human hearing which will fill in any gaps in the sound, and that the actual bit-rate of the wireless sound is very low, about the speed of a dialup modem. This does not constitute much bandwidth and can be sustained on extremely poor wired and wireless connections. It’s a lie to use this as a measure of signal strength. A piece of string and a couple of tin cans, can sustain a voice channel (analog) with no compression at all. Fidelity is in the ear of the beholder, not a scientific result.
When testing the real through put from the defective Apple antenna, the measured bitrate, collision, and retransmitting errors must be tested on a calibrated bench instrument, not by someone clamping their hands around it and listening for the shoe to drop. Someone should tell the ‘Experts’ at AntennaSys to try the same test on a string between two cans, and see if the ‘the call remained solid’ .