Back in the Bar with Steve Jobs

In my earlier post, I speculated on the iPhone software change that Apple would make in response to “antennagate”.

I suggested that, with the new formula, bars would be assigned linearly across the full range of signal strength, from what I called zero, where a connection could not be established and maintained, up to some maximum.

With the new software now released, and after independent analysis, we are able to check my prediction.

Quantifying the Prediction

My prediction was deliberately more qualitative than quantitative. To be strictly quantitative requires that we first agree what “zero” and “full strength” mean. This is somewhat arbitrary. The zero point is determined in part by the performance of proprietary digital signal processing algorithms. The maximum measurable signal strength is a design point.

Nevertheless, to make a comparison, we must first quantify the prediction. At this point, an astute engineer will see that I could carefully choose these two points to make my prediction more closely match reality. I will resist any temptation to do that, and allow the community to set those points for me.

The online consensus seems to be that “zero” is -113 dBm and “full strength” is -51 dBm, a range of 62 dB. Dividing that range uniformly across five bars gives 12.4 dB per bar. We would thus expect the bars to change at -101, -88, -76 and -63 (rounded to the nearest whole dB).

Reality Check

The following chart shows the prediction vs the before and after reality:

Chart of iPhone bar counts vs minimum signal strength
Where I was right

With the exception of a small correction on bar number two, I was correct in predicting linearity.

Where I was nearly right

I came close on the range increase. The range went from 22 dB to 45 dB.  My prediction was (in effect) 49.6 dB.

Where I was half right

I predicted an increase of 28 dB for the fifth bar. In fact the increase was only 15 dB. While my theory was correct, I was wrong to take it to the extreme.

Where I was not wrong

I did not predict that the zero point would be lowered, as it has been, by 8 dB, to -121 dBm.

This is perhaps one of the most astounding changes. 8 dB represents a drop in signal strength of more than six times. It is impossible to imagine that the iPhone digital signal processing performance is that much better than the competition.

The only explanation I can see is that, consistent with my prediction, Apple have redefined bar number one. Whereas it used to mean “you have at least a half-decent signal”, it now means “you have some signal, but not much”.

Where I was overwhelmingly right

In his brilliant fireside chat, Steve Jobs made it abundantly clear that he wanted users to learn more about how their own behavior affects the performance of any phone, and that he wanted his bars to provide an accurate, reliable tool for managing that behavior.

I was spot on. Steve went to extraordinary lengths to explain how behavior affects performance.  Apple’s bars now provide a uniform measure of signal strength over a much wider range than before.

I also implied that Apple were intending to set a new standard. Indeed, Steve almost challenged the industry to “follow the leader”.

What shocked me the most

I was stunned when Steve gave out the relative dropped call number. Steve told us that iPhone 4’s dropped call rate was indeed worse than before, but that the change was less than one call per hundred. That numerator is scary.

Think about it. We can assume that the true number is quite close to one (in a hundred), for if it were much lower, Steve could have said something like one in two hundred, or one in a thousand.

Steve also implied that this was a relatively insignificant change. But for that to be true, the real dropped call rate must in the double-digit numbers per hundred! That is atrocious! Dropped call rates should be much smaller, barely measurable in calls per thousand or even calls per ten thousand, not calls per hundred!

AT&T smartly refused to allow Steve to release the dropped call rates, but they were foolish to allow him to report even the delta.

Post a comment or leave a trackback: Trackback URL.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: