Power use of bluetooth vs Wi-Fi vs wired audio on the iPhone 5

▼ With all this bluetooth audio going on, I wondered: how much power does playing back audio over bluetooth use? So I tested this. The tests then snowballed as I got curious about audio over Wi-Fi, or having bluetooth and/or Wi-Fi on when the phone is idle. All the tests were done on an iPhone 5 with a freshly replaced battery under iOS 7. The results:

• Playing audio through the headphone jack, airplane mode, background app refresh off. In 7:59 the battery went from 50% to 34%, 2% per hour.
• Playing audio through the headphone jack, Wi-Fi off, bluetooth off, background app refresh on. In 8:04 the battery went from 50% to 31%: 2.4% per hour.

So in airplane mode, you should be able to play audio for 50 hours and with 3G enabled but not Wi-Fi or bluetooth and no background app refresh to trigger CPU and network use, it's a bit over 40 hours. I guess this also means that just running the 3G radio uses 0.4% of the battery per hour, so the standby time of an untouched iPhone 5 would be 250 hours. (A week and a half!)

Now let's add some bluetooth and Wi-Fi:

• Playing audio through the headphone jack, Wi-Fi on, bluetooth on, background app refresh off. In 8 the battery went from 44% to 23%: 2.6% per hour.

With these settings, you'd be able to play audio for over 38 hours, three hours less than with Wi-Fi and bluetooth off. But what if we start actually playing audio over buetooth?

• Playing audio over bluetooth, Wi-Fi off, bluetooth on, background app refresh on. In 8 hours the battery went from 72% to 45%: 3.4% per hour.

Now we're down to 29 hours of audio playing. So if you really only use your iPhone to play audio and you need it to last as many hours as possible between charges, using the headphone jack certainly makes a big difference. But you'd have to charge your bluetooth headphones two or three times before the iPhone runs out of juice! In normal use, this is more than good enough to use bluetooth for audio when that's convenient and not worry about the iPhone's battery.

However, there's another way to stream audio from an iPhone: over Wi-Fi.

• Wi-Fi streaming, bluetooth on, background app refresh off. In 9:55 the battery went from 89% to 65%: 2.4% per hour.

So streaming audio over Wi-Fi uses the same amount of power as playing through the headphones. Who knew?

Last but not least, what's the impact of having Wi-Fi on on the standby time when the iPhone is idle?

• Standby, bluetooth on, wifi off, background app refress off. In 8:49 the battery went from 23% to 18%: 0.6% per hour.
• Standby, bluetooth on, wifi on, background app refress off. In 9:44 the battery went from 18% to 11%: 0.7% per hour.

So with just 3G and bluetooth on the standby time would be a week and with 3G, bluetooth and Wi-Fi on six days. In other words: having bluetooth on uses an extra 5% of your battery each day, and bluetooth + Wi-Fi 7% per day.

Please note that I did every test only once and the accuracy and precision of the battery charge indicator are limited, so this is just a rough estimate, don't be fooled by the fact that there are numbers behind a decimal point. But I think the overall conclusion that you can simply leave Wi-Fi and bluetooth on is a pretty safe one.

Permalink - posted 2014-10-21

Disabling logging in to a Mac with an Apple ID

▼ If you have more than on Mac, it can be very useful to mount another system's drives or share another system's screen. Obviously you have to provide a username and a password for this, or the entire internet would be able to get at your Mac once you enable these services in the Sharing pane of the System Preferences.

A few years ago, a new way to log in to another Mac appeared: with your Apple ID. This can be convenient, but it has a big issue: as Mat Honan learned, others may be able to get at your Apple ID using vulnerabilities at third parties and social engineering. And maybe Apple can be forced to release your Apple ID to various governments.

That's bad enough, but now under (Mac)OS X 10.10, it's no longer an option—you're automatically logged in to the other computer if your accounts on both computers are logged in to iCloud using the same account. Which means that there's no good way to log in using different accounts.

So I wanted to disable this behavior, but there's no obvious setting anywhere that governs it. The only way I was able to turn this off was to sign out of iCloud. I'm not a big iCloud user, but it has a few useful services, so this wasn't the solution I was looking for.

Turns out, there is a somewhat hidden way to do this. First, open Users & Groups in the System Preferences. You then probably have to click the lock and authenticate in order to make changes. Then right-click (or click holding the control key) on your user name in the list on the left, and select "advanced options". This brings up the following information:

After deleting the two entries in the list of aliases, I was no longer automatically logged in when connecting to this computer, and the option to log in using an Apple ID was no longer offered.

Permalink - posted 2014-10-20

Bluetooth earbuds: Jaybird Bluebuds X

▼ With the impending arrival of the Apple Watch, it's time to start looking for some new bluetooth headphones/earbuds.

When I checked my RSS feeds this morning, I saw that Marco Arment had reviewed three sets of portable bluetooth headphones. He pretty much came to the same conclusion as I did when I reviewed the Sennheiser MM450 for Ars Technica back in 2011:

The big question is whether it's worth all the trouble. Having to charge another battery pretty much every day, being surprised by "bad wireless" days, the poor audio quality in HSP mode, the video-killing lag, and the extra cost all scream "No!"

Still, getting rid of that wire is really nice. Walking into the kitchen while listening without even bothering to pocket my iPhone is good. And not seeing the iPhone fly through the air because the headphone cable got caught behind a doorknob once again is even better. Apparently, the answer is a qualified "yes" after all.

As per the reasons above, when the MM450 broke I wanted to get another set of bluetooth headphones, but although I liked the MM450, I didn't like the price and durability. So I got the much cheaper FSL Zero instead. The FSL Zero are great for indoors, but the size and sweaty ears issue make it a non-starter for use outside when it's warm or when exercising.

Enter the Jaybird Bluebuds X bluetooth earbuds ($140 at Amazon).

Yes, they're a little bigger than most earbuds, but only a little. They weigh 13 grams. That's half of what a typical AA battery weighs. But the battery life is 8 hours. I don't know how they did it, but the iPhone shows a very tiny battery indicator for the Bluebuds next to the bluetooth icon. Another nice touch is that there's a voice that tells you the battery is running low, as well as telling you when the earbuds are turned on, connected using bluetooth or in pairing mode. Pairing was super simple. There's a red/green LED, but it stays off during normal use. The Bluebuds come with a carrying case and a short USB-to-micro-USB cable, but no charger.

The Bluebuds compared to Apple Earpods and the Sony DREX12IPB

With regular earbuds, I always use the smallest tips because apparently I have small ears. On the Bluebuds these are a bit bigger than normal (and not interchangeable with ones from my wired earbuds) but they still fit—and didn't fall out over the course of several hours' use. Like Apple's Earpods or the Sony DREX12IPB that I normally use, the Bluebuds have three buttons: volume up, volume down and play/pause. With wired headphones/earbuds you skip ahead by double pressing the play/pause button and back with a triple press. On the Bluebuds this is accomplished with a long press on the + and - buttons. A long press on the play/pause wakes up Siri and a really long press turns off the Bluebuds. The volume buttons do the exact same thing as the volume buttons on the iPhone itself, which is good.

Audio quality is very good for podcasts: voices sound natural. As with most earbuds, there is less bass and more high end, but there's still a reasonable amount of bass and the high frequencies aren't harsh. Obviously you wouldn't want to use these to do serious music listening, but the Bluebuds certainly aren't terrible for music. You can even use them when watching video, there isn't any noticeable lag.

You wear them either using the "traditional under-ear fit option" where the wire between the earbuds runs down your cheek and the controls are on the right, or with the wire running up and then behind the back of your head, with the controls on the left. Unfortunately, there's no option to reverse the left and right channels. Jaybird claims the wireless reception is good enough to leave your phone in your gym bag when you work out, or wear it anywhere on your body. Looks like they're not exaggerating: I had barely any dropouts out in the open, where my old MM450s would always struggle. Walking into the kitchen is also entirely doable.

I'm impressed with the Bluebuds. Without any wires or over-ear foam pads but with better than average bluetooth reception, they allow for lots of situations and activities where other types of headphones or earbuds just don't cut it. However, only get them if you can wear in-ear earbuds comfortably, as they are a bit bigger than the wired variety.

Permalink - posted 2014-10-14

Creating an iPhone-optimized version of this site

▼ I've been pretty happy with the redesign of this website that I did last year. It's clean but still has some character. I kept the width of the text to 540 pixels because long lines are hard to read, and the text size is a bit larger than average. Still, on the iPhone the site looks tiny:

Double tapping to zoom in on the text helps a bit, but it's still not really readable. So I decided to make an iPhone-optimized version of this site. As often in the web world, it turned out there's a bunch of tools to help with this, but it's still not an easy thing to do!

Safari on the iPhone renders websites as if they're displayed on a window that's 980 pixels wide. However, this can be adjusted by using a "viewport" setting. I use a viewport of 568 pixels. That's the width of the iPhone 5 screen in landscape. (Well, it's 1136 really but iOS quietly doubles all pixel sizes behind the scenes and 568 x 2 = 1136.)

The result is that it's no longer necessary to double tap to zoom, as Safari now renders the site as if it's displayed on a window that's 568 pixels wide. However, on devices such as iPads everything would get too big, so I also use a maximum-scale setting of 1.0. This means that if the screen is actually narrower than 568 pixels, the device will shrink the page to fit. But if the screen is wider than 568 pixels (such as 768 pixels for an iPad in portrait orientation), the page won't be zoomed.

All of the above is accomplished using the following line in the HTML of every page:

<meta name="viewport" content="width=568, maximum-scale=1.0">

But even with these settings in place, in portrait mode, the text is just too small to read. Turns out you can use "media queries" to adjust your CSS stylesheet when the page is shown on certain devices. The styles below are applied only when displaying on a screen if that screen is no more than 360 pixels wide and in portrait orientation. (The iPhone 5's screen is 320 pixels wide before doubling.)

<style type="text/css" media="only screen and (max-device-width: 360px) and (orientation: portrait)">
  body { font-size: 200%; }
</style>

So on these small screens the font size gets doubled. On devices with screens between 361 and 720 pixels, such as the iPhone 6 at 375 and the 6+ at 414 pixels, the font size is 60% bigger than the standard size:

<style type="text/css" media="only screen and (min-device-width: 361px) and (max-device-width: 720px) and (orientation: portrait)">
  body { font-size: 160%; }
</style>

(There are some more settings for headings and footers.)

This worked for the most part, except that the body text was actually slightly larger than the heading text. Turns out that mobile Safari automatically bumps up font sizes if it thinks those are too small. As I'm trying to control all of this myself I turned this off with the following CSS setting in the body section of the main stylesheet:

-webkit-text-size-adjust: 100%;

The result:

Nice and readable, even if I do say so myself. Turns out that the font size I used is exactly the same as the default one in the "reader" mode in Safari. And when you turn the phone sideways, in landscape you get the exact same view you get with the desktop version of Safari:

Compare the image above to that first paragraph:

After six years of using one, it's not often that I stumble upon an iPhone trick I didn't know before. But here's one. When looking at the lock screen when audio is playing, press the home button and the audio controls are replaced with the large clock that the lock screen normally shows when no audio is playing.

Looking good or not so good on your system? Send me a screenshot!

Permalink - posted 2014-10-01