iPod mini battery tests | iLounge Article

Article

iPod mini battery tests

In order to meet our publication deadline, we left only one issue unresolved when publishing our recent reviews of Apple

« Final: iPod mini battery tests

Taking apart the iPod mini »

Related Stories

Comments

1

Sounds to me like the random playback on the mini is not queuing up the random files in the cache, but is randomly picking songs, only the first of which will be in the cache.  The 3G iPod is nearly identical for both tests, which suggests that it is making more efficient use of the cache.

I wonder if the software update to the iPod mini addresses this?

Posted by Dennis Lloyd in Irvine, CA on February 26, 2004 at 8:45 PM (CST)

1

Not only is it not reading ahead on the random play list to save battery, but if its filling up its 32 MB buffer for antiskip protection, its wasting on average 28 MB worth of disk access every time it goes to the next song.

No wonder the battery life was so bad.  Sounds like its time for a software upgrade.

Posted by Dennis Lloyd in Irvine, CA on February 26, 2004 at 10:24 PM (CST)

1

Scottman + Mcchris - we should have included a link to the original iPod mini Power User review on this point. It actually doesn’t seem like either one of those is the case; as we noted in the Power User review, the mini is definitely using the cache for shuffle play also. It’s not accessing the hard drive after every song, that’s for sure.

Perhaps the shuffle play is not making the hard drive get accessed more often, but when it does get accessed, the mini has to search more on the hard disk for the songs it’s going to pull. Or perhaps there’s something else that accounts for the disparities. We’re still investigating, and test #7 (another iPod mini test on shuffle play) is running right now with again surprisingly different results.

Posted by Dennis Lloyd in Irvine, CA on February 26, 2004 at 11:10 PM (CST)

1

Actually your Type-A is exactly how i use my iPod.

I have almost never any interaction with it, since I use smart-playlists and SoundCheck.

That’s probably why I am still satisfied with my 1G 5GB iPod. With a good smartplaylist I don’t need bigger HD capacity or an extraordinary battery.

Posted by Dennis Lloyd in Irvine, CA on February 26, 2004 at 11:54 PM (CST)

1

mcchris wrote:
“Not only is it not reading ahead on the random play list to save battery, but if its filling up its 32 MB buffer for antiskip protection, its wasting on average 28 MB worth of disk access every time it goes to the next song.”

I am kind of suspecting that the iPod and iTunes are unable to use the buffer (streaming or antiskip) in between tracks:

If you use iTunes on a remote computer you can see that when a new song starts (random or not) there is a pause and a surge in network activity as the new songs is loaded to the buffer.

Very weird that Apple did not include something better there.

Posted by Dennis Lloyd in Irvine, CA on February 26, 2004 at 11:59 PM (CST)

1

Note that there is virtually on difference between the two 3G iPod tests, which is how it should be, because in test A, the hard drive spins up, loads up the first 32MB of songs on to the RAM (which is a preset order that the iPod ramdomly selects as soon as you hit play on the playlist), plays them, and then repeats the process. In test B, the same should happen, except it plays them in alphabetical order. The mini is doing something very different, so Apple needs to fix the firmware fast I would say.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 2:32 AM (CST)

1

Battery life of my iPod mini generally exceeds that of my 3G iPod.

I think some people like to look for trouble where none exists and then when they don’t find it, they make something up!

Hey folks, while you’re at it, why don’t you perform the Type-C and Type-D tests!  Maybe they’ll negate your Type-A and Type-B tests.  If that is the case, you may need to perform a Type-E test.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 5:58 AM (CST)

1

The recent software update on the mini was simply the “restore” function made available for those who didn’t boot up the CD enclosed with the device.  I didn’t when I first downloaded iTunes to the mini.  My iMac recognized the mini and had no trouble downloading.
Batteries need to be conditioned before they reach their optimal charge.  Use your mini for a while then test the playback time.  Apple says “up to 8 hours” not “guaranteed 8 hours”.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 6:45 AM (CST)

1

Jeremy,

Just to clarify: my tests of the battery life of the iPod Mini, as published in my Wall Street Journal column of Feb. 11, did indeed involve the use of shuffle mode during playback of the entire library. Also, I had the volume set at about 75% and was using an EQ setting—Vocal Booster. So my test was close to your type A test than to your type B. And I performed mine on two different Minis.


Walt Mossberg

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 7:00 AM (CST)

1

There will always be an inverse correlation with battery lifespan and bitrate. The higher the bitrate and quality, the lesse the battery charge lifespan.

I think you need to use some graphs, and plot time vs bitrate (right up to uncompress AIFFs) for random play until exhauustion.

These graphs would look a bit like CPU utilisation against multiple resolutions, as plotted for some CPU benchmarks.

That way we get to see complete profiles of the iPod’s battery behavior for different playback scenarios.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 9:17 AM (CST)

1

I’ve had my mini for a week now, and I’m getting very nervous about the battery situation.  My normal usage is similar to your type-B test.  Last night I left the mini to charge thru the AC adapter.  This morning the batt indicator said “charged”.  I discnnected and started listning to one album with no shuffle and no EQ, and about 20 minutes later, my batt indicator showed only 70% full! to me that is unacceplable for a portable device that has been charged over night.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 10:54 AM (CST)

1

I’ve found that the battery indicator usually isn’t a good indicator at all.  I have shown empty battery quite a few times and still gotten two or more hours of play.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 11:04 AM (CST)

1

What bitrate was used btw? Results should differ between 128 and higher bitrates, since those are usually bigger in size, therefore fill the cache faster - and result in a more active HD and more power consumption.

Don’t think 128 kb mp3s are still standart.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 11:26 AM (CST)

1

There are known problems with the battery indicator.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 12:06 PM (CST)

1

It looks to me like the battery monitoring circuit tracks battery load over interval and updates the display accordingly.  It, consequently, behaves more like a time-remaining indicator (in the form of a battery icon).

If I advance several times through a shuffle, I’ll note the battery image ‘draining’ at a startling rate.  As soon as I settle onto a song a want to hear, the battery image recovers somewhat.  I’ve noted shifts from 90%...then dropping to 25% during my indecision…and then returning to near 90% once I’ve settled.

I’ve decided to view the battery icon as a “Estimated time remaining based on current interval usage (of ~8hrs total time)”

:)

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 12:52 PM (CST)

1

To go along with the higher bit rate leading to lower battery life theory one must also consider that long songs would also fill the buffer faster and cause more hd access leading to lower battery life.  I think that if the test is going to be fair you have to make sure that all the songs from one test to the next are the same file size.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 12:53 PM (CST)

1

The “higher bitrate - shorter life” is not a theory. I have tested it myself.

It’s just common sense. Spinning up and down the hard drive is the most battery intensive thing the iPod does. Also, the PoirtalPlayer chip can ratchet its operating frequency from 32KHz to 90MHz on-demand. This is part of its power-reduction strategy. More data to be decompressed because of higher bitrates requires higher clock speed.

Finally, you can quickly verify the compression/battery relationship yourself. Make a few AIFFs. Play them back. This demands almost continuous disk access by the iPod based on zero compression. You will be astonished at how quickly your battery indicator drops.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 1:02 PM (CST)

1

This may be a moot point. But what percentage of the hard disk space is occupied on the iPods? Hard disk performance in general takes a hit when they are reaching 100% capacity. Is there more unused disk space on the 3G? This may account for needing more spin cycles to access the needed data in the mini when the hard drive is being read.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 1:25 PM (CST)

1

I’ve noticed something strange with my mini. When I charge it via the AC adaptor, the large “charging” icon eventually fills up, but the little battery icon never fills up. the highest I’ve ever seen it is 80-90%

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 1:40 PM (CST)

1

I know it’s for an Archos and not an iPod, but I can’t find any equally exhaustive analysis people have done on mp3 player power consumption. Outside of NDA stuff, of course:

      HDD off, backlight off, idle 94 mA
      HDD off, backlight off, play 97 mA
      HDD off, backlight on, idle 129 mA
      HDD off, backlight on, play 131 mA
      HDD on, backlight on, play 230 mA
      HD on, reading, backlight off ~ 600 mA
      HD spin up before read max 800 mA

Conclusion:

Using a disk sucks around 6 times as much power as the CPU. Now, the iPod’s disk is smaller, so it has lower inertia and power consumption, so maybe the factor is reduced to 3:1 or 4:1. The disk power consumption on the iPod Mini will be even smaller, so the ratio is reduced further. Then again, the 4GB Hitachi uses regular atmospheric pressure to keep its disk aligned, whereas the 1.8” and 2.5” disks used in regular iPods and Archoses are vacuum sealed.

Posted by Dennis Lloyd in Irvine, CA on February 27, 2004 at 1:41 PM (CST)

Sign up for the iLounge Weekly Newsletter

Email:

Recent News

Recent Reviews

Recent Articles

Sign up for the iLounge Weekly Newsletter

Email:

iLounge is an independent resource for all things iPod, iPhone, iPad, and beyond.
iPod, iPhone, iPad, iTunes, Apple TV, Mac, and the Apple logo are trademarks of Apple Inc.
iLounge is © 2001 - 2014 iLounge, Inc. All Rights Reserved. Terms of Use | Privacy Policy