Disappointing battery life on 2011 Core i5 11" air...

[elfenlied102938]

I have the 2011 Macbook Air 11" Core i5 1.6GHz 4GB RAM machine, and I'm slightly disappointed at the battery life.

Anandtech is reporting 5+ hours (380min) of light, casual browsing, but it seems I get much less.

My test setup:
Using suasua.com to refresh 4 tabs of Safari on gizmodo.com (front page, with flash), 50% LCD brightness, lowest illuminated keyboard setting. All apps closed except for Safari browser. 100% starting battery. This test is arguably lighter than Anandtech's "Light Web Browsing" (which includes iTunes on repeat playing in the background, but flash is turned off)

After 30 minutes, the battery read "3:37 remaining" and "87% full". This translates to ~4 hours of battery life, more than 1 hour less than what Anandtech is seeing.

I ran the same test with Chrome browser. After 30 minutes, the battery read "2:16 remaining" and "84% full". Interestingly that's over 1 hour of battery life difference but the actual percentage difference is only 3%...Unfortunately I prefer Chrome to Safari, so I may have to live with this result.

Can anyone run the same test as mine and report back the results? It'll only take 30 min Experimenting with/without flash might also be interesting.

P.S. If I actively "take control" of the computer (moving mouse, typing, changing screen, etc), the battery drain is faster than just leaving it alone running the automatic suasua refreshes...

 

[PaulWog]

I called it! People were thumbing down my post about how "Andandtech is misleading people with their review", but it's true. You were mislead.

Alright, here's the thing: Andandtech is plain incorrect. They have reported false numbers... obviously accidental. The 11-inch Macbook Air sports a 35 Watt-Hour battery. The 13-inch Air sports a 50 Watt-Hour battery. That means the 13-inch Air (aside from having a larger screen) should be achieving higher battery life. In Andandtech, however, only reports a 36 minute difference in battery life... a 36 minute difference in battery life between a 50 watt-hour battery and a 35 watt-hour battery? I think not.

Anyways, don't bother wasting your time expecting the numbers andandtech has set out. You won't be able to reproduce those numbers even if you run a 20-second refresh loop on websites without flash, and run iTunes music on repeat. Believe me, you simply won't.

Anyways... more on-topic to what you're looking for:

I have done similar tests before. Here is how I get 5 hours average: Run screen at 50% brightness (give or take depending on if you want a slightly better battery life or slightly worse). In my case let's just say 50% brightness, as that's what I ran. Do that with no keyboard light on. Have flashblocker on. Only use Safari with no other programs in the background (other than iStat which I had running as a wigit). Spend 30 minutes just browsing and posting on Macrumors.com (I refresh the page relatively often... probably once every 15 seconds on average give or take, play with the touchpad sometimes, and type relatively fast). I get about 20% battery life lost per 1 hour, which equates to 5 hours battery life. In fact, most times I've tested the battery life like this, I come out with 18-19% used up, which leaves room for a bit of youtubing (a few minutes per hour) in the mix while still maintaining a 5 hour battery life.

I figure if I *really* wanted to, if I were browsing super lightly (not sure why I'd need to), I could get 5 hours and 30 minutes-ish out of the Air with 30% brightness and very very very light browsing.

On the other hand, playing Angry Birds with the sound on at about 30%, and the brightness on at about 40%, I get about 3 hours and 40 minutes worth of battery life by testing 1 hour and looking at the %. That could obviously fluctuate up or down, since I'm guessing Angry Birds isn't going to take up the same amount of battery life every time, depending on how it's played.

 

[abhibeckert]

Mate. You installed flash.

There is a reason macs no-longer ship with flash pre-installed: it destroys the battery life.

Uninstall flash, or install a flash blocker, and then you'll see dramatic improvements. Modern CPU's keep power consumption down by under clocking the CPU while it's idle. With flash placing steady demands on the CPU you're disabling that feature.

 

[PaulWog]

Mate. You installed flash.

There is a reason macs no-longer ship with flash pre-installed: it destroys the battery life.

Uninstall flash, or install a flash blocker, and then you'll get dramatically better battery life.

My Macbook Air 11-inch runs flash straight out of the box. I didn't install anything. ; I installed flashblocker though.

And I hope you aren't implying that Andandtech's numbers (6 hours and 20 minutes) is anywhere near a realistic number for an 11-inch Air. >_<;

 

[abhibeckert]

Perhaps your browser installed flash behind your back? Are you using Chrome?

Ars Technica only tested the 13" model, and they got between 10 hours 42 minutes, and 5 hours 30 minutes of battery life, depending on usage pattern.

Ars also commented "After a while, it became obvious that even things that use only a few percent of CPU time in the background, such as iTunes, have a significant impact on battery life". In other words, if you allow flash banner ads to loop endlessly, your battery life will be closer to 5 hours than 10 hours (for the 13" model).

So yes, I do think the numbers from anandtech are plausible. I can't verify them, as I don't have an 11" air. But they look ok for me, considering they didn't install flash.

Battery life is highly specific anyway, you can't trust any figures except the ones you get under your own usage patterns. So it doesn't really matter what the specific numbers are when someone does an artificial test.

 

[PaulWog]

Perhaps your browser installed flash behind your back? Are you using Chrome?

Ars Technica only tested the 13" model, and they got between 10 hours 42 minutes, and 5 hours 30 minutes of battery life, depending on usage pattern.

Ars also commented "After a while, it became obvious that even things that use only a few percent of CPU time in the background, such as iTunes, have a significant impact on battery life". In other words, if you allow flash banner ads to loop endlessly, your battery life will be closer to 5 hours than 10 hours (for the 13" model).

So yes, I do think the numbers from anandtech are plausible. I can't verify them, as I don't have an 11" air. But they look ok for me, considering they didn't install flash.

Battery life is highly specific anyway, you can't trust any figures except the ones you get under your own usage patterns. So it doesn't really matter what the specific numbers are when someone does an artificial test.

Safari. Everything straight out of the box. Perhaps it did, I don't know. I know out of the box I didn't have to do anything for it to play flash though.

 

[MRU]

^ My MBA 2011 received yesterday did not play Flash out of the box, when visiting a site it instantly told me I needed to download Flash Player 10.

I downloaded Flash Player 11 Beta instead (shortly followed by the Click2Flash extension in Safari).

My test setup:
Using suasua.com to refresh 4 tabs of Safari on gizmodo.com (front page, with flash), 50% LCD brightness, lowest illuminated keyboard setting. All apps closed except for Safari browser. 100% starting battery. This test is arguably lighter than Anandtech's "Light Web Browsing" (which includes iTunes on repeat playing in the background, but flash is turned off)

After 30 minutes, the battery read "3:37 remaining" and "87% full". This translates to ~4 hours of battery life, more than 1 hour less than what Anandtech is seeing.

So you didn't actually run your battery down to the bottom at all - you are relying on the OSX battery indicator/guestimator to tell you your remaining battery life, which is always fairly hit and miss - especially on new machines where the battery has not been run in for a while.

So your presumption is based on 30 minutes of use and is about as scientific as Anandtech's, which you are complaining about.


-------


I timed mine exactly yesterday (an 11" i7) - fully charged to 100% and then ran down to battery low message before plugging in.

I got 4hr 28mins with brightness at 40%, Bluetooth off, Wifi On, iPhoto, iWeb updating a blog, Office 2011 (both word and excel), Twitter & Safari running. I had few tabs open in Safari but mostly sites like this and other text heavy sites and Facebook rather than flash embedded sites.

It was pretty much on Par with the MBA 2010 1.4 which I under clocked with coolbook to improve battery.

Timed my use with my phones stop-watch rather than relying on the battery indicator/guestimator



Whilst that is certainly less than anandtech - it's better that I would have thought.

 

[Abstract]

Either run the test under the same conditions, or don't expect the same results.



And speaking of results: Use a watch.

The only thing you have proved with your experiment is that you don't know how to run a proper experiment.

 

[elfenlied102938]

Either run the test under the same conditions, or don't expect the same results.



And speaking of results: Use a watch.

The only thing you have proved with your experiment is that you don't know how to run a proper experiment.

It seems Anandtech actually ran the test until the battery was completely drained. I didn't want to babysit my computer for 4-6 hours, twice (safari vs chrome).

I did use my iPhone alarm set to 30 min.

Please don't insult other people. I did include a detailed description of what I did in the original post, which anyone can duplicate. Reread my setup - I'm not sure how more specific I can get. This was the point - to see if other people can duplicate my test and compare results.

Perhaps I can't rely on extrapolating data from a 30min test (this assumes battery consumption is linear), but at least it is easy and quick. My other goal was to compare Safari vs Chrome. If I disable Flash on both, I may be able to squeeze 6 hours on Safari. I'll try this experiment later today. I'll also try lightly browsing from 100% to 0% this weekend to see how much I get.

Thanks everyone for the feedback.

P.S. I was hoping to use Chrome as my main browser, but it looks like it's taking quite a toll on battery life

 

[VMMan]

I got about 6 hours with my 2011 MBA 11" i5 with light web browsing.

1 Safari Window with 4-5 tabs
NO Flash (never installed it)
NO Bluetooth
WiFi On
Brightness 3 bars
NO keyboard backlighting

I didn't do this test for a few days, to make sure any indexing was already done.

I'm very pleasantly surprised by how good the battery life is so far.

 

[The Expert]

Try turning off bluetooth, that should give you approximately 30 minutes.
Try lowering the brightness which is the leading cause for draining battery life
Try switching to Firefox and use No script and ad-block plus.

 

[komoornik]

It seems Anandtech actually ran the test until the battery was completely drained. I didn't want to babysit my computer for 4-6 hours, twice (safari vs chrome).

I did use my iPhone alarm set to 30 min.

Please don't insult other people. I did include a detailed description of what I did in the original post, which anyone can duplicate. Reread my setup - I'm not sure how more specific I can get. This was the point - to see if other people can duplicate my test and compare results.

Perhaps I can't rely on extrapolating data from a 30min test (this assumes battery consumption is linear), but at least it is easy and quick. My other goal was to compare Safari vs Chrome. If I disable Flash on both, I may be able to squeeze 6 hours on Safari. I'll try this experiment later today. I'll also try lightly browsing from 100% to 0% this weekend to see how much I get.

Thanks everyone for the feedback.

P.S. I was hoping to use Chrome as my main browser, but it looks like it's taking quite a toll on battery life

Sorry, but I can only agree with others.

Your tests proves absolutely nothing.

Discharge the whole power from batteries and then you can complain about what times you get.

 

[Azathoth]

After 30 minutes, the battery read "3:37 remaining" and "87% full". This translates to ~4 hours of battery life, more than 1 hour less than what Anandtech is seeing.

I ran the same test with Chrome browser. After 30 minutes, the battery read "2:16 remaining" and "84% full". Interestingly that's over 1 hour of battery life difference but the actual percentage difference is only 3%...Unfortunately I prefer Chrome to Safari, so I may have to live with this result.

At least you're trying to do a meaningful test (seeing as 90% of the posters on this forum dont have a clue), but people need to understand that this "percentage" is +/- a few percent error (deviation from linear line) *at least*, and likely to be most at the top end and the bottom end of the battery discharge curve.

A meaningful test would be at least 20% discharge from between 20% and 90%, preferably repeated 2-3 times to show variability

 

[VMMan]

Ambient temperature, which in turn may affect the laptop temperature could also be a variable.

 

[bill-p]

I think it's important to drive home a pain point: Lion sucks battery life. Dramatically so.

I tried to simulate a healthier situation on my 2010 MBA 13": 4 tabs, refresh every 1, 4, 8, and 16 minutes each, Safari, battery from 80% to 50%. Result:

Lion w/ Flash: 1:45 remaining.
Lion w/o Flash: 3:15 remaining.
Snow Leopard w/ Flash: 2:48 remaining.
Snow Leopard w/o Flash: 4:45 remaining.

This was conducted all yesterday evening and throughout the night. The last test was set up before I went to sleep with the battery at 100%.

 

[jackyyeow]

1 Safari Window with 4-5 tabs
NO Flash (never installed it)
NO Bluetooth
WiFi On
Brightness 3 bars
NO keyboard backlighting

Uh 3 bars my eyes will go bad pretty quickly. It's not good to try to read things on insufficient lightings no?

Great battery life nonetheless!

 

[VMMan]

Uh 3 bars my eyes will go bad pretty quickly. It's not good to try to read things on insufficient lightings no?

Great battery life nonetheless!

Ourdoors, I have to usually go to 5 bars, but indoors, I find 3 bars actually pleasant.

If the room is dim (e.g. single lamp in the evening), I actually find 1-2 bars acceptable to okay.

I'm not sure why, but it seems that a lot of forum members need 50% to 100% brightness for acceptable viewing...

 

[elfenlied102938]

Sorry, but I can only agree with others.

Your tests proves absolutely nothing.

Discharge the whole power from batteries and then you can complain about what times you get.

Don't be rude. I posted to this forum to see if anyone could duplicate my test (with clear instructions) and see if they also get similar results. I wanted to know if it's just mine or this is normal behavior.

I agree with the fact that I should run the test by fully discharging it (Li-ion battery is nonlinear) , but I as I stated before, I didn't want to babysit my machine for hours. I will probably do this in the weekend.

Besides, for comparison's sake the test is still meaningful if you control the variables (start at 100%, run same procedures) across different users machines. I chose to start at 100% since it's a convenient (and well-controlled) starting point for both safari and chrome.

By the way, I do notice there's a huge difference between just letting it sit and run the refresh script via suasua.com and actually using it to browse websites. This is why I'm more worried about the "real life" batter life.

 

[johannnn]

a 36 minute difference in battery life between a 50 watt-hour battery and a 35 watt-hour battery? I think not.

13" screen requires more power than an 11" screen. I don't say a 36 minute difference is correct, but don't make stupid assumptions.

 

[freeman727]

Don't forget Mail app. It searches for mail in the background. If you want the absolute best battery life close the app.

 

[iViking]

Flash is the culprit

This is a layman's confirmation that Adobe Flash is the culprit.

I have had 2 of the 2011 11". The first one I installed Flash on. Battery life wasn't so great. I had to return it for a reason that is unrelated. When I got the second one, I did not install Flash.

My battery life has since been roughly 30-35% better.

Flash really does suck battery life on these things.

 

[kilonet]

How do you guys watch YouTube videos without flash?

 

[wickoo]

Try http://www.youtube.com/html5