Serene Life

by garik

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

I use a Sony NEX-5R, and my single biggest problem with the camera is that it often mis-focuses under low light. I shoot a lot of low-light photos, and I work around it by using manual focus.

I have been told that cameras with larger sensors are more likely to mis-focus under low light, because of the physics / optics. Is that true?

I notice that my iPhone almost never mis-focuses no matter how dark, while my NEX does so often enough for it to be a problem.

And reviews say that even Sony's and Fujifilm's top-of-the-line APS-C cameras (A6000 and X-T1, respectively) don't focus as well as micro-four-thirds cameras. For example:

[...] As the light levels became lower the A6000 (like the XT1) slowed down and in dim conditions often struggled to lock-on, whereas the latest Olympus and Panasonic models continued to autofocus successfully even in very low light. For comparison, Sony quotes the A6000 AF works down to 0 EV whereas the latest Panasonics will still operate at -4 EV [...]

So, is it correct to say that larger sensors mis-focus more often under low light?

If so, the conclusion for me would be that full-frame would not be an upgrade over the NEX, since it does not address my biggest problem with the NEX, and is likely to make it worse. Yes, it would work better at higher ISO, but it would focus worse, which is one step forward and one step back -- I don't want that tradeoff.

Addendum: I'm not worried about the speed of focus -- I can certainly wait a second for it to focus. What I care about is what fraction of photos end up mis-focused.

share|improve this question
    
When comparing your iPhone to your APS-C camera, are you comparing shots with equal depth of field? I'm guessing "no", which makes your experiment mostly bogus as you're not using the same conditions in both cases. –  Philip Kendall Jun 8 at 6:46
1  
Further, this post seems to be a combination of hearsay ("I've been told that" without any references) and unwarranted extrapolation (the A6000 doesn't autofocus as well as the E-M1, therefore the problem is the sensor size). –  Philip Kendall Jun 8 at 6:50
1  
That's the whole reason to post this question -- to confirm what I was told. If I was sure it's reliable, there would be no need to post this question. It's not unwarranted extrapolation -- both the APS-C sensor cameras (Sony A6000 and XT1) focus worse than the small sensor cameras (micro four-thirds and iPhone). If you have a better explanation for this than sensor size, I'd like to hear it as an answer. It doesn't make sense to criticize it as unwarranted extrapolation. –  Kartick Vaddadi Jun 8 at 7:13
    
Your guess is wrong. I have shot exactly the same scene from the same point with both the iPhone and the NEX, and the iPhone autofocused, while the NEX didn't. And it's not just one scene -- the iPhone almost always focuses correctly, as I said in the question, which you seem to have missed. –  Kartick Vaddadi Jun 8 at 7:15
1  
What are you shooting in the dark that AF performance overrides all the other factors that make a full-frame camera better than the alternatives? –  ElendilTheTall Jun 8 at 9:37

2 Answers 2

up vote 3 down vote accepted

I think the key thing for here is that smaller sensors inherently have more depth of field at the same aperture number and framing. That means that with a larger sensor, focus is more critical. It's not that the larger sensor is really worse. With a camera phone, in low light, the result will be very noisy (with automatic, non-optional noise reduction smearing out the details), and usually with a very long shutter speed causing motion blur. That further masks any focus problems (and of course makes the result less nice). The larger camera's larger sensor lets you use a much larger physical aperture with the same perspective and framing, which means more real total light in — an inherent advantage, but with DoF as the price. (TANSTAAFL.)

The ratio between the width of the iPhone 5S sensor and the one used in your NEX is about 1:5. And the iPhone shoots at a fixed f/2.2. That means that for the same framing and viewing size, you need to set your NEX lens to f/11 to get the same DoF, and that's not even considering the other blur factors (blur from other sources tends to increase the perception of greater DoF).

Correspondingly, in your trials, the iPhone is probably trying it's best and saying "eh, good enough", while the NEX knows it isn't. Even if you do stop down, the NEX probably does not take that into account — you could file this with the other answers under the heading "cameras: not very intelligent", but as a practical matter, in low light situations you are likely to either use a wide aperture or you have time to focus manually.

And if the entire scene is within the hyperfocal distance, that's exactly what you should do: focus manually and take the shot — another thing you can't do with a phone.

I don't think the same factor is at fault in comparing various APS-C mirror less cameras to Micro Four Thirds (although it may have a slight impact). The sensor sizes are in the same ballpark, and there are APS-C DSLRs (like the Pentax K-3) which are rated into the negative EOS. I think here its just a matter of the various companies' technology, and tha will continue to be a area of intense competition over the coming years.

share|improve this answer
3  
As a general comment on upgrading to a full-frame camera: from your questions, you really want the camera to apply a lot of computer power to doing the work for you. There's nothing wrong with that, but you will find that as you go up the spectrum to more expensive bodies, more and more knowledge is assumed, along with the assumption that you will want more manual control (and all this auto stuff just gets in the way). So, with that in mind, I don't think you do want a full-frame camera. –  mattdm Jun 8 at 10:41
    
Thanks very much, mattdm -- this is all very useful. Regarding DoF, I have the mis-focus problem even when the lens is wide-open (f/2.8), either when set that way in aperture-priority mode, or when the camera chooses to open the lens fully, in program mode. I'm not asking the camera to use a particular DoF and then complaining that it can't handle that. I just wish that it picked SOME DoF and then takes a usable photo with that DoF rather than a blurry mess that goes straight to Trash. –  Kartick Vaddadi Jun 8 at 14:58
1  
Regardless of what aperture is selected for taking the shot, the camera meters and focuses with the aperture wide open. –  Michael Clark Jun 8 at 17:47
    
Oh, yes, thanks for reminding me of that. Does ISO affect focus accuracy? If I set the ISO too low (given a fixed shutter speed and aperture), am I more likely to end with mis-focused images, because there isn't enough light in the image for the autofocus to work? –  Kartick Vaddadi Jun 9 at 4:27
    
What's "TANSTAAFL"?? –  BBking Jun 9 at 7:00

No, larger sensor cameras are not more likely to mis-focus - if you take the Canon 1DX (with a modern lens) for example, it's a full-frame camera that's about as far away from "likely to mis-focus" as possible.

But when a large sensor camera mis-focuses it's more noticeable, especially when most tiny sensor cameras (cellphones) have wide angle lenses.

The thing is that a small sensor paired with a wide angle lens has extremely wide depth of field, the iPhone for example, in it's normal mode (not extreme closeup), always has everything from a few inches from the lens all the way to infinity in sharp focus - it's not that the iPhone is that good in low light focus, it's just that it doesn't have to be.

share|improve this answer
    
Fair enough, so leaving the iPhone aside, would you say that when an APS-C camera mis-focuses, it's more likely to be noticeable than a micro four-thirds camera mis-focusing? I hear what you say, but to me the camera mis-focusing and the result of the mis-focus being more noticeable are the same thing at the end of the day. –  Kartick Vaddadi Jun 8 at 14:27
4  
APS-C and µ4/3 are much closer in size than either compared to the tiny iPhone sensor. The conventional wisdom that the µ4/3's higher DoF would hide more focusing errors than the APS-C camera could be more than offset by the specific implementation of CDAF in each case. A good APS-C system could very well outperform a decent µ4/3 system. –  Michael Clark Jun 8 at 17:57

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.