Camera Shutter Speed

[Phil Kulak]

I've noticed from EXIF data on some of the early samples that, despite having OIS, the camera never takes the shutter below 1/22. 1/10th should really be about where it is with 3-stop OIS comparable to the Nokia 920, and would really improve low light performance. Anyone have any ideas on this?

 

[Matthew Merkle]

I've noticed from EXIF data on some of the early samples that, despite having OIS, the camera never takes the shutter below 1/22. 1/10th should really be about where it is with 3-stop OIS comparable to the Nokia 920, and would really improve low light performance. Anyone have any ideas on this?

Considering a seriously steady hand and DSLR-level stabilization usually isn't worth a crap below 1/20-1/15, I doubt this. 1/10 of a second is a really long time for motion.

 

[Phil Kulak]

Considering a seriously steady hand and DSLR-level stabilization usually isn't worth a crap below 1/20-1/15, I doubt this. 1/10 of a second is a really long time for motion.

My S3 takes sharp photos at 1/30 and will drop to 1/15 (regrettably) in low light to take blurry, POS shots. This makes sense since the focal length in these things is 28mm and the rule of thumb is that you want a shutter speed of at least 1/focal length. Nokia advertises 3-stop OIS and Canon's lenses are 4-stop. 3-stops below 1/30 is almost 1/3, so 1/10 should not be out of the question (as long as your subject isn't moving).

 

[Matthew Merkle]

As usual, what they advertise and what is achievable are vastly different things. Places like DPreview back that up.

 

[BigDinCA]

I'm no Ansel Adams, but I wonder if it doesn't need to leave the shutter open that long? With the larger pixel size, a similar shutter time will produce much brighter pictures than the S3, maybe too bright. If your S3 can't do it in 1/15 I would bet the One can do it fine in 1/22 or 1/30 with a pixel that is 3X the size. The more light I can get more in, the sooner I can close the shutter. I know that's over-simplifying it, but that's the easy guess.