iPhone 4S produces more accurate colors. Overall, I'm delighted how great the image quality if for all cell phones; astonishing how far we've come in just a few years.
There's no absolute color. Color is what we humans see when photos bounce off objects. The spectrum of light bouncing off isn't absolute. You have no idea what those objects really look like. The lighting in the room matters. I have both CFL and old fashioned bulbs in my house. If I took a photo of something you knew in both rooms you'd be unable to tell me which is more accurate, because only I know the lighting in my rooms and you don't.
You may be able to say "this photo looks right" but that's pretty darn subjective. Turns out we humans are kinda useless for tests like these. Past a few megapixels this stuff really doesn't matter because things like focus, lenses, and lighting are so much more important than resolution. The photos are "sharp" enough and none of these phones support mounting lenses or lights on anything. So its kind of arbitrary. Unless one of these phones has an especially shitty lens or focus or bad software, they'll probably just look mostly the same. I consider these long solved issues and don't expect much difference.
Definitely not an Android issue here, as I know for a fact Samsung uses its own camera app and capture framework. Not sure about the underlying hardware. If it's a v4l2 driver you might have a case against the kernel infrastructure.
While not indicative of real-world use, comparison shots like this, IMHO, should be done with a stand... unless the only variable being tested is "likelihood of shakiness due to camera(phone) form factor."
I generally agree with you, except in this case one of the specs they touted in the iPhone announcement event was capture speed. It captures much faster than the iPhone 4, and would be a good explanation as to why the 4s appears less shaky.
I don't think capture speed has any effect on the sharpness of the image, sharpness is more down to the lens/shutter speed, I'm assuming the "speed" touted by Apple refers to the startup time of the camera app as well as the responsiveness of the UI and the ability to take consecutive shots faster.
Looking at the shots again, it does seem that the sharpest image, that of the original iPhone 4, is also the noisiest, iPhone4 autoexposure set it to high-iso/fast shutter I suppose?
Yes, that's definitely the cause here. The iPhone 4 Picture is much darker, meaning it was made with a higher shutter speed and therefore is much shaper when holding it by hand. If you have a good camera lying around you can try it out by playing with iso/aperture/shutter-speed: Make two pictures in a dark(ish) place, one with low speed and low iso, the other with high speed and high iso (but same aperture). Depending on how dark it actually is and how steady your hands are, you'll see a big difference in sharpness.
The other thing are the colors, which seem much better in Samsung's picture. This, too, has nothing per se to do with the quality of the camera but it's settings. This time it's the white balance that's different. This is one of the things, where camera automatics are still often wrong. A small change of the subject can cause big differences. Moving a few centimeters more to the left is enough...
Again, you can try that with a "real" camera. A DSLR will let you choose the color temperature, but even cheap digital cameras have sometimes different modi for "sunlight"/"cloudy"/"artificial light". I won't go into details here (see [1]), sufficient to say that this can greatly affect a picture.
Also note that the picture from the Galayx II and the one from the iPhone 4S have roughly the same brightness and sharpness. Judging from the noise and resolution, the two cameras seem quite similiar (and are maybe even produced by the same factory).
TL;DR
This comparison is not meaningful at all. For a professional result you would need to control the cameras completely manual and take a lot more pictures in different settings (light, reference colors, tripod, ...).
>For a professional result you would need to control the cameras completely manual and take a lot more pictures in different settings (light, reference colors, tripod, ...).
Well, it's also worthwhile testing how well the automatic settings work -- that's how most folk will use the phone.
(All I really want in a camera phone is non-shitty low-light photos.)
I believe this is called the "Robert Scoble"-effect. One could liken it to social media spam: I don't want to know Robert Scoble, don't find him notable, yet his name appears everywhere.
Such people know how to insert themselves in the daily conversations and can count on the support of their followers to carry their views far and beyond. Most of the time those views are not much more than a catchy social hook. You Googling his name and reading this reply will likely cement Robert Scoble in your brain for a while longer.
I find it an interesting, albeit at times annoying, phenomenon. Like the Quora drama (in which Scoble was able to permeate into a lot of topics):
They are focused on different points. The Galaxy is focused closer than the 4S; the details on the Belkin card reader (in particular, the edge contacts in the slots) are clearer on the Galaxy, and the difference is even more marked on the keyboard letters. The Galaxy also has a higher exposure. On the book paper, the lack of focus combined with overexposure has causing some bleed into the type; but it shows better detail in shadow areas.
Much of a muchness, IMO. The two are very close, and aside from things like color temperature (which I personally would discount and tweak later if desired - though I wonder if the white balance is configurable on either phone), they are all but equal.
This a dreadful comparison. There is nothing to show depth of field; there's no comparison between performance in low-light and harsh-light conditions; no comparison between flashes. All-in-all, pretty useless.
Also, camera shake - just look at the screws. I'm amazed that Scoble didn't take the time to use a tripod - the main comparison on display is how steadily he held each phone!
The really bad quality of iPhone 4S shot may be due to his hands shaking, we really wouldnt know untill somebody uses a tripod to do comparative shots.
Nice thing about Galaxy S2 software though is that you can specify ISO, focusing type as well as use a timer to deal with "shakiness" of the shot.
In less than ideal light situations, Galaxy suffers from shutter lag, i havent tested the iphone picture taking that much though.
33 comments
[ 3.5 ms ] story [ 72.2 ms ] threadI think that the 4S shot is too red/yellow compared to the Galaxy II. This sort of comparison is pretty subjective though.
(aside for evanwalsh: for the record I hate my Nexus One and have a 4S on the way)
Without making sure parameters & the software used were the same, such a casual comparison doesn't seems too useful to me.
You may be able to say "this photo looks right" but that's pretty darn subjective. Turns out we humans are kinda useless for tests like these. Past a few megapixels this stuff really doesn't matter because things like focus, lenses, and lighting are so much more important than resolution. The photos are "sharp" enough and none of these phones support mounting lenses or lights on anything. So its kind of arbitrary. Unless one of these phones has an especially shitty lens or focus or bad software, they'll probably just look mostly the same. I consider these long solved issues and don't expect much difference.
While not indicative of real-world use, comparison shots like this, IMHO, should be done with a stand... unless the only variable being tested is "likelihood of shakiness due to camera(phone) form factor."
Looking at the shots again, it does seem that the sharpest image, that of the original iPhone 4, is also the noisiest, iPhone4 autoexposure set it to high-iso/fast shutter I suppose?
The other thing are the colors, which seem much better in Samsung's picture. This, too, has nothing per se to do with the quality of the camera but it's settings. This time it's the white balance that's different. This is one of the things, where camera automatics are still often wrong. A small change of the subject can cause big differences. Moving a few centimeters more to the left is enough...
Again, you can try that with a "real" camera. A DSLR will let you choose the color temperature, but even cheap digital cameras have sometimes different modi for "sunlight"/"cloudy"/"artificial light". I won't go into details here (see [1]), sufficient to say that this can greatly affect a picture.
Also note that the picture from the Galayx II and the one from the iPhone 4S have roughly the same brightness and sharpness. Judging from the noise and resolution, the two cameras seem quite similiar (and are maybe even produced by the same factory).
TL;DR This comparison is not meaningful at all. For a professional result you would need to control the cameras completely manual and take a lot more pictures in different settings (light, reference colors, tripod, ...).
[1] http://en.wikipedia.org/wiki/Color_balance
EDIT: He actually made more than these three: https://secure.flickr.com/photos/scobleizer
Well, it's also worthwhile testing how well the automatic settings work -- that's how most folk will use the phone.
(All I really want in a camera phone is non-shitty low-light photos.)
His stream, which has some other test photos with commentary: https://plus.google.com/111091089527727420853/posts
Such people know how to insert themselves in the daily conversations and can count on the support of their followers to carry their views far and beyond. Most of the time those views are not much more than a catchy social hook. You Googling his name and reading this reply will likely cement Robert Scoble in your brain for a while longer.
I find it an interesting, albeit at times annoying, phenomenon. Like the Quora drama (in which Scoble was able to permeate into a lot of topics):
http://scobleizer.com/2011/01/30/why-i-was-wrong-about-quora...
http://www.quora.com/Dan-Kaplan/Sorry-Scoble-Quora-is-not-yo...
Heh, I love that. I'm going to use it.
Much of a muchness, IMO. The two are very close, and aside from things like color temperature (which I personally would discount and tweak later if desired - though I wonder if the white balance is configurable on either phone), they are all but equal.
http://campl.us/f349
Not a comparison, but it really does show that the iPhone4S camera is pretty damn good.