Cameras have always been close to my heart. It all started when my dad bought me a disposable camera so I could take my own pictures during our family vacation to Hawaii. That was the year of 1997 when I was only 5 years old. The 27 frames that the 35 mm film held, introduced me into the world of photography. Soon after I had my very own first camera, a Nikon F60 SLR, it was a Christmas present, a good one at that! Being a photographer and a technology enthusiast, I am always interested in advances related to cameras.
Today, SLRs are almost obsolete, only people that truly respect and appreciate the technology still use them. DSLRs occupy the market now; there is no more film, it's all digital. In fact, even that statement is a stretch. DSLRs are the only thing on the market, but almost nobody buys them anymore, only professional photographers do. So what do ordinary people take pictures with you ask? Phones! People don't want to carry multiple devices on them, and they don't have to! Ever since the introduction of the iPhone in 2007, people use their smartphone cameras.
Just like anything else in the world of technology, smartphones, and their cameras, are constantly being redesigned and improved. Since majority of people own smartphones and use their cameras everyday I wonder how good those cameras are. Obviously, they have to be pretty good because everyone uses them! I am also interested in how the smartphones compare to each other and whether their rate of advancement is adequate enough. So let's get to it!
I carry around an iPhone 5 which was released in September of 2012. My brother carries my old hand-me-down iPhone 4 which was released in June of 2010; 2 years apart. Let's talk about the internals of their cameras.
Both iPhones have a 1/32" (4.54x3.42 mm) sensor size and a 5x digital zoom. iPhone 4 has a 1.75 µm pixel size, 3.85 mm focal length and a F/2.8 aperture, while the iPhone 5 has a 1.4 µm pixel size, 4.10 mm focal length and F/2.4 aperture. Given these figures we can calculate the amount of mega pixels iPhone 4 and 5 have: (2592 x 1936) 5 MP and (3264 x 2448) 8 MP, respectfully.
So what do these numbers tell us? In the world of smartphones, they are competitive. You would find about the same cameras in other phones around those two years. But in comparison to actual consumer-grade DSLRs they are about 2 or 3 years behind. By no means are they revolutionary or comparable to professional cameras but they are a tenth of a size and considerably cheaper. That's where you have to draw the line. If you were in photography contest you wouldn't be using your phone, but most people are not professionals. This is why smartphone cameras occupy majority of the market; they are more than enough for quick pictures that most people take.
For my comparison of the two iPhones I will take 5 kinds of pictures with 3 zoom intervals. I will take a picture of
- a white sheet of paper
- a black sheet of paper
- half black half white
- a well lit outdoor picute
- a dim outdoor picture
- indoor picture
Once I have the pictures taken, I will compare their file sizes, ISO and shutter speeds.
This week I will compare the pictures of white, black and half white half black. Since black contains all of the colors, I am curious if it's file size will be much bigger than whites and in the half white half black picture if the file size will he half smaller. Obviously I do not have optimal lighting conditions so white may look gray at times, adding more colors than just pure white and contributing to a difference in file sizes, but I did the best I could. Let's get started.
iPhone 5 iPhone 4
 | |||||||||
| 1.2 MB ISO 3200 Shutter speed: 1/15 no zoom |
 |
| 772 KB ISO 1000 Shutter speed: 1/15 no zoom |
 |
| 688 KB ISO 3200 Shutter speed: 1/15 half zoom |
 |
| 638 KB ISO 1000 Shutter speed: 1/15 half zoom |
 |
| 485 KB ISO 3200 Shutter speed: 1/15 full zoom |
 |
| 492 KB ISO 1000 Shutter speed: 1/15 full zoom |
 |
| 2.2 MB ISO 200 Shutter speed: 1/20 no zoom |
 |
| 1.8 MB ISO 80 Shutter speed: 1/17 no zoom |
 |
| 1.3 MB ISO 200 Shutter speed: 1/20 half zoom |
 | |||||
| 1.1 MB ISO 80 Shutter speed: 1/20 half zoom |
 |
| 929 KB ISO 200 Shutter speed: 1/15 full zoom |
 |
| 689 KB ISO 80 Shutter speed: 1/20 full zoom |
 |
| 1.2 MB ISO 1600 Shutter speed: 1/17 no zoom |
 |
| 1.7 MB ISO 1000 Shutter speed: 1/15 no zoom |
 |
| 1.0 MB ISO 1000 Shutter speed: 1/15 half zoom |
 |
| 1.2 MB ISO 800 Shutter speed: 1/15 half zoom |
 |
| 817 MB ISO 800 Shutter speed: 1/15 full zoom |
 |
| 656 KB ISO 1000 Shutter speed: 1/15 full zoom |
I'm actually surprised with the results. First and foremost, I was expecting pictures of black color to be bigger in file size than of whites, but in fact they were about 50% smaller. Black contains all colors whereas white doesn't contain any. I was curious if that would be reflected by the file size, and apparently it is, but it's complete opposite of what I thought it would be.
iPhone 4 has 5 megapixels and the iPhone 5 has 8 megapixels, a 37.5% increase. I was curious to find out whether or not the file sizes would show the same correlation. They don't, the file sizes range anywhere from 15% to 35.7% increase.
Another interesting observation, which I was anticipating, is that pictures of white have a very small ISO speed whereas blacks have very high ISO speed.
Onto the zoom, obviously the further you zoom the more granulated the picture becomes, especially with digital zooms. I was expecting the file sizes to decrease about 50% at mid and full zoom and for the most part they do, it's not exactly 50% each time but based on this data I think it is save to make the assumption that they generally do. There is one outlier to this generalization, on iPhone 5, the picture with no zoom and half zoom on half white half black did not change in file size; I am not exactly sure as to why.
Last but not least, the half black half white pictures. I wanted to find out if they will be half the size of a pure color. As I said before, I thought it would be black that's larger than white in file size but that's reversed. The pictures are not half of the file size, but they are in-between the two, which makes sense.
All in all, I got a lot of unexpected results. Now that I have what I think is a foundation, I wonder if I will see similar results with actual pictures of objects next week. I wonder if dimmer pictures will have a bigger file size and higher ISO speed... We'll see!
Until next week!
-Mario
Update: Someone pointed out to me that I saw inverse results with the white/black comparison because white is the color that contains all colors whereas black does not. I did some more research.
There are actually two answers here, depending on how you interpret the question.
1)Light: With light white is all of the colors. white light can be split by a prism into the rainbow and its parts. Black is no light and therefor no color.
2)Paint/physical mediums: White paint is no colors, black paint is all of the colors, this is also explained by the absorption principal of light.
Hopefully this eliminates some confusion, there is no right/wrong answer here.
I agree with your comment about light and physical media explanations for black and white. However, I disagree with your thought that one being made up of lots of colors and one not having an effect on the file size. Theoretically, the black image would store a pixel (in bits) as 00000000 00000000 00000000 meaning no RED, no GREEN, and no BLUE color components. This lends itself very well to a compression algorithm and can be stored as 24(0), or a larger number if surrounding pixels are the same - in this case the entire image could be compressed to two numbers. The number of bits of 0's (all of them) and 0 to indicate the value.
ReplyDeleteAgain, in theory, the exact same thing would happen with all white. The pixel would be represented by 11111111 11111111 11111111 and compressed likewise. As you can see from the images, the assumption about the black being all 0's is much more likely to be true that the assumption about the white being all 1's. Once the values for white becomes something like 11101001 11101011 11100011 - meaning not pure white, the compression algorithm does significantly worse 3(1)1(0)1(1)2(0)4(1)1(0)1(1)1(0)5(1)3(0)2(1). Note that this is just one kind of compression and other kinds react differently to fluctuations in values. I just wanted to point out that the details of the file size are more complex than it might seem.