Time to be with your loved ones

Time to be with loved ones

by sat

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.

There are lot of sites, which compare digital cameras by parameters, user experience and etc.

I looking for a website which has some kind of procedure to test image quality and compares digital cameras by the result. Are there any?

share|improve this question
1  
What exactly is "image quality"? How would you compare it objectively? –  mattdm Jan 9 '11 at 0:38
    
There's several factors that I could see this including. They include the amount of time it takes to take a picture, the amount of noise on the image and other qualifiers, and probably some others as well. –  PearsonArtPhoto Jan 9 '11 at 1:47
1  
I would add: colour accuracy, possibly higher ISO noise. –  AJ Finch Jan 11 '11 at 11:27
    
Noise, real sensitivity (time/aperture/ISO for fixed light), resolution, aberrations are factual comparable for compact cameras (as some of these parameters are for lenses for SLR). –  Leonidas Apr 27 '11 at 0:20
add comment

7 Answers

http://www.kenrockwell.com/ is a bible for choosing a lens or a camera !

share|improve this answer
1  
Please see photo.stackexchange.com/questions/10980/… –  mattdm Apr 29 '11 at 12:36
    
Very interesting thread, thanks for the link. –  Matthieu Napoli Apr 29 '11 at 12:46
add comment

First off, comparing images on the Internet is a misleading business since the resolution is usually too low to show up differences. Images from my old 2.1 Mp Canon P&S look not much inferior to my present 15 Mp DSLR when displayed in a normal web page.

However DPReview.com have, very fortunately for us, had a policy over many years of taking photos of the same resolution chart, under the same conditions and storing the original image for download. It is a marvelous resource. There you can find resolution images going back many years for all the major camera models.

Technical image quality has several dimensions and all need to be measured if you want a complete result. For many people though, resolution and chromatic aberration are the most important results. Luckily these things are easy to measure if we use the DPReview.com resolution image library.

Imagine we want to compare the following cameras (selected for no particular reason)
You can find the converted raw images under the Resolution heading of the camera reviews:

Procedure

  • Click on the links to download the full size images of resolution charts.
  • Download a copy of quickmtf. You can find it on Quickmtf.com. You can get 50 free measurements before you have to pay for it. We use quickmtf because it uses the ISO standard for measuring resolution according to the SFR slant edge method.
  • Open each image in Quickmtf. You will see a number of 5 degree slanted edges at different points in the chart. Choose one of interest to you.
  • Using the cursor tool, draw a rectangle to enclose part of the slanted edge.
  • A graph will open up showing the MTF curve. Now you can read off things like line pairs per picture height.
  • In the tool bar click on the buttons for Edge Spread and Line Spread. You will be shown the edge spread and line spread functions. Here you can read off the edge blur and the chromatic aberration.

Results

Horizontal edge, 15% from the center.
Edge blur in pixel (chromatic aberration in pixel).
For reference - 1.27 pixel is the best attainable result.

  • Nikon D7000........2.5 px (0.8 px)
  • Nikon D5100........2.6 px (0.6 px)
  • Nikon D3100........2.1 px (0.7 px)
  • Canon 600D.........2.4 px (0.6 px)
  • Panasonic DMC-G2...2.0 px (0.2 px)

I show edge blur because it is immediately understandable to most people. Reading MTF graphs is challenging. The important thing to note is the ringing in the graphs for Canon and Panasonic. This indicates that the images have been sharpened more than the images for Nikon, which distorts the results. Ideally the images should not have been sharpened, but we have to work with what is available.

The graphs are shown below:
Nikon D7000
enter image description here
Nikon D5100
enter image description here
Nikon D3100
enter image description here
Canon 600D
enter image description here
Panasonic DMC-GF2
enter image description here

share|improve this answer
add comment

Comparometer

This tool lets you compare reference images from digital cameras side-by-side. These images are JPEGs straight from the camera, and were taken under carefully-controlled conditions, to provide valid comparisons of camera capabilities in actual shooting situations. You can also download the images (using your browser's "save image as" function) and output them on your own printer, to see how the cameras involved will perform in your application.

http://www.imaging-resource.com/IMCOMP/COMPS01.HTM

share|improve this answer
    
Love this site. Once the pictures and data convinced me not to buy the Canon A620 instead of the A610 - sensitivity was lower, picture quality was not better. –  Leonidas Apr 27 '11 at 0:17
add comment

It's impossible to objectively rate image quality, as image quality is quite subjective. You could look at example images, but they probably won't reflect the situations you would use your camera in.

Thankfully, most modern cameras have excellent sensors that outresolve all but the most demanding of photographers.

share|improve this answer
add comment

It probably doesn't matter. Taking great looking photos has much more to do with the photographer than the camera. A good photographer can take great photos with any camera.

I suggest you decide what kind of camera to get based on the kind of shooting you do or intend to do, then visit snapsort.com to find a good model in that class. If you're going to take a lot of low light photos, get something with a fast lens; if you'll be shooting over a long distance, get something with a good zoom, and so on.

However, I agree with the suggestion to visit dpreview.com. They have extremely comprehensive reviews, which includes image quality and comparisons against other cameras.

share|improve this answer
1  
While I agree with the basic sentiment, I think it's worth noting that different cameras are better for different purposes. For an extreme example, consider a Sinar -- undeniably a good camera, but I sure wouldn't want to use it to shoot a basketball game! –  Jerry Coffin Jan 9 '11 at 3:08
    
This is exactly what I suggested: decide which broad category of camera he wants, then use SnapSort to recommend a specific model. –  ieure Jan 9 '11 at 6:10
3  
It's a nice sentiment but you've clearly not tried to use the camera on my old cell phone. It was truly terrible. In the hands of the greatest photographer alive you might get some interesting shots, but I think the same photographer would get even better photos (and a lot more of them) with a high quality camera and lens. With this in mind, why not strive to get the best camera in addition to being a good photographer? –  Matt Grum Jan 9 '11 at 12:29
    
That's irrelevant, since he isn't going to buy the camera on your old cell phone. While I'm not disputing that there are quality differences between cameras, I think the image quality between newly purchased cameras isn't drastic enough to be the single issue on which a purchase hinges. –  ieure Jan 9 '11 at 22:54
    
It often does matter a great deal and anyway artistic quality deserves to be displayed with the highest technical quality. I take/print many large panorama photos. Here I find exceptional sharpness and flatness of field really does matter. I observe people as they view my prints. They first step back to take in the print as a whole and then, without exception, step closer to view the details close up. And then they react with pleasurable surprise to see the fine detail faithfully preserved. –  labnut Apr 27 '11 at 10:15
show 1 more comment

DxO Mark provide objective data on the output of camera and cameras in combination with lenses.

The data comes from the testing they do on cameras and lenses for their RAW developing software "DxO Optics" so it should be fairly unbiased. I don't know enough about the technical aspects of digital photography to comment on the validity of their methodology.

share|improve this answer
4  
While DXOMark is a nice metric that has some measurement value, it encompasses very little of what constitutes image quality. –  Itai Jan 9 '11 at 2:28
1  
If there were some way by which I could vote +100 to what Itai says here, I would! –  mattdm Jan 9 '11 at 16:37
1  
Interesting. I expected there to be objections, but I thought they would relate to how the data was collected, not the metrics themselves. Surely, for example, the signal-to-noise for a given ISO value has a strong correlation to image quality in environments where you're forced to use higher ISOs? Similarly, anyone who shoots under (for example) bright stage lights would find a camera with a large dynamic range would improve the image quality in their highlights and/or shadows? –  Scott Carroll Jan 9 '11 at 17:43
    
To my knowledge DxO Mark conduct the most comprehensive, thorough, careful and exact set of photographic measurements that are published on the Internet. My main criticism is that they could use better methods for measuring lens resolution. For example use a microscope to examine the aerial image or better still, use Hartmann-Shack wavefront analysis. –  labnut Apr 28 '11 at 8:39
    
@mattdm, do you not believe that, in principle, careful measurements bring greater knowledge and insights to many subjects, including technical image quality? –  labnut Apr 28 '11 at 8:42
show 3 more comments

The website dpreview.com has detailed reviews that include sample images. You can preview sample images from various models to get an idea of the results each produce.

Also, check out flickr.com you can see images sorted by camera that took them as well.

(The links I provided are for viewing results for the Nikon D90)

share|improve this answer
    
Note that the reviews at dpreview.com not only contains sample images, but also test images of standard scenes as well as comparisons between how different cameras handles the test shots. –  Guffa Jan 9 '11 at 1:48
    
using flickr is an excellent suggestion –  AJ Finch Jan 11 '11 at 11:28
add comment

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.