Canon currently produce two camera bodies that have quite a high degree of similarity, based on their general specifications. The two bodies are the Canon 7D, which is an APS-C sensor camera, that has 18mp of resolution, 8 fps drive and the Canon 1Dmk4, which has a larger APS-H sensor inside, with 16mp and 10 fps drive speed.
I often am asked as to what some of the main differences are in image quality between these two cameras. With this in mind I set up the two cameras to shoot some comparative images from both in order to compare some of their characteristics. In an earlier post I carried out this exercise with both cameras set at ISO 400. You can find that post by looking in the Blog Archive on the right side of this page under the February tab.
For this comparison I set them to ISO 1600.
I used a Canon EF 70-200 f/2.8L IS II lens, mounted on a tripod. Only the camera body was changed each time. No sharpening was applied other than a default setting of 25 in Adobe Camera Raw. Raw images opened in Photoshop CS5, downsized and saved using File Save For Web and Devices at 90kb file size.
The images were taken in good light, as can be seen from the exposure details. The black and white zebra made a good target for comparing detail in blacks, whites and fur, especially around the edges of the ears. I also included the dark green and shadow patches in the background for comparing noise levels.
|Canon 1Dmk4. ISO 1600. 1/500s at f/5.0, as taken. 16mp, no cropping.
|Canon 7D. ISO 1600. 1/5000s at f/5.0, as taken. 18mp, no cropping.
The first image is that of the Mk4, with its 1.3x crop effect. The image directly beneath it, and above this text, is that taken with the 7D, with its 1.6x crop effect.
Another way to better understand the size difference between the images is to relate them to the sensors in each camera.
The APS-H sensor in the 1Dmk4 is 27.9mm long and 18.6mm high. When the image circle is projected from the rear of the lens, more of it falls onto this sensor then the smaller sensor in the 7D, hence the wider field of view, with more background visible.
The APS-C sensor in the 7D is 22.3mm long and 14.9mm high. When the image is projected from the rear of the lens, this smaller sensor can only record over its smaller surface and you lose more of the captured image around the edges of the image circle.
These two pictures serve to show the difference in subject size between the image obtained by the 1DMk4 and the 7D then, out of the camera.
The Mk 4 image is made up of 16 megapixels, and each pixel is 5.7 microns in size.
The 7D image is made up of 18 megapixels, and each pixel is 4.3 microns in size.
None of that is apparent whilst looking at them now, only the fact that the 7D is giving you a bigger zebra.
In the next step I have cropped the same two images.
|Canon 1Dmk4. ISO 1600. Cropped from 16mp to 6.8mp.
|Canon 7D. ISO 1600. Cropped from 18mp to 12mp.
In order to make it a bit easier to see differences at what might be considered a normal crop size, I cropped the bottom image, from the 7D, to 12 megapixels, thus discarding 6 megapixels.
Then, I ‘equalized’ the two cameras by cropping the Mk4 image, until the zebra was the same size, approximately, as the zebra in the 7D image. This left me with a cropped 1Dmk4 image of close to7 megapixels, still pretty much usable for most requirements in resolution.
Comparing the quality of the two images on my 23″ screen side by side, I cannot see any massive quality differences between them. To my eye the blacks look a little better on the Mk4 image, and the background may be showing very slightly more noise on the 7D image.
For the last example, I again cropped the two images further, keeping the zebra pretty much the same size.
|Canon 1Dmk4. ISO 1600. Cropped from 16mp to 2mp, downsized to 800×533.
|Canon 7D. ISO 1600. Cropped from 18mp to 3.5mp, downsized to 800×533.
This time I cropped in heavily, to try and make it easier to judge image quality almost at pixel level.
The cropped 1Dmk4 image comes in at 2.0 megapixels, from a 16mp original, and downsized to 800×533 for the test.
The cropped 7D image comes in at 3.5 megapixels, from an 18mp original, and downsized to 800×533 for the test.
In this post I cropped a little less aggressively than on the last pair of images in the ISO 400 post. I chose here to crop the Mk4 image to 2mp which is really for illustrative purposes only.
For my own workflow for web I prefer not to go under about 4mp in resolution with any image.
To my eye the Mk4 file is still slightly superior in terms of noise, the blacks, and the general tone of the whites in the image. On the other hand the 7D file is retaining more detail in most areas, due to the Mk4 file running out of pixels.
This comparison shows that in good light, and at ISO 1600, it is possible to crop the 1Dmk4 image until it matches the “subject size” of the same image taken on the 7D, and that there is not a great deal of difference in the image quality of the two cameras when this cropping is carried out. The Mk4 image shows a slight contrast advantage at moderate cropping levels. At extreme crop sizes, the higher resolution of the 7D gives it a definite advantage.
If you are surprized at how close the two sensors are in these comparison images, it needs to be remembered that they come from the same era, with similar levels of technology applied, and we have taken away the advantage that the larger Mk4 pixels may have by cropping it to equalize subject size.
The comparison images were also shot in bright light, which also narrowed the low-light advantage that the Mk4 holds.
A different method of testing sensor characteristics and image quality between the two cameras would be to compare image quality between the two bodies when longer focal length lenses were used on the Mk4’s bigger sensor, to ‘equalize’ the sensor characteristics, for example shooting the 7D at 100mm focal length and the Mk4 at 130mm, approximately.
I have taken a set of images for this purpose, in very low light and will be posting that comparison soon.
After extensive use of both of these camera bodies, my own findings in actual shooting conditions pretty much supported what comparing these images shows but I would of course not usually be shooting at ISO1600 in such bright light to begin with, though from the results you can see that you could if you wanted to.
Where focal length is a limiting factor, in other words if the subject is so far away that only very heavy cropping will get enough subject in the frame for the final output, and in good light, the 7D has a resolution advantage and its smaller pixel pitch are hard to beat. The comparison test carried out here simulates a situation where focal length was limited, in other words we cropped the subject much closer, in stages, with heavier cropping being applied to the Mk4.
I did not calibrate the two camera bodies to the camera lens. Whilst there were no clouds in the sky, and I shot as quickly as possible between changing bodies, there might be slight changes in available light, though not enough for the camera metering to pick up. I may have also not gotten absolutely sharp focus on the Mk4 iso 1600 shot as the nose of the zebra is not as sharp as in some of the other test frames at different settings with the same camera and body but this will not affect the noise comparison much. I do not have a 7D with me at the moment to repeat this image.
No sharpening applied to these images, but for my purposes, in raw format, I felt that both lens and camera combinations were working well enough.
With proper sharpening applied, there would be lots more detail showing on the subject, even the slightly softer (user-error generated) Mk4 image.