Wednesday, March 21, 2012

Side-by-Side ISO comparison of Canon 5D Mark II vs. 5D Mark III (ISO 100-25600)

I just received my Canon 5D Mark III (after waiting forever for a 5D Mark II replacement) and was anxious to see how this new camera handled high ISO shooting. Since I shoot a lot of events where I have minimal light, this is a critical feature for me. Sure, there are lots of great new features in this camera, but having clean images at ISO 3200 or ISO 6400 is the holy grail for me. So...with that in mind, I set them up for a stand-off!

Here is my test scenario:

* 5D Mark III and the 5D Mark II side by side on tripods
* Focal length set to 200mm (100-400 on the Mark III and 70-200 on the Mark II which could effect sharpness)
* Focused on the same object (a signed football helmet)
* Aperture of f5.6
* Set timer mode to 2 sec (to avoid any shake)
* Used the new Lexar 1000x cards (to test the speed of the UDMA7 card slot)
* Turned off the lights in my home office
* Shot in RAW and imported into Photoshop with no altering of the files
* Zoomed and cropped to show a close-up of a dark and light area

Room setup with both cameras side-by-side (you can not tell from this image, but the 5D Mark III LCD is much brighter and sharper than the LCD on the 5D Mark II)

(Original image taken with the 5D Mark III at ISO 100 - scaled to 33pct)

And here are my results. Note: You can click on any of the following images to see a much larger version.

At ISO 100, both images are very clean. Interesting to see such a difference in the colors and gradations.

At ISO 800, we start to see some noise in the Mark II image but the Mark III stays clean.


At ISO 1600, we see a small amount of noise showing up in the Mark III image, but it is still far better than the Mark II image.

At ISO 3200 we see noise in both images, but after careful comparison, you will see that the Mark III image at ISO 3200 has about the same amount of noise as the Mark II at ISO 800



At ISO 6400, we see significant noise in the Mark II image but the Mark III image still looks very useable.


Now we are taking the Mark III into an area where the Mark II shouldn't go. So, I kept the Mark II at ISO 6400 to show you the comparison of the previous maximum ISO to the new higher ISO of the Mark III. You could argue that these two images, even with very different ISO settings, are about equal in quality.

It ISO 20000, as one would expect, we have significant noise in the Mark III image. But in the rare case where you need to shoot an image with almost no light present, this would yield a useable image, especially for a newspaper or small printed image.

Lastly, we crank it up to ISO 25600 to see what we get. At this point, we are in a pretty noisy world of digital imaging, but the image does not look much different from the ISO 20000 photo. With this in mind, I would avoid the ISO 20000 and go straight to ISO 25600. I did shoot some test images in a room that was almost pitch black, with just some light coming off of a distant computer monitor and was impressed that the camera could focus on the subject. I was also amazed at how much light the camera grabbed. I could barely see the subject, but when viewing the image on the LCD of the camera, it looked like daylight. Pretty amazing.

More tests to come!

6 comments:

Anonymous said...

so I am confused, because according to your samples photos, the new 5d Mark iii is worse in terms of noise. Is this correct or did you label the pics incorrectly?

Jeff Cable said...

N0 - the 5D Mark III is definitely better than the Mark II. Not sure which images you are referring to. I think they are all correct.

Jim Harmer said...

Jeff, I think he looked at the last photos where the II was still on 6400 and the III was higher, without reading the text.

So would you agree with the conclusion that the III gets about an extra stop of light for the same noise level?

Anonymous said...

Sorry, the fact that you used the same ISO setting for each camera up to 6400 and then only raised the mark iii's iso through me off. I thought the iso setting were equal for all shot's, but now I see my error.

I will say that I am concerned that the mark iii looks markedly softer. I have pre-ordered from Amazon and I will be very upset if my new mark iii is softer than my mark ii. Can you put my mind at ease? Have you compared the two using the same lens?

Jeff Cable said...

Honestly, I have not had time to test any further. I do believe that the softness is due to the lens and not the camera.

Android Indonesia said...

Thanks for review, it was excellent and very informative.
thank you :)