Lens/Body Testing Results and Thoughts

[dandaluzphotography]

Hi Guys,

My friend and I spent the better part of 4 hours last night testing our lenses on our three different bodies and I wanted to share our finding with you guys. We were looking for flaws with autofocus with three different bodies and 4 different lenses:

7D
60D
XSI

70-200 f/4L
70-200 F2.8L IS II
24-70 F/2.8
17-40 F/4

All the shots were taken on a tripod with the following settings: 1/200th, F4, ISO 100, 580 EX II flash set to auto. We tried all the lenses with all the bodies. I took pictures of my friends face zoomed in half way with the focal point over his right eye.

We found the best combination, as far as sharpness with autofocus, to be the 7d/70-200 F/4. All other combinations produced soft to very soft images when viewed at 100%. A couple were even blurry.

We took all the shots again using manual focus and this time every shot was extremely sharp; sharper than the 7d/70-200 f/4 combination using autofocus. There were some shots where I missed the focus, but I adjusted and it was awesome. The sharpest combination was the 7d/70-200 f/2.8 with the 70-200 f/4 a close second followed by the 24-70 and then then 17-40, all on the 7D. The 60D was also great but the images from the 7D were slightly sharper.

These were the results for these lenses and these bodies. As I understand it, other copies of these lenses and bodies could produce completely different results as they are all slightly different even within the same body/lens type.

The way I see it, I cannot rely on my 60D and 17-40 and 70-200 f/4 lens using autofocus. It just doesn't work well. If I'm going to be taking pictures that won't be viewed larger than, say 1920x1200, which is pretty big already, then it might not matter too much. As far as the shots that I took using autofocus that were soft, they can be fixed with sharpening in CS5 but I feel like maybe I shouldn't have to do this.

Are my expectations too high? Should I send the camera and lenses to Canon to be calibrated? I'm leaning towards no. At least not for my current needs. I don't see myself needing to have images viewed on screen or in print for that matter, at full resolution where the autofocus issue is really evident.

Thoughts?

Danny

 

[Derrel]

My thought is that you seem new to digital SLR photography and the many differences it has when compared against shooting film. I think you might not be used to the sharpness-killing aspect of anti-aliasing filters used on modern, high-MP d-slr cameras.

As far as the Canon 17-40 not producing sharp images...it's a wide-angle zoom...its images cover a lot of physical real estate (due to the wide-angle viewing!!!!) and so the pixels typically are spread across a fairly large area of the real world, and with a tele-zoom, the area covered is normally a fairly small piece of the real world...when the two differing lenses' images are looked at side by side, the lens that has its pixels covering a smaller, physical area usually looks "sharper".

 

[dandaluzphotography]

My thought is that you seem new to digital SLR photography and the many differences it has when compared against shooting film. I think you might not be used to the sharpness-killing aspect of anti-aliasing filters used on modern, high-MP d-slr cameras.

As far as the Canon 17-40 not producing sharp images...it's a wide-angle zoom...its images cover a lot of physical real estate (due to the wide-angle viewing!!!!) and so the pixels typically are spread across a fairly large area of the real world, and with a tele-zoom, the area covered is normally a fairly small piece of the real world...when the two differing lenses' images are looked at side by side, the lens that has its pixels covering a smaller, physical area usually looks "sharper".

I'm absolutely new to this. I've been into photography for about 1 year now, so I've never even shot film. I'm trying to learn all these different things!

 

[Big Mike]

It is certainly possible that different lenses (even different copies of the same lens) and different cameras will give you different results. I know a photographer who is supper picky about accurate focus, and as such, he's owned six different copies of the 24-70mm F2.8 L.
There are so many little aspects that can lead to (or away from) that 'perfect' sharpness that we want. For me, I know that not all my shots are going to be as sharp as I would like...but I have developed a tolerance of what is acceptable.
Part of the problem is the way that we evaluate images these days...which is usually to have it on a computer monitor and zoom in to 100% or closer. All images will display 'faults' if we zoom in close enough.

I think that a better (or more realistic) test should be something like this....print out an 8x10 and hold it at arm's length.

 

[dandaluzphotography]

It is certainly possible that different lenses (even different copies of the same lens) and different cameras will give you different results. I know a photographer who is supper picky about accurate focus, and as such, he's owned six different copies of the 24-70mm F2.8 L.
There are so many little aspects that can lead to (or away from) that 'perfect' sharpness that we want. For me, I know that not all my shots are going to be as sharp as I would like...but I have developed a tolerance of what is acceptable.
Part of the problem is the way that we evaluate images these days...which is usually to have it on a computer monitor and zoom in to 100% or closer. All images will display 'faults' if we zoom in close enough.

I think that a better (or more realistic) test should be something like this....print out an 8x10 and hold it at arm's length.

I mentioned before how this is probably fine for my current needs. Plus most of the stuff I do is going to be viewed on a monitor. I think it's fine.

 

[Derrel]

Oh, my mistake Danny...I was under the impression that you were an older "film guy", and thus a person unfamiliar with all the oddities and quirks of digital imaging.

 

[dandaluzphotography]

Oh, my mistake Danny...I was under the impression that you were an older "film guy", and thus a person unfamiliar with all the oddities and quirks of digital imaging.

LOL. I'm a younger "digital guy" unfamiliar with all the oddities and quirks of digital imaging. But I learn more every day!

 

[GooniesNeverSayDie11]

The 7D also has the capability to micro adjust for individual lenses. The other 2 cameras do not. There are many differences between focal lengths and sharpness as was already explained. Typically wide angle lenses are less sharp. If you shoot a yard with a tree 40 feet into the scene, it will be much less detailed when shot with a 10-22 than if you shot it with a 70-200. Thats just typical. So in some regards you may be expecting too much. You also shot a subject that has the potential to move, even if in the slightest way. You also shot using a focus point other than the center ( which is less accurate ) or you focused and recomposed which could be your problem in and of itself.

Another aspect is lens and camera tolerances. To make this easy to understand, we'll say a lens is given a grade of 1-5 when its made, 1 being tack sharp and 5 being a tad soft. All are acceptable by quality control. A 6 or higher however is considered faulty and not sold. Lets say its the same with Cameras AF systems. So you may get a lens that is a 1 and a body that is a 1 and you have a match made in heaven. But what if you have a body thats a 2 and a lens thats a 5? Less sharpness. What if you have a body thats a 5 and a lens thats a 5? ( You are going to be pissed if its a combo that cost you nearly 10k ) Thats what canon fixes when you send in a body and lens for them to trouble shoot and adjust.