7D Image quality??

[StanNJ1]

I'm trying to help a friend of mine evaluate why a lot of his images taken with his new 7D look soft. At first I was convinced it was because of his 18-135 kit lens. But when I mounted my 70-200L 2.8 on his camera and looked closely at the images they too looked softer than when mounted to my 5D or 1DMKIII.

Are there any known image quality issues with the 7D?

[gjl711]

Are you evaluating per pixel sharpness? Overall sharpness? Have you MFAed the lens?

Posting a sample which shoes the issue would really help. There are lots of reasons why an image is soft and many are user related and some are not.

[kfreels]

This comes up all the time. The 7D has a rather aggressive anti-aliasing filter as compared to the T2i, 60D, et al. This helps it handle moire a lot better but the down side is a loss in apparent sharpness. The solution is a slight increase in output sharpening.

In-camera jpegs have always seemed a bit on the soft side for my taste so I shoot in RAW almost exclusively. I and several others seem to think that when shooting jpegs, a good deal of data is lost in the jpeg compression process that can't be recovered after the conversion process. So sharpening at this point isn't as effective. The good news is that it can be dealt with easily. It's simply a matter of using the picture style editor and editing your styles with a higher sharpness setting.

As I said, I and a few others think that Canon may have failed to correctly set the default styles up to deal with the stronger AA filter and this often leads people to think there is a softness problem when it is something that is really a settings issue.

[AJSJones]

Originally Posted by kfreels

This comes up all the time. The 7D has a rather aggressive anti-aliasing filter as compared to the T2i, 60D, et al. .

I found its AA filter about on a par with that of the 5D2, perhaps a tad less moire (100L macro direct comparison) - they both had limiting resolution (perhaps 10% MTF) at ~ 90% of Nyquist. Do you have some comparisons for the 7D and other cameras - I'm always trying to collect data on the AA filters - it's hard to eliminate the other variables to test the AA filter directly.

[wayne.robbins]

Mostly, on the 7D, and even on the 5D3, I shoot Raw+JPG. I do this because it's convenient- if the JPG looks good enough SOOC, I leave it, otherwise I have the raw file to play around with. That said, I often bump up the Picture Styles on the 7D - using the Standard setting- but tend to bump up the Sharpness by +2 and sometimes the Contrast by the same. I would guess I am not the only one to do this. I've also heard of people shooting in "Landscape" picture style - because- sharpening is actually +1 over Standard's default settings.

[TeamSpeed]

Originally Posted by StanNJ1

I'm trying to help a friend of mine evaluate why a lot of his images taken with his new 7D look soft. At first I was convinced it was because of his 18-135 kit lens. But when I mounted my 70-200L 2.8 on his camera and looked closely at the images they too looked softer than when mounted to my 5D or 1DMKIII.

Are there any known image quality issues with the 7D?

I have seen more than a couple of L lenses that didn't result in sharp results from a 7D, but did on other bodies. Then again, I have seen EFS lenses that were razor sharp on a 7D as well. Once you have ruled out back or front focusing, and you still have issues, it will most likely be the glass and not the sensor. If you cannot get any image to be sharp regardless of glass, then that could be the sensor.

[StanNJ1]

Thanks for your responses so far. I will try to post some pics later. He has been shooting RAW. Do the output sharpening adjustments effect RAW images?

[TeamSpeed]

Originally Posted by StanNJ1

Thanks for your responses so far. I will try to post some pics later. He has been shooting RAW. Do the output sharpening adjustments effect RAW images?

It does if you use DPP, otherwise 3rd party tools will just apply a default group of settings if you allow them to, and you have to take it from there.

[moltengold]

i can make my 7D act like the T2i ,T3i , 60D
in the menu i adjust the Picture Style to User Defined 1
and changed all numbers to

Picture Style User Defined 1(Standard)
Sharpness 7
Contrast 2
Saturation 2
Color tone 1

this image from the camera not edit by any software
7D + EF100mm f/2.8L Macro IS USM

[kfreels]

Originally Posted by AJSJones

I found its AA filter about on a par with that of the 5D2, perhaps a tad less moire (100L macro direct comparison) - they both had limiting resolution (perhaps 10% MTF) at ~ 90% of Nyquist. Do you have some comparisons for the 7D and other cameras - I'm always trying to collect data on the AA filters - it's hard to eliminate the other variables to test the AA filter directly.

If you would have asked me a few months ago I would have had some good information for you. I had some pdfs that I discovered about a year ago on a mad hunt for this info right after I bought my 7D. I had upgraded from a T2i and I noticed it when I would do severe cropping on some subjects. But I found a way around it by adjusting the sharpness. (Also I bought some longer glass to reduce the cropping). Unfortunately I cleaned up my documents folder a few months ago and deleted a bunch of crap that I no longer cared to have and that went with it.

I do recall some of the text though. In particular it was noted that the main issue was with video. So the first batch of cameras with major video capability - namely the 5DII and the 7D - addressed this with that AA filter. The filter in the two was very similar but the tighter pixel density and smaller pixels on the 7D made it stand out more. Owing to a lot of gripes about "7D Sharpness" the AA filter on the T2i and 60D which came out later was less aggressive. So the strongest AA filter was that used in the 7D and 5DII but the loss in sharpness was most pronounced on the 7D.

One other thing it mentioned was the removal of the AA filter if the user doesn't shoot a lot of video. It said that moire can be addressed in post much easier on stills when necessary with a slight gaussian blur and it can of course be applied selectively vs to the whole image. I've considered having this done but can't quite get myself to do it.

[Lore]

I don't know, I flipflopped between the 5DmkII, 7D, then back to the 5DmkII, going through my flickr, I'm hard pressed to know which is from which unless I actually remember which is from when, the 17-55 & Tokina 50-135 were bangin' sharp on the 7D only common lens I had on both was my sigma 50 & for a short time 24-105L worked great on both.

From what I found, I seem to apply the same amount of sharpening to both RAW files, tho less on the files when I had the 135L, those files needed little to no sharpening PP.

I do know tho when I had the 5D (not mkII), the files were sharper SOOC which clearly is a result of the weaker AA filter.

Edit: I guess if I put them side by side w/ same lens photos I 'can' see the difference & I can see why I ended up with my current camera, course it looks like focus hit the box top.


Day 27 - Cat Fishing by lorephoto, on Flickr
Camera Canon EOS 7D
Aperture f/1.4
Focal Length 50 mm


Horatio by lorephoto, on Flickr
Camera Canon EOS 5D Mark II
Aperture f/1.4
Focal Length 50 mm

[Stone 13]

The 7D needs good to great glass to for relatively sharp output and it CAN require more sharpening that other cameras but I wouldn't consider it excessive maybe an extra +10 in LR for my shots compared to my 450D.

Keep in mind the 7D's pixel density will expose flaws in one's hand holding technique. What was acceptable technique with my rebel was revealed to be fairly sloppy with my 7D and cleaning up my technique made a world of difference. If the OP has that covered, then MFA would be the recommended course of action and a trip to Canon service if that doesn't work out.

Posting a few pics would help out quite a bit.

[StanNJ1]

Here are two shots which aren't identical but close in settings and both were shot with the same exact lens on the same day. It is clear to me that the 1D MKII yielded sharper results

[JakAHearts]

I just looked at two of them.

[kfreels]

OK. A 200% crop on an 18MP image is not equivalent to a 200% crop on a 10.1MP image. I'm willing to bet that if the focal length is the same on these two and the head sizes are that close, the subject in the 7D photo was a considerable distance further from the camera.