*UPDATE* A Bit of EOS 70D Info [CR1-CR2]

[Zv]

I did a few more tests and what I found out is this - even when I cropped the 5D II image to match the 7D one the full frame image looked crisper, was 2/3 of a stop brighter, a bit colder in tone and showed less CA. Due to the effect of compression with a tele lens on the 7D the subject looked bigger in relation to background than the cropped version of the 5D II. Naturally. Though due to this its hard to determine if one is resolving more than the other. To my eye I could see tiny cracks and hairs on the books on both images. If anything I have now realized my 135L works better on my 5D II!

[jrista]

I did a few more tests and what I found out is this - even when I cropped the 5D II image to match the 7D one the full frame image looked crisper, was 2/3 of a stop brighter, a bit colder in tone and showed less CA. Due to the effect of compression with a tele lens on the 7D the subject looked bigger in relation to background than the cropped version of the 5D II. Naturally. Though due to this its hard to determine if one is resolving more than the other. To my eye I could see tiny cracks and hairs on the books on both images. If anything I have now realized my 135L works better on my 5D II!

It would be helpful if you provided full size files with EXIF. Sorry, but your findings just don't fit the facts, the math, or the theory, otherwise I wouldn't ask. Did you configure AFMA for the same lens on both bodies (it is possible the lens was not calibrated properly for the 7D.) Anyway, images w/ EXIF would be nice, as I know my stuff in this area pretty well.

I really need to just bite the bullet and buy the 5D III. I've been waiting for the 7D II to be released, to see if it would fit my needs better, but I'm tired of renting gear all the time to prove my points...

If it helps any, Roger Clark, a well respected scientists with I believe multiple Ph.Ds, has an excellent page demonstrating how much more resolving power the 7D has in focal-length limited scenarios, and the 5D II is one of the cameras tested and compared:

http://www.clarkvision.com/articles/telephoto_reach/index.html

The second image on the page shows the moon shot with the same lens (300mm f/2.8 L) on four different cameras. The differences in resolving power for each sensor are more than clear. The 7D moon is about four times larger than the 5D II moon, and the additional detail is very clear.

Here is another link, where Roger covers noise and high ISO:

http://www.clarkvision.com/articles/pixel.size.and.iso/index.html

To quote his assessment:

Here is my assessment:

In all the images, the 5DII images fail to show the subtle color differences that the 7D and 1D4 show. The color in the 1D4 and 7D are very close (until noise hides it).

ISO 100: 7D noise is small and detail is well above other images. 7D=top, 2nd=1D4

ISO 800: 7D noise is showing, but the detail is still well above the other cameras. 7D=top, 2nd=1D4

ISO1600: 7D noise is becoming prominent, but image detail is still very good. 7D=top, 2nd=1D4, but the difference is narrowing.

ISO3200: 7D noise is becoming objectionable and color is getting lost, in particular in Mare Serenatatis (the large circular dark area in the upper center). top=1D4, 2nd 7D. A good down sampling algorithm (like 2x2 pixel average) could improve the the image.

ISO6400: Noise is too apparent in 7D, and 5DII (which is slightly older technology than the 7D or 1D4). Top=1D4, 2nd=5DII. In my numerous sensor evaluations, I consistently see the 1D series sensors have fewer hot/bad pixels and the images here show that too: the 7D and 5DII images have a lot of "spiky" noise not seen in the 1D4 image.

In all the images, if we boost the low level, we will see that all the 7D and 5DII images have a lot of fixed pattern noise, which decreases as ISO increases. The 1D4 has a little fixed pattern noise at low ISO which quickly decreases at intermediate ISOs. The noise is illustrated below.

It is only when he reaches ISO 6400 that the 5D II even really becomes a contender. At ISO 3200 the 1D IV finally edges out the 7D. At all other settings, the 7D has a considerable resolution advantage that definitely shows. These findings jive with my own experiences...the 7D is an excellent camera that far outresolves most other DSLR cameras (with the exception of only a couple new ones from Nikon only recently released (D3200 & D5200)).

No matter how you slice it...the 7D trounces the 5D II and the 5D III in terms of spatial resolution. Assuming you are not noise bound, in any focal-length limited situation, the 7D will produce better results.

[Zv]

It's just not even worth posting the images with exif info as in the real world it makes little or no difference to me. I'm quite happy with how both cameras perform. It was for my own curiosity that I did the test. I am sure the professors know a lot more than me but when I'm out shooting portraits or weddings it will not matter one bit. The difference at the pixel peeping level is minor for an average shot of a subject 10 feet away. If I want to take pictures of the moon I shall almost certainly use the 7D as it has more reach and will likely out resolve the 5D II in that situation. (I have taken pictures of the moon with the 7D before). Right tool for the job etc etc. To me lighting and good technique are far more important things to worry about. Both cameras resolve quite enough for me!

Have a great weekend everyone and have fun shooting great pictures!

[jrista]

I was curious by all this discussion so I took two shots. One with the 7D and one with the 5DII to see how they compare in overall IQ. I tried to keep things constant including framing and settings. The 5D II file looked sharper and brighter. The 7D file was pretty close though. Pretty obvious test but hey, now I know for myself.

(I should add that the out of focus blur is more prominent in the 5D II image due to the lens being closer to subject and aperture being the same - f/3.5, as on the 7D.)

the 7d has a strong aa- filter, the 5dmk2 has a week aa-filter therefore 5dmk2  looks sharper,  and there are a  Mp difference between them and the lens to 7d needs to be 1,6 times  better in resolution and contrast.
so it is among other things an optical question

I believe the notion that the 7D has a strong AA filter is a myth. I've used the 7D with all of Canon's top-end Mark II L-series super-telephoto lenses except the 400mm f/2.8. In every single case, the sharpness and resolving power of the 7D was considerably greater than any other sensor I've used, either my own cameras or rented cameras. The "softness" of the 7D, in my opinion, has nothing to do with a strong AA filter. It has to do with either poorer-quality glass (i.e. the 100-400mm f/4.5-5.6 L IS is NOT a particularly sharp lens, but is most frequently paired with a 7D), or poor camera handling (camera shake is more likely to cause IQ-degrading softening with the small 7D pixels than the large 1D X pixels.)

I have searched high and low for some actual concrete evidence that the 7D has a strong AA filter. Like someone actually measuring the thickness of the low pass plates in the filter stack over the sensor. I have yet to actually find any REAL evidence that the 7D has a particularly strong AA filter. As such, I'm inclined to say that is just an internet myth, that has arisen from the kinds of lenses most commonly used on the 7D...things that fall into the same budget range such as the 100-400m 70-200, 70-300 L, etc. None of those lenses compare to the new Mark II Great Whites...and when you DO combine the 7D with truly top-end still photography glass, there isn't any softness to be seen anywhere.

Here is a shot with the EF 500mm f/4 L IS II with a 1.4x TC III (700mm) on the Canon 7D:



Here is the 100%, 1:1, pixel peeping crop:



^--- RAZOR-FRIKKIN SHARP!! --^


7D "Strong AA filter" == Total Myth!

tell that to John Sheehy who have shown that that is not the case

I see...and the full resolution visual evidence I've just provided you is...what, all an illusion? LOL, you guys are INCREDIBLE!!! What exactly does it take to prove a point to you guys? Even Zv! It is obvious he sees a difference between the cropped 5D II and 7D, but no images will be uploaded because it would support the only argument that actually fits the math and the theory! That the 7D image is sharper and clearer than the 5D II image.

Zv, the argument is not whether the 5D II is good enough. If it is good enough for your work, that's GREAT! The argument was just about the simple facts and realities of things. I was not trying to make this personal, just keep things honest and factual, and I had hoped since you had both cameras, you could supply a couple images to prove the point. I've never cared for Mikael/Ankorwatt, he's demonstrated a rather astonishing ability to ignore physical evidence dangled a few millimeteres right in front of his nose. I guess I had hoped that most people were more reasonable than that, and that physical evidence would be enough to demonstrate that all that worthless theory and math was not quite so worthless after all. Instead, I just get more backpeddling and obfuscation from you, too?

What in the world happened to a little honest objectivity, ppl? Is the entire world so wrapped up in themselves that they can no longer present or even participate in a reasoned debate, or acknowledge the facts and not take it personally when those facts reach up and smack them in the face? Wow. Just...wow.

[Zv]

Very well if you insist. Here is a side by side shot. one the left is the 7D unaltered image and on the right is the 5D II cropped to look like the 7D. No other adjustments or corrections were done in Lightroom. I took a screen dump and made a jpeg of it. I used a 135L @ f4 and ISO 100, 1/25s. A tripod was used to stabilize the camera and MF via live view mode was used to achieve optimal focus on the face on the Dumas book. This is how it looks without zooming in.

[jrista]

And now lets zoom in and see. What conclusions can we draw from this?

I can only conclude the 5D II is not upscaled properly, as the books are smaller. There seems to be no doubt the 7D contains more information and detail, though...the books are quite a bit larger (which means every aspect of detail, such as the cracks, contain MORE pixels than the 5D II shot.) Anyway, I give up. Everyone seems to have an agenda, and that agenda doesn't include objective fact. Sorry I bothered you.

[jrista]

And now lets zoom in and see. What conclusions can we draw from this?

I can only conclude the 5D II is not upscaled properly, as the books are smaller. There seems to be no doubt the 7D contains more information and detail, though...the books are quite a bit larger (which means every aspect of detail, such as the cracks, contain MORE pixels than the 5D II shot.) Anyway, I give up. Everyone seems to have an agenda, and that agenda doesn't include objective fact. Sorry I bothered you.

All I did is click compare and then zoom, I dunno how LR handles its business but like I said earlier this is likely due to compression. No worries, I am not bothered in the slightest. 

Pop both of those into Photoshop, and scale the 5D II shot so it is the same image dimensions as the 7D. Stick the 5D II upscaled image in a new layer in the 7D image, and toggle it on and off. I think the difference between the two, and what I mean by "resolving power", will become quite clear then.

[Zv]

And now lets zoom in and see. What conclusions can we draw from this?

I can only conclude the 5D II is not upscaled properly, as the books are smaller. There seems to be no doubt the 7D contains more information and detail, though...the books are quite a bit larger (which means every aspect of detail, such as the cracks, contain MORE pixels than the 5D II shot.) Anyway, I give up. Everyone seems to have an agenda, and that agenda doesn't include objective fact. Sorry I bothered you.

All I did is click compare and then zoom, I dunno how LR handles its business but like I said earlier this is likely due to compression. No worries, I am not bothered in the slightest. 

Pop both of those into Photoshop, and scale the 5D II shot so it is the same image dimensions as the 7D. Stick the 5D II upscaled image in a new layer in the 7D image, and toggle it on and off. I think the difference between the two, and what I mean by "resolving power", will become quite clear then.

Wait a min I see what went wrong ... OK yeah you're right I only zoomed and it zoomed an already cropped image so it looks different. I will try the photoshop route.

[Zv]

OK here we go, upscaled on photoshop (which is destructive so .... anyway).

[jrista]

OK here we go, upscaled on photoshop (which is destructive so .... anyway).

I wouldn't necessarily call upscaling "destructive". It is distributive, for sure, as it distributes existing information and fabricates new information...but I wouldn't necessarily call upscaling destructive. It is a better way to clearly demonstrate the difference, which I think is clear now with your latest screenshot. You could downscale the 7D image to the 5D II crop size. The differences won't be as apparent, as downscaling IS definitely destructive, however the 7D shot will pick up additional clarity and sharpness, as well as reduced noise relative to the 5D II shot.

Anyway, thanks for being honest!

[Zv]

OK here we go, upscaled on photoshop (which is destructive so .... anyway).

I wouldn't necessarily call upscaling "destructive". It is distributive, for sure, as it distributes existing information and fabricates new information...but I wouldn't necessarily call upscaling destructive. It is a better way to clearly demonstrate the difference, which I think is clear now with your latest screenshot. You could downscale the 7D image to the 5D II crop size. The differences won't be as apparent, as downscaling IS definitely destructive, however the 7D shot will pick up additional clarity and sharpness, as well as reduced noise relative to the 5D II shot.

Anyway, thanks for being honest!

You're welcome! Glad we sorted that out. 

[jrista]

Jrista
search : John Sheehy 7d aa-filter and you get answer

this is from dpreview, 3 aps and  the 7d , 7d has a stronger aa-filter like Nikon often have before  compared to for example  Pentax APS who use the same sensor as Nikon but lighter filter, d7100 has no filter and the resolution is also higher



dpreview resolution test shows 2600 LPH for the rebel   and 2500 LPH for the 7d

Lets qualify things here. The 7D has a "stronger" AA-filter than the 650D, 60D, and D7100. STRONGER THAN X. I need to be extremely clear here that having a "stronger" AA filter than any one of those cameras by no means indicates the 7D has a TOO STRONG AA filter. The sentiment that you were pushing was that the 7D is losing IQ because of it's AA filter, an AA filter that is "too strong".

I strongly dispute that notion. When it comes to digital photography, we want an AA filter to be ideally suited for the specific sensor you are using. You don't want it to be too strong, and neither do you want it to be too weak, ESPECIALLY if you shoot anything that might produce aliasing or moire. Aliasing and moire are BAD things...and in your example shot, the 7D image is the ONLY one that looks acceptable to me...it reproduces the information present in the subject being photographed better than all the others. It should also be noted that the softening caused by an OLPF is predictable softening, while it eliminates an unpredictable outcome...moire. Softening is global, and thus something we can easily correct in post with a little bit of sharpening. Correcting moire is a far more difficult task, and it must be performed locally rather than globally, only to regions actually affected by it. The concept here is no different than diffraction...stopping down to gain necessary DOF is preferred over shooting wide and ending up with a thin DOF. Correcting for diffraction in post is easy because it is global and linear, correcting for an improper DOF is practically impossible because it is localized and non-linear.

I believe my 100% crop photo of the orange-morph house finch....a bird, with lots of criss-crossing feathers and color detail...looks nearly perfect! No moire at all, no aliasing, however the detail level is exquisite. The 7D has neither a too-weak nor a too-strong AA filter. It has an AA filter that is just about PERFECT for an 18mp APS-C sensor with a 4.3µm pitch.

I stand by my statement. The notion that IQ on the 7D is lost because of an AA filter that is too strong is a myth.

[Zv]

Bryan from TDP notes that if you increase sharpness to a setting of 1 in the camera picture style profiles the image looks just fine and is about same as the 550D, 60D etc. It is a non issue. I think the 7D images are plenty sharp and most of the time I just add a little sharpening in post anyway. He also states that this COULD be due to a STRONGER (as jrista said) AA filter, but nothing about it being too strong.

[GMCPhotographics]

Jrista
search : John Sheehy 7d aa-filter and you get answer

this is from dpreview, 3 aps and  the 7d , 7d has a stronger aa-filter like Nikon often have before  compared to for example  Pentax APS who use the same sensor as Nikon but lighter filter, d7100 has no filter and the resolution is also higher

I personally found my 7D to produce slightly soft per pixel detail when compared to my 5DII/III. I've read that Canon used a stonger than usual AA filter to reduce moire in movie mode. It was a very nice camera to use, but it's IQ wasn't on par with Canon's full frame offerings. It had very high iso noise, slightly odd micro contrast and the pixels were softer than any other DSLR I've used. I loved the 8fps, very capable AF system and great handling.


dpreview resolution test shows 2600 LPH for the rebel   and 2500 LPH for the 7d

Lets qualify things here. The 7D has a "stronger" AA-filter than the 650D, 60D, and D7100. STRONGER THAN X. I need to be extremely clear here that having a "stronger" AA filter than any one of those cameras by no means indicates the 7D has a TOO STRONG AA filter. The sentiment that you were pushing was that the 7D is losing IQ because of it's AA filter, an AA filter that is "too strong".

I strongly dispute that notion. When it comes to digital photography, we want an AA filter to be ideally suited for the specific sensor you are using. You don't want it to be too strong, and neither do you want it to be too weak, ESPECIALLY if you shoot anything that might produce aliasing or moire. Aliasing and moire are BAD things...and in your example shot, the 7D image is the ONLY one that looks acceptable to me...it reproduces the information present in the subject being photographed better than all the others. It should also be noted that the softening caused by an OLPF is predictable softening, while it eliminates an unpredictable outcome...moire. Softening is global, and thus something we can easily correct in post with a little bit of sharpening. Correcting moire is a far more difficult task, and it must be performed locally rather than globally, only to regions actually affected by it. The concept here is no different than diffraction...stopping down to gain necessary DOF is preferred over shooting wide and ending up with a thin DOF. Correcting for diffraction in post is easy because it is global and linear, correcting for an improper DOF is practically impossible because it is localized and non-linear.

I believe my 100% crop photo of the orange-morph house finch....a bird, with lots of criss-crossing feathers and color detail...looks nearly perfect! No moire at all, no aliasing, however the detail level is exquisite. The 7D has neither a too-weak nor a too-strong AA filter. It has an AA filter that is just about PERFECT for an 18mp APS-C sensor with a 4.3µm pitch.

I stand by my statement. The notion that IQ on the 7D is lost because of an AA filter that is too strong is a myth.

[jrista]

it looks stronger than other filter  canon have used together with theirs newer  18Mp cameras

And yet, it is neither too strong nor too weak. To turn the argument around, all three of the other cameras in your sample shot have AA filters that are too WEAK. As a result, you experience moire. Moire is something we have no decent tools to correct in post. The tools we do have are tedious, and must be applied locally. A too-weak AA filter is worse than a just-right or too-strong filter, as when you have a just-right and too-strong filter, all it takes is a little global sharpening to correct in post. Nature is FULL of interfering patterns.

It is relatively rare to not have the need for an AA filter. The cases where you might be ok are most landscapes, or if you  regularly shoot below the DLA (in which case, diffraction will soften enough to act as the AA filter itself.)
However even in nature, there are plenty of repeating or nearly repeating patterns. Birds feathers are an excellent example. Fur on wildlife. Striated detail in leaves or other natural materials that often show up in macro photography. There are also plenty of things that just experience aliasing, not necessarily moire, such as twigs in trees, power lines, strong edges on any surface or structure, etc. These are all things we have to correct in post if we don't have an AA filter or have an AA filter that is too weak.

I've never seen meaningful, visible softening in the 7D that was not due to a lens. When I use top-end telephoto glass, I see no softening whatsoever. That tells me that the bulk of "softness" people experience is due to the lens, or an incorrect AFMA setting. If you use high quality glass, softness just doesn't occur. It may not be hypersharp either, but it is still sharp as a tack every single time, and if I want things to be even sharper, I can always add just a little sharpening in post. Note, here, that the shot of the finch with the 500/4 L II + 1.4x TC was NOT sharpened. That sharpness is strait out of the camera.