5dIII vs 7D shoot out for bird photography at 600mm and cropping

[AlanF]

There are many, many comments in the the threads about the 7D having a 1.6x longer reach for bird and nature photography than the FF. Those of us who have both FF and 7Ds (Neuro et al) reckon that in practice the difference in reach isn't much in practice and the higher IQ of the 5DIII or 1Dx etc gives overall better images. This morning, I decided to have a shoot out with the 300mm f/2.8 II + 2xTC III on a 7D vs a 5DIII. Both cameras were hand held, but resting on the shelf of a hide (blind) at sufficiently high shutter speeds that there was no camera shake. Iso was at 640 and the aperture at f/5.6. I post 6 collage pairs out of many shots, which were are representative. The 7D images are at 100% crop (the actual number of pixels). The 5DIII have the number of pixels increased by 1.5x in each of width and height. All images were taken in RAW, and the sharpness and luminance set at 25 units in PS. The crops were cut and pasted into PS jpegs, and the collages sharpened using USM at 0.5 radius and 100%.

The results parallel what I have found many times in the past.

In terms of resolution
1. At very far distances where the cropped image on the 5DIII is about 200-400 pixels high or wide, the 7D is marginally better.
2. For larger images of say ~500-800 pixels high or wide on the 5DIII, it is at least as good as the 7D, if not better.
3. For images greater than a 1000 pixels, the 5DIII is very clearly superior.

The noise is better on the 5DIII.

 

[AlanF]

Three more. The photos are nothing special, and not very good. They are just to illustrate in a semi-scien tific manner.
The heron in the lake is the closest shot, in the reeds, further away, and the widgeon is at the limits of photography, just to have something that is pixel-limited, like the duck in the previous.

 

[Click]

Very interesting post. Thanks Alan.

 

[candc]

http://iwishicouldfly.com/iwishicouldfly/journal/html/020112b.html

this was an article that was posted in another thread where this subject was being discussed, its a little different in that it compares the5dii, 1div, and 7d. all the images are cropped somewhat to get to a target. the 7d cropped the least 1div in the middle and the 5dii the most. if you read the article you will see that the results are pretty much identical.

 

[Ruined]

Very insightful test.

One additional point to consider - if you were taking say 500 photos, it would be a lot more work to have to go through and crop all of them in post compared to simply framing them in camera as you wish to begin with. So even if quality is no better the crop does have an advantage for this type of photography if you plan to take a significant amount of images.

 

[Don Haines]

Your results are similar to what I found on a comparison between the 5D2 and the 60D using a 300F4..... The 5D2 was better in closer while the 60D was better on far away items.... What I really found shocking was that for very far away subjects, the SX50 outperformed both!

 

[privatebydesign]

Pretty hilarious post considering Alan was one of the biggest 1.6 tele advantage advocates out there, until he got a FF camera.

I have been given so much grief and negative feedback for saying the same thing for years and showing my 7D and 1Ds MkIII crops. Welcome to real world meets theoretical pontificating.

 

[surapon]

There are many, many comments in the the threads about the 7D having a 1.6x longer reach for bird and nature photography than the FF. Those of us who have both FF and 7Ds (Neuro et al) reckon that in practice the difference in reach isn't much in practice and the higher IQ of the 5DIII or 1Dx etc gives overall better images. This morning, I decided to have a shoot out with the 300mm f/2.8 II + 2xTC III on a 7D vs a 5DIII. Both cameras were hand held, but resting on the shelf of a hide (blind) at sufficiently high shutter speeds that there was no camera shake. Iso was at 640 and the aperture at f/5.6. I post 6 collage pairs out of many shots, which were are representative. The 7D images are at 100% crop (the actual number of pixels). The 5DIII have the number of pixels increased by 1.5x in each of width and height. All images were taken in RAW, and the sharpness and luminance set at 25 units in PS. The crops were cut and pasted into PS jpegs, and the collages sharpened using USM at 0.5 radius and 100%.

The results parallel what I have found many times in the past.

In terms of resolution
1. At very far distances where the cropped image on the 5DIII is about 200-400 pixels high or wide, the 7D is marginally better.
2. For larger images of say ~500-800 pixels high or wide on the 5DIII, it is at least as good as the 7D, if not better.
3. For images greater than a 1000 pixels, the 5DIII is very clearly superior.

The noise is better on the 5DIII.

Dear Sir , Mr. AlanF.
Thousand Thanks for you Testing and Spend a lot of your time to Show and Teach us.
Just One more thing that I would like you to do for us again---Use That two Beautiful systems and Shoot the Doll or the Colorful Jar or Bottle of Wine ( The Doll or the Bottle of Wine can not Move), And The Cameras On the Tripods( With Remote Control or Self Timer) = That will show the real Sharp of the same photos with out motion blue of the real Birds.( Yes, Sir, I Know that your Strong Hands and do not create the Motion Blur when your press the shutter)
Again, Thanks you, Sir.
Surapon

 

[candc]

Your results are similar to what I found on a comparison between the 5D2 and the 60D using a 300F4..... The 5D2 was better in closer while the 60D was better on far away items.... What I really found shocking was that for very far away subjects, the SX50 outperformed both!

I have to get one of those, I go on back country canoe trips and it would be perfect for wildlife. I think Panasonic makes one with a faster lens but the canon is 24-1200 equivelant, hokely smokely!

 

[ahab1372]

Very insightful test.

One additional point to consider - if you were taking say 500 photos, it would be a lot more work to have to go through and crop all of them in post compared to simply framing them in camera as you wish to begin with. So even if quality is no better the crop does have an advantage for this type of photography if you plan to take a significant amount of images.

When it comes to Wildlife photos, I always find myself cropping, even if using a APS-C camera The animals are always too far away, and lenses are always too short. So the amount of work is about the same for me

 

[Sporgon]

Both cameras were hand held, but resting on the shelf of a hide (blind) at sufficiently high shutter speeds that there was no camera shake. Iso was at 640 and the aperture at f/5.6. I post 6 collage pairs out of many shots, which were are representative.

Not quite convinced by the method here. The light doesn't look that bright on most of the pictures, so ISO 640 and f5.6 I am guessing would give a shutter speed at best of 1000th, more likely 500th. Unless you were resting on a suitable support I'd be worried that the stability wasn't reliable enough on a 600mm focal length for a very subtle test such as this one.

 

[dgatwood]

No matter what sensor you use, you can never really out-resolve your lens. What I think you're seeing here is that the 300mm L II just isn't a good enough lens to allow the 7D to outperform the 5Dmk3. If you believe the DXO analysis, they rate it at just 21 MP of resolution on a full-frame camera. Assuming those numbers are correct:

1. You're only effectively getting about 8.4 MP of usable resolution out of your 7D when used with that lens.
2. The 5Dmk3 exceeds the maximum angular resolution of the lens, so no other sensor can significantly outperform it.

A 7D's sensor, if scaled up to be full-frame size, would provide a whopping 45 megapixels. Thus, to fully take advantage of the 7D's resolution, your lens would need to be capable of resolving 45 megapixels on a full-frame sensor. If a sufficiently sharp lens existed, in theory, the 7D would mop the floor with a cropped image from the 5Dmk3 when used with that lens. Unfortunately, I don't think such lenses exist yet, and probably won't until Canon decides to build a high-megapixel full-frame camera.

Of course, in practice, the DXO analysis is, IIRC, limited by the resolution of the camera used to take the measurements, and may not adequately reflect the true maximum angular resolution of the lens itself. In other words, take everything I just said with a grain of salt.

 

[candc]

Very insightful test.

One additional point to consider - if you were taking say 500 photos, it would be a lot more work to have to go through and crop all of them in post compared to simply framing them in camera as you wish to begin with. So even if quality is no better the crop does have an advantage for this type of photography if you plan to take a significant amount of images.

When it comes to Wildlife photos, I always find myself cropping, even if using a APS-C camera The animals are always too far away, and lenses are always too short. So the amount of work is about the same for me

I don't like shooting with the intention to crop later, I would rather use a longer lens but there is a limit to what you can do. I am using the sigma 120-300 with the canon 2xiii to get to 600 on a crop body and that gets pretty far out there and the results are pretty good as long as you don't have to crop further, then its not acceptable to me. I would really like to get the 600ii I think you can crop 50% with that lens and still get good results on either format.

 

[chauncey]

I do a fair amount of BIF images of GBH's and Snowy Egrets using an older 300 f/2.8, with/without a 1.4 TC, mounted on an equally old Ds3, were those images mine...they would have been discarded in the initial culling process. But then, I'm one anal retentive, pixel peeping, SOB.
To be fair, I just got a new 2X IIII TC yesterday and haven't used nor micro-adjusted the focus yet but, if it preforms like the above examples...it'll be returned forthwith.

 

[candc]

those are 100% crops, with a 2x converter, that's excellent results

 

[Canon1]

What was your shutter speed Alan?

 

[Hannes]

I imagine the results may look different if you didn't have the 2x converter on which will limit resolution a lot

 

[takesome1]

It is not just the camera hardware you should compare but the firmware as well.
Testing at the same PP settings do not do this.

The 7D files have more room for improvement and sharpness in post processing.
A test that would compare the two bodies fairly would be a test of the photographs after normal PP.

That is not to say you will find anything much different that others have already found or that you believe you found.
These type of tests have been done over and over ever since the 7D was released.

When and if a 7D II ever arrives I am sure we will get to see a whole new series of comparisons and tests.

 

[Skulker]

Interesting post. Your results are much like what I get in so much as I generally like the results from the 5D3 over the 7D.

The only one I found surprising was the second were the 7D seems to have an advantage in low light. That's not what I find.

Thanks for sharing your results its interesting to see real world results rather than people trying to justify their opinion of the spec sheet. ;D

 

[Don Haines]

Your results are similar to what I found on a comparison between the 5D2 and the 60D using a 300F4..... The 5D2 was better in closer while the 60D was better on far away items.... What I really found shocking was that for very far away subjects, the SX50 outperformed both!

I have to get one of those, I go on back country canoe trips and it would be perfect for wildlife. I think Panasonic makes one with a faster lens but the canon is 24-1200 equivelant, hokely smokely!

While I like the incredible zoom range on the SX-50, be warned that as a p/s camera it has a long lag time when you press the shutter, it takes a long time between pictures, the ergonomics S__K compared to a DSLR, you WILL have to do noise reduction on SX-50 pictures, and it's EVF is most definitely NOT current technology.... That said, you can pull off some good shots if you get used to it...

The following comparison shots are between a 60D with a 70-200 at 200mm and the SX-50 set to 1200mm(equivalent). The first shot is the full frame from the 60D, the second shot is the full frame from the SX-50 (both scaled to fit here), the third shot is the jay from the 60D cropped to 800 pixels wide, the fourth shot is the jay from the SX-50 cropped to 800 pixels wide.

There are conditions where a p/s can beat a DSLR and a chunk of L-glasss