As as been mentioned before, the choice of RAW converter affects how sharp the files from 7D come out.
With equal settings, DPP produces softer images from the 7D than CaptureOne does. The same does not occur for the 5DII, so it is not a matter of a general bias in one converter.
I posted this conversion of RAW files from IR, 1Ds3, 5DII and 7D in the 7D master thread a while ago:
Remember those images are with equal framing, and notice how little resolution advantage the 5DII has over the 7D then.
Please forgive me for beating this horse over and over again, but it seems that a lot of both users and reviewers use DPP without being aware of such differences, and conclude that the 7D is not capable of delivering sharp images.
Thanks for the comments alundeb and acknowledge that different converters can favour certain cameras (as shown in your linked post). In comparisons like mine I have tried to equalise the processing with a tool that is availble to all photographers - DPP (that, and I don't have any other RAW converter!).
Another test that can be done done is to use the available tools and techniques to squeeze the most out of each image. That would show what each camera is capable of doing, but it then depends on the skill of the person processing the images. If a photographer cannot match the processing (or doesn't have the software) then the results for them are theoretical. I have attempted to use commonly available software to make the test a basis for further comparison.
I could use my (amateur) skills to process the 5D2 and 7D image each as well as possible but then my technique will become part of the comparison and I don't think my visual acuity for subtle changes are up to it.
If I could reshoot (and use f/5.6 as suggested) and make the RAWs available to a skilled person then a good comparison can be made. This would address your issue of RAW processing differences.
That said, my comparison provides a basis for understanding the camera performances in a test that is not often done but is of great interest to reach/lens limited photographers of distant objects. You know through your experience, that you can do better with the 7D image and so that the 7D would come out more clearly on top. Others, with minimal post processing skills/tools, may conclude that the images may be sufficiently similar, so that the 5D2 has sufficient image quality for their purposes.
I doubt there can be a test like this that can give a definitive answer because it comes down to the skills/tools and needs of each individual photographer. The key to such comparisons/tests is to provide the images and the explanations of process so that photographers can decide for themselves.
The above tests are a base comparison with identical RAW processing. No conclusion is made about the validity or usefulness of it, except that it provides a starting point and a rough guide as to how the 5D2 compares to the 7D. (In knowing that the 7D files can do with more/better processing that shown here it still surprised me how well the 5D2 did - I didn't know that before the test).
A further test can be made using DPP to get the best out of each image with that available tool on different settings.
Next a test could involve any RAW tool available to get the very best out of each image and a comparison.
Each of these series of tests can give photographers valuable (to them) information, and I think that each test has its own merits, and I think it would be great if each test could be done. The value of the comparisons will then be down to the individual pphotographer.
I guess your trying to justify a 7D over 5D2, the only problem is in real world bird shooting you wont often be at iso 100 and thats where the 5D2 will stand out.At iso 1600.
Ross
This appears to be a contrarian comment, but I see noticeable differences between the crops when looking at the gray portion of the rooftop.
The curved lines along the right edge are noticeably smoother and more detailed, and I can detect horizontal lines within the tiles in the 7D crops that are not noticeable in the 5D crops. I'm not a birder, but it occurs to me that the parallel curved lines in particular remind me of bird feather structure, and suggests to me that if I *were* a birder, maybe this difference would mean something to me at large print sizes.
(Other than seeing these differences, I have no horse in this race, and I also thought the comments regarding RAW processing and potential differences at higher ISOs to be insightful.)
I haven't compared my 5D II to my 7D directly, but I do know both cameras can reveal incredible amounts of detail, and I disagree with talk of the 5D II's pixels being of much better quality, at least below ISO 800. BTW the sharpened 5D II 150% crop shows obvious pixelation around the front edge of the tiles and does not stand up well at all to sharpening. I would use a 1.4x TC any day of the week inpreference to using interpolation. This topic has been studied quite a bit and the TC (unless it's rubbish) produces better results than uprezzing. I think their would be little difference between a 7D + 300 (say) and a 5D II + 300 + 1.4x in IQ.
Vontom,
Great test! Really appreciate this post as I have often wondered about the same issue that you raise. I recently debated getting a 40D for extra reach. But then I thought, 40D vs 5d2 cropped, would I really see that much of a difference? Your tests do appear to answer that question for me.
The comparison is about the 5D2 and the 7D with the same lens. (eg not 5D2+TC vs 7D which is a common consideration). If I get a TC for the 5D2 I can use it on the 7D as well.
I agree. Quite telling is the visibility of the highest central peak of the 100km crater Theophilus (in the centre of the shadow line). It's 6000ft high and the Sun is just striking its top. The 7D shows it clearly where it is barely visible in the 5D2 image.
Here is a severe USM in PS of 500%, 0.3, 0 of both:
In my view the 7D shows more resolution at basic DPP settings than the 5D2 at 150%, and it clearly has more room for sharpening in post. I'm not surprised that this is the case (18mp crop vs 21mp FF) but I didn't realise how close the 5D2 would be.
The moon may not be a good subject to test the res differences in FL limited cases. Getting a good shot is very dependent on seeing conditions and if they are not good, any advantage one has may be significantly decreased.
Whatever the conditions the 7D is showing more detail, very much so with fine structure.
I like this comparison vontom. Thanks for putting it together. I have a 5DII and my 7D should be here in about 1 week. I plan on testing the two myself, but your testing has shown some of the aspects I will be interested in when comparing these two bodies. I thought about posting some of my findings until I was how you got flamed in the earlier attempt.
An image will look sharpest when viewed at integers of the original: ie 100%, 50, 25, 200% etc Vs 75, 150% etc. This is especially true for curved lines where your image viewer will have to make some averaging choices.
Thus your tests always have the 5D at 150% which is not an optimal viewing size. To flip the tables 1, I would also try the 5DII image at 100% Vs the 7D image at 62.5% (1/1.6) or 65%.
If this changes which image you prefer, then the issue is a resizing one, not necessarily one due to the cameras.
Mike K
An image will look sharpest when viewed at integers of the original: ie 100%, 50, 25, 200% etc Vs 75, 150% etc. This is especially true for curved lines where your image viewer will have to make some averaging choices.
Thus your tests always have the 5D at 150% which is not an optimal viewing size. To flip the tables 1, I would also try the 5DII image at 100% Vs the 7D image at 62.5% (1/1.6) or 65%.
If this changes which image you prefer, then the issue is a resizing one, not necessarily one due to the cameras.
Mike K...Show more →
A cacophony of confusion, in my opinion. The enlargement factors you cite as advantageous are a feature of the monitor. Resizing images using bicubic interpolation removes monitor-related resolution issues.
Pixel Perfect wrote:
The moon may not be a good subject to test the res differences in FL limited cases. Getting a good shot is very dependent on seeing conditions and if they are not good, any advantage one has may be significantly decreased.
Whatever the conditions the 7D is showing more detail, very much so with fine structure.
Hi Whayne, etal,
I agree that using the moon is a dicey choice for making a comparison, as well. Short of shooting from the vacuum of outer space any celestial imaging will be subject to atmospheric disturbance. Is it possible to make a comparison of optics and/or sensors under good seeing conditions...of course it is. But, due to my own photographic dealings with the instantaneous changeability of the atmosphere, I, personally, would shy away from using such images as my evidence.
patrick_morris wrote:
Whats the point of worrying about things zoomed in so far?
The first post indicated the test was relevant specifically for a "lens limited" situation (what you have with you at the time). Suggestions of using different lenses or adding a converter aren't relevant. There are many folks who like to photograph birds or other distant small animals where the combination of subject distance (limited by the subject's skittishness coefficient) and subject size, for a given lens, results in a need to crop, sometimes substantially. That is the same as "zooming in" and that's the whole point of the thread. I also see significant improvement in the 7D shots, especially of the solar heater coils on the roof - also reminded me of the kind of detail one might see in a bird's plumage. If the shots had been taken with a smaller aperture or even the 300 2.8 at 5.6 or 8, the differences might even have been more a bit pronounced. Yet another indication that "more pixels per duck" can often lead to a better image capture. Thanks
I find those 100% and 150% crops extremely close. The most noticable difference is the deeper red from the 5D. Which one more closely matches the real color? If I had to guess I'd say the 7D.
Also, I'm seeing a tiny bit more detail in some aspects of the shot from the 7D. Take a look at the loops in the solar panel. The 7D seems to have ever so slightly more detail. Is it just me?
