Canon EOS 7D tested, retested and reretested

marc_bergman1 · November 13, 2009

"What a perfect way to generate hits to a website. Come up with a testing method that isn't quite right, yet shows a controversial result, then post the results and wait for the tidal wave of responses.

Brilliant."

Isn't this the Ken Rockwell method?

Oh and. Please send money.

dmanthree · November 13, 2009

At least Rockwell makes no bones about his site, and says flat out that it's total BS and not based in fact or reality. This guy does not and presents results that are, well, questionable, in the very least.

images_in_light_north_west · November 13, 2009

Darwin wiggett is a well respected landscape photographer, if your going to bash some one know who you are bashing.

Ross

py-photography · November 13, 2009

Pointing out errors isn't bashing.

There is no way, a lens will perform the same on every body.. its just not possible.

No two things can be made exactly the same.

Because someone can take photos doesn't mean s/he can perform a valid test.

angel_bocanegra · November 13, 2009

This wiggett joker, judging by his biased review and unfair approach to minimize the 7D, he doesn't get my respect as a reviewer no matter how good his pictures are.

philip_wilson · November 13, 2009

Darwin Wiggett is a top class photographer - if you doubt this read one of his books. I am uncertain of the 7D image sharpness and did my own test. The camera does appear to have a different image quality to other Canon DSLRs. I was surprised that mine was much closer to my 5DII than I expected. I am not sure what conslusions to draw from all this. I suspect that the 7D is very sensitive to post processing and also that the may be considerable sample variation. My conclusion from my tests is that the 7D is clearly nosier than the 5DII, that the difference in sharpness when I pixel peep is less than I thought as the printed 5DII images clearly look better than the 7D images. Finally I think my 5DII is sharper than Darwins examples but it does tend to the same type of "flat" images that his test examples produced.

angel_bocanegra · November 13, 2009

I don't doubt that Darwin is a top class photog, but going all the way to say that the g11 is actually sharper than the 7D? First you cant compare SLR with a point and shoot with a fixed lens. Could it be that Canada got a bad batch of 7D's? Its weird how all three 7D'd tested came out flawed. I must admit I had to exchange one, but now I am absolutely pleased with my new toy.

py-photography · November 13, 2009

I got one of the first 7D's in Canada and mine has been great.

Darwin takes great photos no doubt about that. But I do think his review is based on getting hits to his site, in the hope more people will sign up for his workshops.

It was another marketing ploy, nothing more.

As Angel said, comparing the G11 to a dSLR is ridiculous, whats next comparing detail between the 7D and LF camera...

scott_ferris · November 13, 2009

Angel, Peter, he is not the first to do so. Though I think Darwin did a pretty poor job I don't think Michael Reichmann did.

http://www.luminous-landscape.com/reviews/kidding.shtml

Incidentally whilst I am a big believer in bigger sensors better, always, at smaller sized prints I have some pretty amazing G10 images that do compare well with my 1Ds MkIII, but only in optimum conditions with the right type of subject and at base iso, not where one usually uses a P&S!

py-photography · November 13, 2009

Scott, Thanks for the link, very good read. As you said if everything is right, then they compare. But how often is everything right or optimal? I don't put food on the table taking pictures, and never will. Some of the things Darwin said in his review are ridiculous. Example I know where my lens are the best... That made me laugh and he lost all credibility at that point. I work in manufacturing, (MET) and one of the first things you learn is nothing is perfect and no 2 things are exactly the same. Its simply not possible... Its why we have tolerances and sample size etc

julian_robinson4 · November 13, 2009

Sheesh there's a lot of people who lay blame and see mythical problems with this test. Any test is tricky to conduct but the ones that count are those that reasonably accurately predict what you are likely to find in real-life shooting. This test is a good attempt at that by a competent photographer who knows what he is doing.

Just look at the side by side examples and you'll see reasonable facsimiles of what you well get in real life, and a reasonable idea of how much you might be able to tweak each camera for sharpness etc.

First post David Amberson is exactly right - I too have looked at one 7D and hundreds of photos from 7Ds and none of them is as sharp as it should be, and none of them is sharp enough to justify 18MP. You simply don't gain anything from the extra pixels in this camera. Some examples are awfully soft.

I agree too as a theory that Canon's AA filter seems to be the cause. In fact I'd suggest it may be deliberately so to ensure that 7D photo IQ doesn't impinge too much on 1D's and 5D's. This might be fair enough by Canon if it had no cost (though I don't think it is fair) but even if so, then they should not burden photographers with extra MP's that are essentially unusable -- Canon should be honest and say "we have detuned this and made it a 12MP camera (or whatever)".

There is some support for this from the-digital-picture.com 's interesting comparison page.

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?SampleComp=0&FLI=0&API=0&Lens=458&Camera=673&FLIComp=0&APIComp=0&LensComp=458&CameraComp=460

In my view the 7d is worse at a pixel level than every other camera shown. That is, on a pixel to pixel basis and not allowing for how many pixels. What this shows is that the machine has been de-tuned from maximum possible sharpness, or some other limitation is preventing the advantages of more pixels being utilised, or there is a problem with the test images being compared.

If you look at the images from each camera with a given lens and f-stop, you'll see that the 7D is softer per pixel. The diagonal lines show what seems to be stronger AA filtering than others. This SHOULD be changed by Canon.

The softening shown there seems a lot less than Darwin found so more investigation needs to be done.

Ben Goren says " The only fair test is to make same-sized prints with the various systems, which is exactly equivalent to resampling, and which we all know requires sharpening." I disagree - an equally fair test is to render the same FOV onto an equal screen area and view it on screen. Resizing to show same FOV is identical to printing to same size. The screen is where many images are viewed these days. It is relevant only if your image on screen is heavily cropped because reducing a full size image from any old camera to screen size will give reasonable apparent sharpness so nothing would be learned. But heavily cropped photos will demonstrate true camera performance, the same as you might learn from a large print.

I do think Darwin has buggered up some of the comparisons by comparing images not re-sized which I would argue don't tell you much, but the equal or nearly equal FOV samples are powerful evidence and tell you nicely what one camera is capable of compared to another. The weakest point of any comparison is removing the misleading effects of unequal sharpening, but this applies to any comparison and if you look carefully you can usually see how much detail is actually in the image regardless of sharpening.

THe other thing we need to remember is that Canon are NOT good at production control and QA of critical performance aspects of consumer grade cameras. Well, any cameras at the moment. The variable results found in reviews and tests can often be attributed to the reviewer getting a surprisingly good copy, or a surprisingly bad one (or two or three). Don't shoot the messenger!

In summary - Darwins is a good test that agrees with the general thrust of many other analyses of the 7D's critical performance. Canon should address this test and assist Darwin (and us) to find out what's wrong.

walter_strong5 · November 14, 2009

<a href="/photodb/user?user_id=426229">Marc Bergman</a> <a href="/member-status-icons"><img title="Subscriber" src="http://static.photo.net/v3graphics/member-status-icons/sub8.gif" alt="" /></a>, Nov 12, 2009; 12:17 p.m.

"To be fair they may live in countries where post-processing is illegal".

WHAAAAAAAAT?

marc_bergman1 · November 14, 2009

Julian,

I don't think the link to the-digital-picture.com shows anything in your favor.

Compare the results to the tests in DPReview for the Canon 7D and the Canon XSi.

Canon 7D

----------------Absolute Resolution-------------Extinction Resolution

Horizontal -------- 2500--------------------------------3100*

Vertical-------------2450--------------------------------3050*

Canon XSi

---------------Absolute Resolution--------------Extinction Resolution

Horizontal - ------2300--------------------------------2500*

Vertical------------2200--------------------------------2500*

* Moire visable

Now compare to the examples on the-digital-picture.com. Look at the actual chart values. Look for clear lines. With the 7D I am seeing good data up to around 30 on the chart. With the XSi I am seeing good data just past 24 on the chart. Both tests show comparable data. They clearly show the 7D has higher resolution. The tests from Image Resource also show comparable data.

Now to show the effects of diffraction make both cameras the 7D. Set one side for f /4.0 and the other to f /16. Note how the resolution is better at f /4.0. You can also see how the contrast is lower at f /16.

The Canon 7D clearly shows higher resolution than the XSi and the 50D. Canon has been able to provide this resolution while keeping the noise in check. I look forward to seeing examples once Adobe Camera RAW is beyond the beta stage and the files are properly prepared.

marc_bergman1 · November 14, 2009

Walter,

That is what we, in my country, call a joke. They aren't always funny, if you know what I mean.

keith reeder · November 14, 2009

First post David Amberson is exactly right - I too have looked at one 7D and hundreds of photos from 7Ds and none of them is as sharp as it should be, and none of them is sharp enough to justify 18MP.

</blockquote>

He really isn't. Both you - and he - need to put your prejudices aside take a proper look. There's a wealth of evidence and opinion out there that clearly and convincingly contradicts your comment (including many of the examples at The Digital Picture - and it's pretty obvious too that the ones that look less sharp are focused differently).

The fact that you choose to ignore the existence of that evidence is telling...

You simply don't gain anything from the extra pixels in this camera. Some examples are awfully soft.

</blockquote>

In which it differs from any other camera you care to mention how exactly? I've seen "soft" from every camera out there - it doesn't mean they're soft cameras, and to imply that it does absolutely screams bias and hidden agenda.

Even in this thread there are comments like this:

I received my 7D on 10/1/09. At first I was apalled at the focus issues and noise. Then I learned how to use the camera, being a new focus system. One of the things I shoot are stunt hawks. I am consistently impressed with the focus tracking ( when it is set correctly for the job). I came from a 40D, so I am also impressed with the resolution. I was never able to get away with the extreme crops I have to do sometimes with these birds.

</blockquote>

Which unequivocally undermines any validity the review in question has.

Of course, the most important thing in that quote is:

Then I learned how to use the camera.

</blockquote>

Some of the bashers here might take note...

keith reeder · November 14, 2009

Sheesh there's a lot of people who lay blame and see mythical problems with this test.

</blockquote>

Nothing mythical about diffraction - Wiggett complains about softness at f/16, despite the fact that at that aperture the airy disc will be two pixels wide!

It's just a bad review, where the camera is simply not being used properly given its specification and design characteristics.

keith reeder · November 14, 2009

As an aside, I've just read the 7D review in the UK's PhotoPlus magazine.

It is awarded 97%.

I've never seen a higher score in that magazine.

The reviewer raves about its image quality across the ISO range (the "harshest" thing they have to say there is that above 6400 ISO you lose some detail!); the AF is as good as anything out there and better than most (a UK bird 'tog who contributes to the review says she is "at last" able to keep up with Peregrine falcons at speed); metering and exposure are "pretty faultless"...

And it goes on in the same tack throughout the review.

It also makes a point of acknowledging that it's very clearly a camera best suited to sport, action and wildlife photography - it's not a "low cost alternative to the 5D Mk II, it's different kind of camera... better suited to subjects that need long telephoto lenses ".

In other words, while it will shoot landscapes, it is not a landscape camera.

This is utterly obvious, yet Wiggett et al complain when it doesn't work like a 5D Mk II!

rogernoel · November 14, 2009

As a prospective buyer of the 7D, I read the extremely, lengthy Wiggett review, plus all of the comments to date. I have also read several other reviews of this camera, most of them were very positive. It should be noted that almost at the very end, Wiggett responded to the firmware question, that he did not have the latest update when he tested the camera, but nothing further.

I would hope that this review will not be a reason for not purchasing the camera. If the camera were as defective as Mr. Wiggett's review indicates, I doubt that Canon would have marketed the camera. I shoot all my shots in RAW and JPEG and am totally addicted to post processing. I will upgrade from my 30D to the 7 D shortly.

angel_bocanegra · November 15, 2009

JULIAN ROBINSON wrote;