Review of 7D by Darwin Wiggett

Started Nov 11, 2009 | Discussions
highpriest Contributing Member • Posts: 989
Re: Dear alundeb......

David Hull wrote:

Other people figured this out, examples have been provided here and elsewhere. He could have figured it out as well had he had the time and done a bit of reasearch. When he tested the third unit, he should have gotten a clue but he missed it. It is really that simple.

So what did he miss? I'm not being awkward, just genuinely curious.

Cheers
HighPriest

 highpriest's gear list:highpriest's gear list
Canon EOS 5D Mark II Canon EF 85mm f/1.8 USM Canon EF 50mm f/1.4 USM Canon EF 70-200mm F4L USM Adobe Photoshop Lightroom 4 +1 more
highpriest Contributing Member • Posts: 989
Re: he should sharpen the images

tivoboy wrote:

I find the best place to see the OOF or lack of sharp out of camera focusing is in the hair. In the example above the mans mustache and hair is blurred, and no amount of sharpening is going to increase the sharpness if the individual hairs are blurred. It needs to start out better.

I agree. The first image isn't what I would call sharp - even at that size it looks a little mushy. With a high quality lens such as the 24-70, the camera should have done a much better job.

Cheers
HighPriest

 highpriest's gear list:highpriest's gear list
Canon EOS 5D Mark II Canon EF 85mm f/1.8 USM Canon EF 50mm f/1.4 USM Canon EF 70-200mm F4L USM Adobe Photoshop Lightroom 4 +1 more
rondhamalam
rondhamalam Senior Member • Posts: 2,582
Disappointing indeed

elfroggio wrote:

Wiggett compares the 7d to Nikon d300s and the 1sdMk3.

http://darwinwiggett.wordpress.com/2009/11/11/the-canon-7d/

left 7D
right Xsi

I resized the image to make 7D the same size as XSi, didn't help much

rondhamalam
rondhamalam Senior Member • Posts: 2,582
AA Filter is to blame

What else?
diffraction? no difference from 12 to 18MP, and 21MP was into play
camera shake? not on tripod
lens limit? not even on 21MP
pixel blooming? non sense

verdict: AA filter makes it soft

Tim O'Connor
Tim O'Connor Veteran Member • Posts: 5,424
Re: AA Filter is to blame

rondhamalam wrote:

What else?
diffraction? no difference from 12 to 18MP, and 21MP was into play
camera shake? not on tripod
lens limit? not even on 21MP
pixel blooming? non sense

verdict: AA filter makes it soft

Left one could have been out of focus ...:)

This just doesnt mesh with other well respected reviews - even the BJP one says:

"When the system was able to combine correct exposure with perfect focus the benefit of 17.9 million pixels was clear. Animal fur texture was bitingly sharp, and a comparison with 12 million pixel DSLR shots upscaled to 17 million pixel size always favoured the extra resolution."

-- hide signature --

Canon 40D
FujiFilm F20

http://www.pbase.com/timothyo

 Tim O'Connor's gear list:Tim O'Connor's gear list
Nikon D600 Nikon D500 Sony a7R III Nikon AF Nikkor 50mm f/1.8D Sigma 35mm F1.4 DG HSM Art +18 more
snapperZ
snapperZ Contributing Member • Posts: 869
Re: AA Filter is to blame

The tests look quite well conducted to me.Yes apertures are not optimal for a crop but that doesn't explain why the xsi gives better results. DPP and ACR seem to give a similar result too, he even tested a few different copies. The AA filter (or worse still loss of detail from NR in hardware) is the only answer I can come up with but no amount of capture sharpening is going to make these images look like a true 18MP image IMO. Interestingly this fits in completely with the recent 5DI/7D comparison in in these forums and RGs disappointing early pre-production samples. A bit of a shame as I've been considering a 7D for FL limited bird photography.

Having said all that I'd like to see some comparison pics form a 135L at f4-5.6 or similar before really making up my mind.

 snapperZ's gear list:snapperZ's gear list
Canon EOS-1D X Mark II Canon EOS 5D Mark IV Canon EF-S 15-85mm f/3.5-5.6 IS USM Canon EF 24-105mm f/4L IS USM Canon EF 70-300mm f/4-5.6L IS USM +29 more
MAC Forum Pro • Posts: 13,826
Re: Dear alundeb......

some funny points with some truth...

The reviewer is a landscape shooter and doesn't like dull for large print scape shots.

One can see in the following review that 5dii has better image quality for scapes

in the cameralabs review

if you shoot wildlife and some other stuff, then 7d probably suits good IQ. For the highest expectations in scapes and portraits, the 5dii wins

GeorgeML wrote:

Mrs Reality wrote:

And that technique would be?

1. Don't stop down a 7D beyond about 5.6 or maybe 6.3 at the most?

2. Spend a fortune on fast lenses and hope that you get a teensy bit more resolution?

3. Expect to spend a lot more time and effort post processing and don't forget all that sharpening?

4. Always use a sturdy tripod, MLU, and a remote or the timer?

5. Tweak the 7D constantly to get the best out of it for each shot and situation?

6. Micro adjust all your lenses to make up for the manufacturing slop in the camera or the lenses?

7. Try every Raw converter on Earth to find the best one?

8. Exchange or have your 7D and lenses serviced as many times as are necessary to make sure you have a good copy?

9. Don't forget that extra sharpening and more complicated post processing. Extra sharpening fixes everything.

10. Downsample all images to make them look as good as the images from a less pixel packed camera?

11. Test, test, and retest, and then test some more?

12. ALWAYS blame the user, not the camera?

Or just convince yourself that the 7D is the best of the best and only the best, no matter what.

You are joking but these are seriously good tips for convincing oneself that the 7D is indeed an improvement over older sensors. LOL

 MAC's gear list:MAC's gear list
Canon EOS Rebel SL1 Canon EOS 7D Mark II Canon EOS Rebel T7i Canon EOS 6D Canon EF 35mm F2 IS USM +8 more
Eduardo del Solar
Eduardo del Solar Senior Member • Posts: 1,941
Darwin's take on DPreview of the 7D

Sorry Guys, I now see that the dprevie comparison is in RAW (I just saw the JPEG label on the photos)........... But like I said in my review, the 7d does good with close subjects like studio scenes (the way dpreview tests most cameras), the problem is is with more distant scenes and detailed mid distant scnes like landscapes. I like to test the kinds of subjects I will ACTUALLY photograph, not test charts and flower vases.

If anyone is getting results they like from the 7D, then that is all that matters. Mileage may vary. For the way we shoot (detailed and often distant landscape in RAW and processed in DPP), the 7d does not cut it dor us. But don’t believe us or ANY review site, test for yourself with your lenses, apertures, RAW convertors and subjects.

Makes sense to me.

Eduardo

 Eduardo del Solar's gear list:Eduardo del Solar's gear list
Canon EOS 5D Mark III Sony a6000 Canon EOS 7D Mark II Canon EF 50mm f/1.4 USM Sigma 35mm F1.4 DG HSM Art +8 more
xxD Contributing Member • Posts: 726
Even I can do better

There's a problem with that review. Must be the method used or something, because even I and my 7D can do better than those samples.

loafer Senior Member • Posts: 1,561
lack of consensus (consistency)

For both the 50D and 7D the lack of consistency among the reviews is what strikes me. All the reviews have their own fairly well established "work flows" but come up with quite different results.

The only real wildcard that I can detect is the RAW converter issue (making the generous assumption that there is no bias).

On the positive side, I suspect that most owners of the camera will eventually be able to get the same results as the reviewer who comes up with the best results (unless that reviewer has a biased review.)

nothsafoto Regular Member • Posts: 419
Re: lack of consensus (consistency)

And there is the rub. Established workflows. the system has a whole new AF and metering, so the old workflow goes out the window.

These results do not tie in with others who have started from scratch and ignored previous methods or setups.

Robbie
--
Canon Can...Can you??

riknash Veteran Member • Posts: 6,869
Excellent Suggestions

Mrs Reality wrote:

1. Don't stop down a 7D beyond about 5.6 or maybe 6.3 at the most?

2. Spend a fortune on fast lenses and hope that you get a teensy bit more resolution?

3. Expect to spend a lot more time and effort post processing and don't forget all that sharpening?

4. Always use a sturdy tripod, MLU, and a remote or the timer?

5. Tweak the 7D constantly to get the best out of it for each shot and situation?

6. Micro adjust all your lenses to make up for the manufacturing slop in the camera or the lenses?

7. Try every Raw converter on Earth to find the best one?

8. Exchange or have your 7D and lenses serviced as many times as are necessary to make sure you have a good copy?

9. Don't forget that extra sharpening and more complicated post processing. Extra sharpening fixes everything.

10. Downsample all images to make them look as good as the images from a less pixel packed camera?

11. Test, test, and retest, and then test some more?

12. ALWAYS blame the user, not the camera?

Or just convince yourself that the 7D is the best of the best and only the best, no matter what. ;

Excellent Suggestions!

Most of us unwashed common folk are not qualified enough to properly use our cameras and the post processing software and computer hardware needed to bring out the best possible images. We should simply turn the dial to Green Box Auto mode and happily click away creating Jpegs which we then have Wally World print as 5 x 7's. ..

Seriously, in an effort to improve my results, I followed items 2, 3, 4 (sometimes), 5, 6, 9 and most certainly 12. My results are improving... but at great expense.

My previous camera was a Rebel XT which in my opinion takes much better landscape photos than my present 50D. I hated the XT for everything else. Way too slow for BIF which I had discovered I really enjoy doing. I was/am really disappointed in how fuzzy the 50D produces images. The expensive lenses did help along with lots of extra sharpening and time intensive handy-work in PS. Apparently the 7D creates even more exceptionally fuzzy photos unless you understand quantum mechanics...

From everything I have read here and at FM regarding the Darwin Wigget responses, it is clear to me that the 7D has created quite a stir. Being the highest pixel density per sq mm sensor, it has created many challenges that were simply easier to ignore in earlier less pixel dense cropper cameras. I doubt even Canon had anticipated how much this would affect the users. Suddenly most every piece of our post processing software is deficient, our techniques need to be relearned, and even our picture taking methods need to be re-evaluated. Like jumping into a Ferrari from a Honda Civic. They both have wheels and an engine but driving either one properly requires different skill levels.

I still want a 7D but will wait for the QC from Canon to improve along with the photo software to improve enough to allow even my unwashed hands the ability to produce excellent images without 3 hours of post fiddle work per image.

In the meantime, I'm off to get more fuzzy images of the Snowy Owls with my 50D...

http://rixbox1.telusplanet.net/rick/2009_11_11_Two_Snowy_Owls/album/

 riknash's gear list:riknash's gear list
Canon EOS 7D Mark II Canon EOS 5D Mark IV Canon EF 200-400mm f/4L IS USM Extender 1.4x Canon EF 24-70mm F4L IS USM HTC One
David Hull
David Hull Veteran Member • Posts: 6,341
Re: Dear alundeb......

highpriest wrote:

David Hull wrote:

Other people figured this out, examples have been provided here and elsewhere. He could have figured it out as well had he had the time and done a bit of reasearch. When he tested the third unit, he should have gotten a clue but he missed it. It is really that simple.

So what did he miss? I'm not being awkward, just genuinely curious.

Well, the one thing that myself and some of the others are harping about is that he set his aperture too tight and diffraction issues were becoming a significant factor in his zoomed images. This is what it looks like to myself and several others.

Several people have set up similar tests and gotten a much better result. To me the expected result is that the Drebel and 7D images would have been much closer. The best one is the Fred Miranda example where the 7D produced an image almost identical to the 5DII and this is more what I would expect.

However, the real point of my statement was that the guy tried three times, each time exchanging his camera for a new one. After finding the same issue with three brand new cameras , he never looked deeper to find the real root cause of the issue -- he just blamed the equipment. Several other people, upon reviewing his review were able to see the mistake instantly... and not only that, were able to replicate his test correctly and get a more reasonable result.

I am looking mostly at the pictures of the "thanksgiving basket" early in the review for all of these discussing which show the diffraction thing pretty well. I don't have a comment on the haystack photo. That thing just looks strange like it was processed through a Lucius Art tone mapper or something -- no clue what is going on there.

I am not trying to be awaward either (and I don't think you were implying that I was), I am really interested in understanding what is going on with some of this 7D AF stuff and figuring out for myself what the root cause is. I don't think the camera is fundementaly flawed, I am starting to think that there were possibly some misadjusted ones early on and a whole lot of pilot error on the part os some users.

I am not trying to be awkward either (and I don't think you were implying that I was), I am really interested in understanding what is going on with some of this 7D AF stuff and figuring out for myself what the root cause is. I don't think the camera is fundamentally flawed; I am starting to think that there were possibly some misadjusted ones early on and a whole lot of pilot error on the part of some users. This will get me flamed by the religious zealots (like Mrs. Reality), but I really don’t care.

Cheers
HighPriest

 David Hull's gear list:David Hull's gear list
Canon EOS M5 Canon EOS R
David Hull
David Hull Veteran Member • Posts: 6,341
Re: Excellent Suggestions

Nice one, good to see someone put this all into perspective with a sense of humor. Keep in mind that thre ar a whole boat load of people who are getting good photos with both the 50D and the 7D. I own a 50D and I don't get fuzzy pictures (except ocasionaly when my wife uses it -- good thing she doesn't read here ), I don't see why you should either.

 David Hull's gear list:David Hull's gear list
Canon EOS M5 Canon EOS R
MAC Forum Pro • Posts: 13,826
Re: Excellent Suggestions

agree, different cameras for different needs. you have it right -- it is the time consuming hoops that need to be jumped with the processing software that is the disappointment for landscapes -- stopped down at least.. I also think there is some lens and F stop comparison impact in all this. F11 on a crop is about as far as I go, unless macros

riknash wrote:

Excellent Suggestions!

Most of us unwashed common folk are not qualified enough to properly use our cameras and the post processing software and computer hardware needed to bring out the best possible images. We should simply turn the dial to Green Box Auto mode and happily click away creating Jpegs which we then have Wally World print as 5 x 7's. ..

Seriously, in an effort to improve my results, I followed items 2, 3, 4 (sometimes), 5, 6, 9 and most certainly 12. My results are improving... but at great expense.

My previous camera was a Rebel XT which in my opinion takes much better landscape photos than my present 50D. I hated the XT for everything else. Way too slow for BIF which I had discovered I really enjoy doing. I was/am really disappointed in how fuzzy the 50D produces images. The expensive lenses did help along with lots of extra sharpening and time intensive handy-work in PS. Apparently the 7D creates even more exceptionally fuzzy photos unless you understand quantum mechanics...

From everything I have read here and at FM regarding the Darwin Wigget responses, it is clear to me that the 7D has created quite a stir. Being the highest pixel density per sq mm sensor, it has created many challenges that were simply easier to ignore in earlier less pixel dense cropper cameras. I doubt even Canon had anticipated how much this would affect the users. Suddenly most every piece of our post processing software is deficient, our techniques need to be relearned, and even our picture taking methods need to be re-evaluated. Like jumping into a Ferrari from a Honda Civic. They both have wheels and an engine but driving either one properly requires different skill levels.

I still want a 7D but will wait for the QC from Canon to improve along with the photo software to improve enough to allow even my unwashed hands the ability to produce excellent images without 3 hours of post fiddle work per image.

In the meantime, I'm off to get more fuzzy images of the Snowy Owls with my 50D...

http://rixbox1.telusplanet.net/rick/2009_11_11_Two_Snowy_Owls/album/

 MAC's gear list:MAC's gear list
Canon EOS Rebel SL1 Canon EOS 7D Mark II Canon EOS Rebel T7i Canon EOS 6D Canon EF 35mm F2 IS USM +8 more
ojos New Member • Posts: 8
no effort at all

I have the 7D and the 500D. The effort I need is setting sharpness=3 (the default) in the camera and high-iso noise reduction to standard (also the default). Then I use DPP with the raw file and everything is set...

By the way, comparing the processed images reveals a slight sharpness advantage for the 7d and about one stop less noise. So at least my copy is better than the 500D.

Regards, ojo

rjjr Forum Pro • Posts: 14,659
Huh?

Clear7 wrote:

Are you using your hand again? Come on.. this is a family place.

Could you please explain that remark and how it's relevant?

ojos New Member • Posts: 8
Re: Review of 7D by Darwin Wiggett

Yes. My observations with 450D, 500D and 7D indicate this.

Tiny Tim Regular Member • Posts: 229
DPP sharpening = 0 does reflect underlying file sharpness

or so it would appear from my Noddy comparison.

Here is a scene shot with my 7D and 100-400 at 1/2500, f/5.6, 400 ISO. As far as I can judge the camera focused where I expected it to and neither front nor back focus seems to be a problem. At 1/2500 I do not expect camera shake to be a problem. Although hand held, I did take great care to maintain a stable system and I did use IS. This is the whole scene processed in DPP with Sharpening - 0. There have been no adjustments to exposure/levels/contrast. NR is at DPP defaults of 1,3 for 400 ISO.

Now I know the 100-400 is not the sharpest lens in Canon's arsenal, but it is a representative lens for many people who shoot wildlife and have a 7D for exra reach/detail/cropping power.

Here is a screen print from the focused subject viewed in Rawnalyze. I understand this to be an unprocessed file with no sharpening of any kind.....

I have to say that as far as sharpness is concerned it looks pretty terrible. Whether than is a fault of the lens or the camera I do not know, but there is no doubt it is soft. It looks similar in quality to the atrocious results obtained by Darwin.

Here is the same shot as a 100% crop after processing in DPP with Sharpening = 0....

Apart from the colour correction we seem to have pretty much equivalent sharpness. In other words DPP is genuinely not sharpening the file at all with Sharpening = 0. As we all know, a raw file does require sharpening in order to be fit for human consumption. The fine pixel pitch of the 7D is going to result in lower contrast at edges because it will more faithfully track softness from the lens or anything that does not fall precisely upon the plane of focus. By comparison, big lumpy pixels like those of a 5D2 will have much stronger contrast differentials at edge boundaries. It is my guess that the 7D files will require more sharpening in order to achieve equivalent apparent sharpness/microcontrast, compared to cameras with lower pixel densities, especially when viewed at 100% and thus "stretching" edge boundaries over a larger physical dustance.

So, how much should we sharpen? Prior to getting the 7D, Sharpening = 3 was about right for most of my cameras if I had a sharp shot. Here is how the 7D file looks with Sharpening = 3....

Things are certainly heading in the right direction, but perhaps we are still not quite there yet. Let's try Sharpening = 5....

Getting better, but let's see what Sharpening = 10 can produce....

Well, perhaps that's a bit overcooked, but there are no really obscene halos, and on my 132ppi monitor we should not forget that a 7D file viewed at 100% is equivalent to an image 40" wide. On a larger monitor with a lower ppi the image could scale up to 52" across, give or take. In print, at anything between 200-300 ppi (26" to 17"), would there be a problem with this image, even when viewed close up? I think not.

Now where this leaves us with respect to reviews I do not know, but I thought it was worth at least showing what differing degrees of sharpening can accomplish with a 7D file. I don't have an XTi but I do wonder whether DPP also leaves XTi raw files entirely unsharpened when Sharpening = 0. Perhaps it gives them a head start. Just a thought.

Anyway, the full range of DPP Sharpening values for this picture can be seen in an online album here - http://picasaweb.google.com/EezyTiger/7DDPPSharpening# .

 Tiny Tim's gear list:Tiny Tim's gear list
Canon PowerShot G7 X
tko Forum Pro • Posts: 12,787
not possible I think, it's the tester

I don't know why people are ignoring the evidence right in front of our faces - the DPR test results. 7d (left) versus d300s (right), right from our own beloved web site. Basically, on some photos you can see detail you can only dream about on the d300s. If it was the AA filter, ALL 7d photos would be blurred. Instead, it only seems to be the photos from this guy. Other sites support the conclusion that the 7d is the sharpest APC camera out there.

Keyboard shortcuts:
FForum MMy threads