fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1              3       end
  

Archive 2012 · 7d autofocus worked fine, now not!

  
 
Gochugogi
Offline
• • • • • •
Upload & Sell: Off
p.2 #1 · 7d autofocus worked fine, now not!


Perhaps there is a difference in IQ between individual 7D cameras as options about noise and useable high ISO are all over the place. In my case, ISO 640 is very good even pixel peeping although the blue channel is just starting to break up when viewed at 100%. I find ISO 1250 is my normal limit. I avoid ISO 250, 500, 1000 and 2000 as they are particularly noisy compared to ISO 160, 320, 640, 1250 and 2500. ISO 100, 200, 400, 800 and 1600 are so-so but better than the preceding set. Also, avoid auto lighting optimizer if you use RAW in DPP or shoot JPEGs as it adds considerable noise.

I know a lot of the hockey and football shooters say ISO 1600/3200 are good on the 7D but they tend to shoot in more even light (less low mids & shadows) so noise is less of a problem.



Aug 07, 2012 at 01:43 PM
UgashikBob
Offline
• • •
Upload & Sell: Off
p.2 #2 · 7d autofocus worked fine, now not!


I tend to agree with Mike and Jeff. All of my L glass has intermittent softness on the 7D. I have tried it with spot, single point with and without expansion, and zone. When it works its really good but when it doesn't its slightly soft. Same glass seems to be better on my old 1D Mkiii. Just not comfortable with it. Would really like to find out its technique as its a year old refurb.
Several threads on other sites with this same issue.



Aug 07, 2012 at 02:54 PM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #3 · 7d autofocus worked fine, now not!


Some of you have to be using a different 7D than the one I am using. For the last 5 months, I've been visiting an eagle nest almost daily and documenting the nesting process. While the single chick was in the nest and setting up in one spot was easy, I was using my excellent 1D Mark IV and excellent 500/4 from a tripod. Once the chick fledged from the nest a month ago, however, I have continued to follow its progress, but using a more mobile rig was desirable. So, I've been using my 7D and my 400 DO (another piece of equipment that those with "internet wisdom" love to bash) hand held to follow the fledgling through the marsh, over river banks, and through the timber.

First, the AF of the 7D has never let me down. A couple of days ago, I waited patiently to get a burst of shots of the fledgling taking off from its roost coming toward me. The action was fast and erratic, and the background was extremely busy. Yet, the 7D/400 DO combination nailed the focus (even at 100%) for 8 of the 9 shots in the burst, and I'm confident that the 1 out-of-focus shot resulted from my failure to keep the focus point on the bird. Further, when I compare the prints of pictures from earlier in the project that I took with the 1D Mark IV and 500/4 with those I've taken in the last month with the 7D and 400 DO, I can't see any differences in image quality, and, if I didn't know which ones were taken with which equipment, I wouldn't be able even to guess which came from the 1D Mark IV and which came from the 7D. While I'd guess that, if I were inclined to look at the pictures and compare them at 100% or 200% on the screen, I might find differences, in real world use of the pictures, any differences in sharpness, detail, noise, etc. are totally negligible.

Les



Aug 07, 2012 at 03:39 PM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #4 · 7d autofocus worked fine, now not!


Hello

It does seem that my thoughts about the IQ of the 7D is not quite so absurd, as I seem to have some qualified support.

I have two images, it has to be said that they are not quite equvalent, but it is the best that I can do. One is at 133% by Photoshop interpolation and the other is at 100%. The details were ISO400 F7.1 and 1/500 and the other ISO 640 F5.6 and 1/640. Both had default sharpening in DDP of 3 and were resized and saved in Photoshop.

Here are links to them. I am sure that there will be alll sorts of cavils and qualifications, but try to judge them on overall IQ alone.

http://www.dpreview.com/galleries/4402957911/photos/2141659/a1
http://www.dpreview.com/galleries/4402957911/photos/2141658/b1
One is a 5D2 and the other a 7D, so which is which?

Mike Engles


PS oops the captions give it away, oh well



Aug 07, 2012 at 07:33 PM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #5 · 7d autofocus worked fine, now not!


mikeengles wrote:
Hello

It does seem that my thoughts about the IQ of the 7D is not quite so absurd, as I seem to have some qualified support.

I have two images, it has to be said that they are not quite equvalent, but it is the best that I can do. One is at 133% by Photoshop interpolation and the other is at 100%. The details were ISO400 F7.1 and 1/500 and the other ISO 640 F5.6 and 1/640. Both had default sharpening in DDP of 3 and were resized and saved in Photoshop.

Here are links to them. I am sure that there
...Show more

An issue that I immediately see with the 7D photo is shutter speed. In order to produce the final image from the 7D with its 1.6 crop sensor, that image needs to enlarged by a factor of 1.6x greater than the file from a comparable full frame camera. That means that any imperpections, including those caused by camera shake and subject movement, are also going to be magnified by a factor of 1.6x. And, if you intend to crop or to print large (or look at the image at 100% on the screen), the image is going to be enlarged even further and imperfections are going to be magnified even further. And, with your sample 7D image, trying to get a sharply focused image is further complicated by the fact that you have minimal depth of field, and it would also appear that the subject critter was fairly close, which also magnifies the effects of even minor subject movement unless you are using even higher shutter speeds.

One of the first things I learned when I started to use my 7D (and, to some extent, my 1D Mark IV) was that, if I wanted to get the most out a high pixel density sensor, and especially if I was going to take advantage of the the opportunity to crop more and print larger, I needed to rethink my "rules of thumb" about shutter speed. For any subject with even minor movement, I want to keep my shutter speeds above 1/1000 and 1/1600 or 1/2000 is even better. And, if I'm shooting hand held or if I were shooting with a lens without IS, this would become even more important. If I intend to print at only small size or look only at small web-sized versions, the magnified imperfections of a high pixel density sensor are not important, but, if you do want to fully take advantage of the what a high pixel density sensor can do, it becomes critical to reduce the effects from both subject movement and camera shake by using higher shutter speeds.

Below, I'm pasting a few shots from the eagle fledgling burst that I referred to a few posts up. I realize that the size of the images is relatively small, but I'm just pulling them off my website server, and I don't have time to prepare and upload something larger, but I assure you that they are quite sharp. These were shot at 1/1600, and I know that, had I tried to shoot them at even 1/800, I would not have been happy with the results. Also note that, especially with the busy background, the 7D AF had to be up to the task, too.

1.
http://www.wildlifeimagesbyles.net/db_IMG_3539eagle-nest-8-3-121.jpg

2.
http://www.wildlifeimagesbyles.net/db_IMG_3540eagle-nest-8-3-121.jpg

3.
http://www.wildlifeimagesbyles.net/db_IMG_3543eagle-nest-8-3-121.jpg

4.
http://www.wildlifeimagesbyles.net/db_IMG_3544eagle-nest-8-3-121.jpg

5.
http://www.wildlifeimagesbyles.net/db_IMG_3545eagle-nest-8-3-121.jpg

6.
http://www.wildlifeimagesbyles.net/db_IMG_3546eagle-nest-8-3-121.jpg

And, even when the bird has been relatively static, I'm still looking for shutter speeds in the 1/1250-1/1600 range to minimize the effects of even minor subject movement and hand-held camera shake.

http://www.wildlifeimagesbyles.net/db_IMG_3447a-eagle-nest-8-1-121.jpg

And here is an image similar to your two examples. The mink was a quick grab shot when I caught it out of the corner of my eye while waiting for the eagle fledgling to do something. With the relatively close distance and fairly shallow depth of field plus any minor subject movement, I would likely not have been happy with the results had I been using a shutter speed any lower than something in the 1/1250-1/1600 range.

http://www.wildlifeimagesbyles.net/db_IMG_3335mink-7-30-121.jpg

Les



Aug 07, 2012 at 07:52 PM
Jeff Nolten
Offline
• • • •
Upload & Sell: Off
p.2 #6 · 7d autofocus worked fine, now not!


I agree UZ2, one has to be mindful of the shutter speed. My tendency has been to use the lowest shutter speed I thought I could get away with to keep the ISO/noise as low as possible. However even relatively stationary birds can move their heads quickly and abruptly. That was certainly true of the wary clapper rails I was trying to capture the other day. So I've been practicing with the following rules in mind:

1) Don't use spot focusing in AI mode even for small birds,
2) Use ISO 160, 320, 640 or 1250 (per Gochugogi, but I've been doing this since 40D days)
3) Keep the shutter speed above 1/1000 sec.

My results have been much more consistent these last two days. My 7D does not have a focus issue and noise is not really a problem using Lightroom. However I maintain that the 7D would have been better served with a 12 MP sensor. Here is a shot from today. First is the full image followed by a 100% crop of the full image followed by a 100% crop of the image reduced to 12 MP using a bicubic sharper reduction. The last image looks like what I get from my 5D2 or 40D at 100%. But hey that's 12 MP where the 40D is 10 and the 5D2 crops to 8.

















Aug 07, 2012 at 10:48 PM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #7 · 7d autofocus worked fine, now not!


Hello

The subject image with the 7D image was according to Bridge properties 8.36m. At that distance the DOF would be 4cm, 2cm front and 2cm behind. The shutter speed was 1.6 times the focal lenght of the lens 400mm=1/640. The 7D is a 100% crop.
The 5D2 was at 6.64m,but was enlaged to 133%. The overall result to my eyes gives it to the 5D2.

As I have said before, the in camera NR removes most all of the HF image detail, leaving Low and Mid frequency detail, which just looks coarse. This is even before it reaches the RAW file, so we are dealing with compromised data.

Noise and subsequent noise reduction is a killer of resolution, in any medium of data recording. The best way to get round it is to downsample an oversampled, but compromised RAW file. The noise is interpolated away and the result is a 1Mp image, ideally suited to the web. At higher ISO there is no fine detail left to sharpen as the image is band limited. It is like trying to eq up the cymbals in a piece of music that is band limited to 5Khz, it just brings up the noise.

The 7D does very well at ISO 200 and 320 and at a pinch at 400. Detail is still a bit coarse, but will respond to moderate sharpening, as the original posters images show.
Perhaps the latest set of images that he was so displeased with was done at high ISO.

PS please show the 100% crops , as well as the web image, otherwise, there is no way of judging.
In order to really judge we need to pixel peep, something it seems 7D owners are very loath to do.

Mike Engles



Aug 08, 2012 at 06:05 AM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #8 · 7d autofocus worked fine, now not!


mikeengles wrote:
Hello

The subject image with the 7D image was according to Bridge properties 8.36m. At that distance the DOF would be 4cm, 2cm front and 2cm behind. The shutter speed was 1.6 times the focal lenght of the lens 400mm=1/640. The 7D is a 100% crop.
The 5D2 was at 6.64m,but was enlaged to 133%. The overall result to my eyes gives it to the 5D2.

Mike Engles


But, what I and others have found, is that the old rule of thumb about 1/focal length or even 1/(focal length x 1.6 for an APS-C sensor) is not sufficient as pixel densities increase. I typically try to get at least 1/2 times focal length for even static subjects and 1/3 times focal length or more for subjects with more than minimal movement, and that is with lenses with IS. After having used my 7D for almost 3 years, I've found that, when I am not at those levels at a minimum, the results are, at best, inconsistent. When I exceed those minimums, the results become consistently what I want. Further, the more that I may need to crop increases the need for higher shutter speeds even more.

For the eagle fledgling take-off sequence above, I spent over 100 hours over several weeks waiting for the opportunity to get those shots and to get them with the light hitting the bird the way I wanted it to do so. And, while I was waiting, I was well aware of the fact that if/when I got the opportunity, it was likely going to be my only opportunity. I know that, if my shutter speeds were at, say 1/800, I might have gotten lucky, but knowing how much time I put into getting those shots, I was not about to leave getting them up to chance.

You can either accept the need for higher shutter speeds to get consistent results or you can reject it, but many 7D users have discovered through experience, as well as through logic, that high pixel density sensors require faster shutter speeds to achieve consistent results.

Les





Aug 08, 2012 at 06:24 AM
Andrew J
Offline
• • • • •
Upload & Sell: On
p.2 #9 · 7d autofocus worked fine, now not!


A lot of people criticizing the 50d and 7d are out 4 points or more in MA. 4 points out on the 40d didn't bother many people but those days are gone. The 7d is really a 44 MP camera and you need to be within a point or two with micro adjustment. Live view AF can prove this theory in almost every case.
In less than half an hour with the program FoCal you should see a big difference. If not a trip to Canon is in order.



Aug 08, 2012 at 07:22 AM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #10 · 7d autofocus worked fine, now not!


Hello

It is one thing calibrating a lens using a target for calibration and another getting a lens/camera combination to act consistently with all kinds of targets and lighting.
Most all these calibration aids, rely on a statistical method of calibration, so the error in any calibration can be + or- 4 units. You would do better to use 1 stop of aperture.

MikeEngles



Aug 08, 2012 at 11:18 AM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #11 · 7d autofocus worked fine, now not!


Andrew J wrote:
A lot of people criticizing the 50d and 7d are out 4 points or more in MA. 4 points out on the 40d didn't bother many people but those days are gone. The 7d is really a 44 MP camera and you need to be within a point or two with micro adjustment. Live view AF can prove this theory in almost every case.
In less than half an hour with the program FoCal you should see a big difference. If not a trip to Canon is in order.


While I agree that more accurate lens to camera calibration is required with higher pixel density bodies, I don't believe that is the issue with the 7D or, for that matter, 5D Mark II images that Mike linked to. With both of those images, I don't see an issue with either front or back focus. Instead, I see a general softness. And it would be my guess that the general softness is the result of either minor camera shake or subject movement or both.

Also, while micro adjustment can be a valuable tool, I'm also confident that there are many who make their cameras into ones that cannot focus by doing careless or haphazard micro adjustments.

Les



Aug 08, 2012 at 12:48 PM
dorian
Offline
• • • •
Upload & Sell: Off
p.2 #12 · 7d autofocus worked fine, now not!


hi folks

wow! this generated a lot of conversation! as i look back at some of my 7d files i do think that there is a general softness to them that is not observed with the 1d. that being said i was able to make many nice images when using the 7d previously so i am not sure what the issue is now. maybe i am just out-of-practice with the body as others have pointed out; this is entirely possible. i will have to get a day with nice light and take both bodies into the field and see what i can achieve with each of them. thanks for the help!

cheers

dorian



Aug 08, 2012 at 01:42 PM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #13 · 7d autofocus worked fine, now not!


Hello

The 5D2 image I used had been upsampled but no one has commented on the actual
IQ of the images, as to which might actually be better.

I cannot think of how either subject movement or shake contributed to the softness of either image. They both used the same lens 100-400 zoom, which has IS, so even if the IS only added 1 stop improvement, then actual shutter speeds would be the equivalent to 1/1000 for the 5D2 and 1/1250 for the 7D. These are well within all the margins suggested but other posters. In both instances apart from breathing there was no subject movement.

Dorian, I suggest that you do not wait for good light,as the 7D can do rather well in good light and low ISO. It is in not good light and having to use ISOs higher than 640, that all the softness becomes apparent.

Here is a good picture at ISO400 downsampled for the web 1/800 F8 with 100-400 zoom and IS

http://www.dpreview.com/galleries/4402957911/photos/2044635/hawk-moth_mg_0846

and here is a 100% crop from the original. NO NR or sharpening all other processing ACR6.7 default. It is noisy and soft, but does clean up well and sharpen well.

http://www.dpreview.com/galleries/4402957911/photos/2143275/hawk-moth100_mg_0846

Mike Engles



Aug 08, 2012 at 06:25 PM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #14 · 7d autofocus worked fine, now not!


mikeengles wrote:
Hello

The 5D2 image I used had been upsampled but no one has commented on the actual
IQ of the images, as to which might actually be better.

I cannot think of how either subject movement or shake contributed to the softness of either image. They both used the same lens 100-400 zoom, which has IS, so even if the IS only added 1 stop improvement, then actual shutter speeds would be the equivalent to 1/1000 for the 5D2 and 1/1250 for the 7D. These are well within all the margins suggested but other posters. In both instances apart from breathing there
...Show more

It wasn't my intent to get into an argument about shutter speeds, technique, etc. It was my intent only to pass along what I've learned from 3 years of experience shooting with a 7D.

Thus, since it would appear that you have your mind made up, I'm going to back out of this discussion with 2 final comments.

If you can get consistently good results with those kinds of shutter speeds at 400 mm with a high pixel density sensor, then, I give you credit for being a much better photographer than am I. I know that, when I try to use those kinds of shutter speeds at that focal length with the 7D (and even with an IS lens), I'll sometimes get lucky, but my results will be very inconsistent. If I shoot at shutter speeds that are considerably faster than what you are using, my results will be good and very consistent.

Finally, I would suggest that, if you are convinced that your shutter speeds aren't the problem and that it is the 7D that isn't up to the task, instead of continuing to frustrate yourself, you would probably be best served by selling your 7D and buying a 5D Mark III, 1D Mark IV, or some other camera in which you might have more confidence.

Les



Aug 08, 2012 at 06:39 PM
Jeff Nolten
Offline
• • • •
Upload & Sell: Off
p.2 #15 · 7d autofocus worked fine, now not!


mikeengles wrote:
...and here is a 100% crop from the original...


That image portion looks good to me for having no PP. The BG noise is reasonable and the proboscis is sharp. The body has some motion blur that might have been improved with Uz2's suggestion of a higher shutter speed.

I'm going to practice some more but I feel fairly confident that I can reliably capture the critter I'm after. I have a trip to Costa Rica this fall and I want to be ready. Lightroom even has a handy reduce to 12 MP on export that should suit my viewing needs. This has been a useful thread. Thanks.



Aug 08, 2012 at 07:09 PM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #16 · 7d autofocus worked fine, now not!


Hello

You are correct, I have made up my mind and have stopped using the 7D. It was a refurbed camera and I think that I will ask Canon for my money back.

My complaint is that the IQ of the 7D at ISOs 640 and above in essence makes the 18mp camera, into a 9mp camera. It is a form of misselling.

I have been doing photography for many years, printing my images in the film days. I thing that I know when image degradation is due to shutterspeed/movement and when it is caused by, in the film days excessive grain and in the digital days by excessive noise. I have no complaint with 7D at low ISOs. The images are soft compared to a 5D2, but do respond well to sharpening and NR. I am old fashioned and expect that when I print I have enough pixels at least for 150dpi, preferably 200, ideally 300, which is really not possible with the 7D at high ISO when images need cropping, as NR just reduces the overall resolution quite drastically.

As for buying another camera, well they do not give the pixel density to allow for cropping and that they at the moment are not better than the 5D2, which I have.
Canon need to come up with a 7D with IQ of a 5D2.

All the best

Mike Engles



Aug 09, 2012 at 03:48 AM
TeamSpeed
Offline
• • • • •
Upload & Sell: Off
p.2 #17 · 7d autofocus worked fine, now not!


I have shot with 7Ds for almost 3 years now, and have a different view, and align more with Les. If you have a perfectly working camera, and know how to set up the camera for success, and have a good set of post processing tools, the 7D actually does quite well, and is not quite as bad as you are making it sound.

My workflow is such:
0) Microadjust all lenses to any respective 7D I shoot (I have had 4 now)
1) Set up picture style and high ISO NR values so that they are ready as a starting point in DPP when I bring the raw in
2) I shoot to the right, ALL THE TIME, I never shoot to the left and bring up underexposed parts of images on purpose
3) I take the raw into DPP, make 3 or 4 little tweaks that can only be done in DPP and not in camera and export to JPG
4) I take the JPGs into CS3 and crop/compose accordingly
5) I run a batch action to clean up noise and sharpen and saturate on the JPGs

I have shot well over 10K images at ISO levels > 800 through these 4 years, shooting sporting events, band competition, drama, etc and have done very well with the results. I also make sure my shutter speeds are at least 1 / (1.6 x focal), but sometimes go for faster to guarantee no motion blur, because I do crop a bit, and I need less blur.

My results work well for my expectations and any clients (parents of kids, player families, etc). I understand others have higher expectations of their work, so what I consider acceptable may not match yours. However ISO 640 or less is all you get out of your 7D leads to either an issue with the camera, method of post processing, or your expectations are insanely strict.

ISO 1600

link

ISO 3200

link

ISO 6400

link

ISO 12800

link



Edited on Aug 10, 2012 at 05:29 AM · View previous versions



Aug 09, 2012 at 07:50 AM
mikeengles
Offline
• • •
Upload & Sell: Off
p.2 #18 · 7d autofocus worked fine, now not!


Hello

I gues that I do a completely different kind of photograpy from you. I only use my 7D with a 100-400mm for bird photography for the potential extra reach it is supposed to give. This means that the subject is usually a small part of the whole image, unless I can get really close. This also means that the end image needs considerable cropping and that is when noise becomes very noticable. That it why I try not to use ISOs greater than 640 and therefore have to limit the shutter speeds to generally below 1/1000 to allow for an aperture of 7.1 at most.

In the UK where weather and good light is unpredictable, I usually end up using 1/640 at F5.6 and ISO640. Higher ISOs allow for higher shutter speeds that make for shake free NOISY images, a photographic Catch 22.

I have a 5D2 for all my other photography.

PS I do think that showing web sized images that are supposed to show IQ is not very helpful. 100% crops are as far as I am concerned the only way to judge IQ. Perhaps I am just old fashioned in this

Mike Engles



Aug 09, 2012 at 03:49 PM
TeamSpeed
Offline
• • • • •
Upload & Sell: Off
p.2 #19 · 7d autofocus worked fine, now not!


mikeengles wrote:
Hello

I gues that I do a completely different kind of photograpy from you. I only use my 7D with a 100-400mm for bird photography for the potential extra reach it is supposed to give. This means that the subject is usually a small part of the whole image, unless I can get really close. This also means that the end image needs considerable cropping and that is when noise becomes very noticable. That it why I try not to use ISOs greater than 640 and therefore have to limit the shutter speeds to generally below 1/1000 to allow for an
...Show more

On all boards where I do IQ comparisons and ISO management training, I always supply both the web size and full-size images.

As to the lens you are using, the 100-400 is one of those lenses that have alot of variation from copy to copy. It may work well on other bodies, but may not be all that sharp or resolve well on the 7D. I have seen this with the 24-105L as well, both of which are very old design lenses, and that could be part of the issue. Out of 3 copies, only one was very sharp on the 7D, the other 2 were quite soft, regardless of MFA, etc. I kick myself quite often for having gotten rid of that lens, even to this day. One of my favorite images from that one sharp 100-400L on the 7D is included at the end of this post. Have you tried (ie. rented or borrowed) any other long telephotos, prime or otherwise?

Finally, to your comment about 100% views, I will have to respectfully disagree. One of the single large mistakes people do make is to take each and very image from different bodies and then zoom in to 100% view to compare. That is really quite inappropriate. The only way I do my check for IQ is to first process the image and then resize it to the print size I expect to make at the DPI I need, then I zoom to 100% on that image. If you just take a 7D image and view 100%, it will undoubtedly show imperfections that won't please, however they won't show up on prints, so it is a false positive on the IQ of the image.


link

I do some birding, but am not very good at it. Here is an ISO 6400 500mm duck shot, which seems to be the defacto test shot that seems to accompany discussions around camera bodies!


link

Here is a less aggressive shot at 1600. This is also at 500mm, using a Sigma 50-500 OS lens.

link



Aug 09, 2012 at 06:51 PM
uz2work
Offline
• • • • • •
Upload & Sell: Off
p.2 #20 · 7d autofocus worked fine, now not!


TeamSpeed wrote:
Finally, to your comment about 100% views, I will have to respectfully disagree. One of the single large mistakes people do make is to take each and very image from different bodies and then zoom in to 100% view to compare. That is really quite inappropriate. The only way I do my check for IQ is to first process the image and then resize it to the print size I expect to make at the DPI I need, then I zoom to 100% on that image. If you just take a 7D image and view 100%, it will undoubtedly show
...Show more

I couldn't agree more with the statement that I quoted above. With the 18 MP sensor of the 7D, when you are looking at a 100% crop, you are looking at a very, very tiny portion of the total image. The images that you posted in your earlier post are a great example. When you look at the ISO 6400 shot and look at the large version to which you linked, the noise appears to be horrid. However, when you look at the full image, it may not look as clean as one shot at ISO 200, but it looks very good, and I'm sure that it would print quite well at a reasonable size.

Any comparisons of images from two cameras with different pixel count or density comparing both at 100% is a completely invalid and meaningless comparison that will always make the camera with the lower pixel count or density look better than it should. If, for example, you compare a 7D image to, say, a 1D Mark II image both at 100%, the comparison might well make the 1D Mark II look better with regard to noise. But, because the 1D Mark II starts out with fewer than half of the pixels that the 7D has, each 1D Mark II pixel is going to have to be enlarged by twice as much to create a final image at any given print size, and, the more you have to enlarge each pixel, the more you are also going to be magnifying noise. If, on the other hand, you compared images from both cameras at comparable output sizes, I'm confident that the 7D image would be the clear winner. Even if you were shooting with a 1DX, which would seem to have a huge noise advantage over the 7D, if you have to crop a 1DX image down to, say, 4 or 5 MP to get the same field of view as that from a moderately cropped 7D image, when you make a print (or look at the image on the screen at fit-to-screen view), each of those 4 or 5 MP remaining 1DX pixels is going to have to be enlarged so much and so will noise be magnified to an extent that much of the noise advantage of the 1DX will be lost. I routinely shoot my 7D at ISO 800, and, if I have to, I'll go up to 1600 for wildlife shooting. I find, though, that, if I can't get the shutter speeds that I want at ISO 800, that generally means that the overall quality of the light isn't good enough to create pictures that I'm going to be happy with regardless of noise. And, when I look for noise at 100%, in those images, I'm probably going to be alarmed, but, yes, it does disappear in an actual print or even in screen viewing at fit-to-screen size.

Les



Aug 09, 2012 at 07:25 PM
1              3       end




FM Forums | Canon Forum | Join Upload & Sell

1              3       end
    
 

Welcome back
Log in to your account