Mallard in Flight

One of my first decent BIF shots, but the group of mallards were 120 yards away and going towards the sun which was low in the sky. As a result of the distance these are 100% crops which is generally not a good idea.

The lens used was my EF400mm 5.6L on my 7D. These are ISO 800 images. Other shooting information -- shutter speed was 1/1600 second and aperture was f/8. I used single point expansion for tracking and focus with the AF-ON button only. Since my holding is not very steady, I used a tripod with a Wimberly Sidekick gimbal mount. My good tripod is broken so I was using a cheap limber-legged tripod that shakes like a leaf in the breeze.

The first image has minimal processing -- I accidentally used the ACR defaults for tonal adjustments in the basic tab and used my nominal sharpening values (45, 0.7, 25, 25). The only noise reduction that I ever apply during RAW conversion in ACR is color noise reduction and very little in this case -- 28. I never use Luminance NR because it softens images. In Photoshop, I converted the profile from ProPhotoRGB to sRGB, cropped the size, and converted to 8-bits. I am really impressed with ISO 800 in good light.

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

Here are the lead ducks from the same image as the previous post. This one uses a very tiny bit of Neat Image NR and sharpening.

HOSTED PHOTO
please log in to view hosted photos in full size.

Nice shots Bill, I particularly enjoy the "double"... good shooting.

db

I have noticed the wrinkles in my shots too, when I do almost any cropping of the image. Not sure if thats from the reducing algorythm in CS3 or what, but the 7D images don't hold up as well as my MkIIN or 40D images did. Which makes no sense to me.

Nice in flight shots

Great captures, Bill.

Thanks for the comments db, Mitch, Ken, and Stan.

Mitch, can you clarify the "wrinkles" that you mentioned? I suspect that you are talking about the feather pattern on the duck's back, but that is just a wild guess.

I forgot to mention earlier that I was not quite happy with the beta profile for the 7D that is presently used in Adobe's DNG Converter as well as in ACR and Lightroom (they all use the same RAW conversion engine). Because of that, I created my own 7D profile (actually a group of profiles) and the images seen above were done using my custom 7D profile. The profile that I created tends to have less noise than the beta profile because, relatively speaking, it shifts tonal values downwards while the beta profile seems to shift darker tones upward.

One other thing that I have found different about 7D images is that color balance in the dark areas is slightly different than in the highlight areas, therefore, for the first time I am now using the Split Toning tab in ACR to tweak the color of the shadows. For my camera, I leave the highlights adjustments at 0, 0. I set the balance to +90 and for the shadows, I set hue to 50 and saturation to 45. Different lighting situations require some tweaking of these numbers.

Mitch, there was no rescaling downwards of the images. They are 100% crops. Here is what they looked like in the context of the whole image.

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

Bill Boehme wrote in post #9048785
Thanks for the comments db, Mitch, Ken, and Stan.

Mitch, can you clarify the "wrinkles" that you mentioned? I suspect that you are talking about the feather pattern on the duck's back, but that is just a wild guess.

Remember the Little Chickadee thread of mine from the other day? Look at those shots close. All of them have the wrinkles right next to the bird, like a little layer around them. Compression artifacts, I think from resizing after some small cropping. I also don't like the noise from even a small crop, then sized to 1000x667, so I ran the blur tool over the clear background parts. Not the bird or wood stuff.

canonloader wrote in post #9052035
Remember the Little Chickadee thread of mine from the other day? Look at those shots close. All of them have the wrinkles right next to the bird, like a little layer around them. Compression artifacts, I think from resizing after some small cropping. I also don't like the noise from even a small crop, then sized to 1000x667, so I ran the blur tool over the clear background parts. Not the bird or wood stuff.

Thanks, Mitch. I guess that I missed seeing your chickadee thread. Those are some really fine images. I have used the technique that you describe on a few occasions, but not as often as I should on bird images. For most other type of less critical images such as landscapes, it is not a significant issue. The image above were not resized and they looked the same before I cropped them. I have also seen the same thing in DPP images. It is the result of two things:

• The first thing that causes this is the RAW converter itself. The first part is the demosaicing -- as it reconstructs a color image by interpolating data between adjacent photosites, edges in the image becomes rather soft. The RAW converter uses edge detection algorithms to help reduce color smearing across edges that it is able to identify, but in so doing, the process introduces artifacts of their own which is most of what you see in these images.
  
• The other thing that will cause a problem is edge sharpening using the unsharp mask. Downsizing an image frequently make it a bit soft along the edges and typical practice is to add a bit of USM. However, for things like bird images with their fine feather details, USM can easily make things look much worse.
  
• One other thing. Sometimes people say that they did not sharpen an image. In actuality, there is not such thing as an image without some sharpening being applied automatically. And, we really would not want one with no type of sharpening done during RAW conversion.

What I am saying though, is I never got these compression artifacts form MkIIN or 40D RAWs. I use the same editing workflow, except for exporting a tiff from DPP. I'm going to use DNG on one and see what happens.

Nice shots, Bill, and they'd be better still with some noise reduction on the background only (layer it and protect the bird).

jgrussell wrote in post #9052420
Nice shots, Bill, and they'd be better still with some noise reduction on the background only (layer it and protect the bird).

Thanks, Judy. I may give it a try.

Bill, at 100% crops these shots are very passable, the slight noise in the background is non-existent on my computer screen.

canonloader wrote in post #9052409
What I am saying though, is I never got these compression artifacts form MkIIN or 40D RAWs. I use the same editing workflow, except for exporting a tiff from DPP. I'm going to use DNG on one and see what happens.

That was my initial impression when comparing the images to the XTi images before I shifted my paradigms to realize that I am seeing a magnified view compared to the other models, especially the 1D MkIIn. Using the same lens means that the 7D images will appear less sharp because of the magnification caused by greater pixel density and then you throw in a few edge artifacts on top of that.

These earlier models did not have full coverage sensors so they are, in essence, taking point samples and expanding what it sees out to the edge of the adjacent pixels. This is a form of aliasing that effectively hides or masks out a lot of detail and makes an image appear smoother. So, everything has its good and bad points.

As far as edge artifacts are concerned, they are a fact of life regardless of what RAW converter you use and the camera model makes no difference. Compression artifacts are not a result of downsizing, but happen when the image is converted to a JPG. However, all downsizing algorithms use algorithms to improve edges in order to avoid stair-stepping. The artifacts appear in all images -- I have seen them in images made with several different camera models because it is NOT the camera (unless you shoot in-camera JPG images), but the editing software that creates it. Here is a magnified example of edge artifacts from my XTi. The artifacts are still noticeable at normal viewing size. Something else -- the lower pixel density and its point sampling which aliases details causes a blown up view to look very crumbly despite having a very sharp lens. If you click on the image, you can se a normal sized image that will show how the edge artifacts look.

HOSTED PHOTO
please log in to view hosted photos in full size.