Why the jaggies in this shot?

I'm also trying to not suck but having fun anyway!
 
I was using the Standard Picture Style with 3,1,0,0 parameters. When
you say oversharpened are you referring to my original shot or the
one that was modified by Darryl?
I'll play around with the sharpening (it was shot in RAW) and see if
that helps.
I was referring to your originals. There are sharpening halos around the contrasty edges.

Sharpening always enhances jaggies.

--
John

 
I played around with the image today but nothing would completely eliminate them, the best I could do was when EC was set to -2 but that only made them harder to see.

We have another storm moving through the Bay Area tomorrow and I'm hoping for exceptional atmospheric conditions this weekend. The rain and wind push everything out and night shots can be quite nice. I'll try different settings if the weather works out.

Just for grins, here's the shot with everything set to zero (sharpness, contrast, saturation, color tone) and a straight conversion to JPEG via DPP. The jaggies are muted quite a bit but still there.

The original image (large file warning!):

http://farm3.static.flickr.com/2393/2284393563_d0a4214710_o.jpg

Here's the cropped image, same settings as above:



Definitely an improvement over the first shot with sharpening set to 3, I think.

Is there a way to upload RAW images to websites? I always hate to post JPEG conversions but flickr won't except much else!
 
Do you have a different RAW converter to try like Adobe Camera Raw (in Photoshop and Lightroom)? In my experience, DPP is more prone to moire and jaggies than ACR.
--
Matthew
 
John's right, the image was showing a bit more noise than it should have. I used noiseware to remove the noise and then blurred it up a bit and boosted brightness & contrast and violia!
 
Just ZoomBrowser and DPP. I have Elements 6.0 but have to convert the image to TIFF or JPEG before it will open the file so I'm kind of stuck with the Canon software for the time being.
 
You are seeing artifacts of the bayer sensor and demosaicing algorithms.

Each row of a bayer sensor has either Red and Green pixels, but no Blue pixels or Blue and Green pixels, but no Red pixels. This leads to the 2 pixel jagginess.

The demosaicing algorithms then has to fill in the other pixels for a color channel. If you open the image in PS and look at the separate color channels, you can see each one looks different.

--
http://www.pbase.com/chibimike
 
Beautiful shot!
Did you take it from Treasure Island?

What tripod did you use?

Jeff
--
Capturing the beauty of God's creations in daily life!
 
A long series of storms had just moved through the Bay Area with high winds, then the temperatures dropped so the conditions were just about perfect. I was using the Slik Pro 700DX, a decent tripod but not without problems. I have to give most of the credit to the 50mm f1.4 lens used for the shot, it really brings it home.
 
I'm still a newbie at all this, what could I do differently when taking this shot? Or is this one of those things we deal with PP?

This series of shots is the only place I've seen this behavior so I'm not worried about the equipment. I figure it's one of those rare occasions where the image tweaks the sensor in ways it just can't handle.
 
Prevent.......doubtful, minimize the effect, yes

Try a different raw processor.........or don't sharpen the image in DPP and save the image out of DPP as 16 bit tif, and sharpen using USM in photoshop elements (or another program). I've found that, on occasion, sharpening in DPP produces strange results and USM in PSE does a much better job. Different sharpening algorithims perform their magic in different ways and one might be better than the other. One trick might be to mask the areas of concern, invert the mask, and sharpen everything but the bright lights.

If you could post a "very large image" image, without any sharpening in DPP, I'd be glad to try the USM in PSE to see if there are any differences.

Very nice photograph, it'll be worth the effort to resolve the issue.

Mike
 
You are seeing artifacts of the bayer sensor and demosaicing algorithms.
That might be part of it, but the same camera model shoots all kinds of strips of light without this happening, so there must be something special here. If I were the OP, I'd shoot these buildings again with high magnification to see if there is any pattern in the lights. I'd also shoot with an IR-pass filter if I had one, to see if the lights in those edges have more IR than the other lights in the image.

--
John

 
I mean on the one hand you have people complaining about the "softness" of Canon's AA filter, and on the other we see "jagginess" which is a real artifact from pixel based digital pictures. It's just unusual to see it in a Bayer sensor as it tends to be more soft by design. Blown out contrast aside, it's a real compliment to your camera/lens combination to render the sharpest image possible. I suspect that your image scales very well to larger print sizes.

--
I have plenty opinions ...
http://www.epinions.com/user-theuerkorn
 
There isn't really anything you could do to prevent this, just fix it in post.

You just happened to find an instance that shows bayer flaws. The bright lines are just a few pixels wide and at a slight angle which magnifies the visibility of the pixel or two of color slop that happens with a bayer sensor. The horizontal and vertical lines that are not at an angle are fine.

--
http://www.pbase.com/chibimike
 

Keyboard shortcuts

Back
Top