"Latitude for focus errors?" Tue 12-Jun-12 04:47 AM by Clint S
Chula Vista, US
How much can a photo be out of focus and still produce what looks like a decent image. This can be a confusing question because of the variables, Circle of Confusion and size of the final print are just two.
But here is an example, the first image is the full frame reduced to 1200 pixels longest edge. It could have been as long as 1800 pixels.
But here is the surprise.
This is a 1200 x 774 pixel 100% crop.
The images are in my gallery if you want to see the 1200 pixel versions.
Now does anyone have an excuse for out of focus images? LOL
#1. "RE: Latitude for focus errors?" In response to Reply # 0
This is an excellent question Clint, and your statement about variables is also significant. I can add: How large an image are you printing (are you printing at all), from how far away will it be viewed, and of course, while not so much associated with the D800 (because it has tons of it), how will the resolution play into the focus issue?
Focus is generally my main concern and I am frequently disappointed in the job that some camera-lens combinations perform (I won't go in to specifics here), But I am buying the D800 because what I have heard good things about its AF capabilities with different lenses in varying lighting conditions (even though I have now heard countering claims).
Another issue that I may find to be true (it seems obvious, but I haven't proven it yet) is that an increased pixel density will display evidence of a poorer focus better than a lower pixel density. To date, my D700 seems to focus better than my D7000 (which has a little less pixel density than the D800). I will certainly know in a couple of days if the D800 can compete (although I suspect there will be no question).
#2. "RE: Latitude for focus errors?" In response to Reply # 1 Mon 25-Jun-12 12:45 AM by Clint S
Chula Vista, US
>Another issue that I may find to be true (it seems obvious, >but I haven't proven it yet) is that an increased pixel >density will display evidence of a poorer focus better than a >lower pixel density. To date, my D700 seems to focus better >than my D7000 (which has a little less pixel density than the >D800). I will certainly know in a couple of days if the D800 >can compete (although I suspect there will be no question).
The answer to this is what I was trying to state my findings for as I had thought the same thing since getting the D7000, which was contrary to my medium format film shooting experinces. But I apparently stated in a poor fashion.
With my D7000 any focus error shows up in 8 x 10 or larger prints and even in large (1000 pixel) web images.
But with the D800 and focus error I can print the image, without cropping, at 8 x 12 and appears to be focused. Here is the full image at 1200 pixels which when viewed on 100% on a monitor looks similar to an 8 x 12 print (just bigger).
What is amazing is how out of focus the dandelion was, again needs to be viewed at 100% for the full effect.
That is very unlike the D7000 in my experiences.
With the D7000 I have to work at making sure I get a well focused image for prints of any size, but with the D800 I do not have to work for a well focused image, unless I am after images that will print at 16" longest dimension or larger, in which case I just need to make minor changes in my shooting style .
The D800 had brought back a pleasure to shooting photos! And I have now learned that there is such thing as "good enough" when shooting.
#3. "RE: Latitude for focus errors?" In response to Reply # 1
New Bedford, US
Hi Tom, I agree with you on the focus between the D700 and the D800. Today was the first time I was able to get out to shoot with the D800. After looking at the images it seemed, in my opinion, the D800 is less forgiving on focusing errors. A lot of the shots were taken at f/5.6, my usual aperture for shooting with the Nikon 24-70 on the D700. I think I will be using the tripod more as my handholding technique when using the D800 seems to need improvement. -------------------
#4. "RE: Latitude for focus errors?" In response to Reply # 0
New York, US
As you say, a lot of variables are involved. I never shot MF myself, but I used to shoot LF, and I expect the principle was the same: the larger the negative, the less magnification you needed to make your print. Magnification increases the size of the image, obviously, but it also increases the size of any errors / flaws in the image as well.
Now, let me state this upfront: I'm lousy at the mathematics of re-sampling images, so I may have botched this big-time. That said, my understanding is that when you are in the digital realm, the concept of magnification doesn't work the same way we older (not old) photographers who cut our teeth in the film era think about magnification. (For simplicity, let's keep image cropping out of this discussion.)
If you're printing an 8x12 image at 300 dpi, then that requires 8.6 million pixels, or ~1/4 of the resolution of the D800. For the web, which is typically something like 72 dpi, then (by my potentially flawed math) you’d need something like a 102x68-inch monitor to view a D800 image at 100%! For a 1200x800 pixel image displayed on the screen (72 dpi), you’re using less than 1.4% of the pixels a D800 captures. At that size, I think even a pinhole camera will give you a sharp image.
For me, the long and short of it is 100% pixel-peeping with the D800 is only necessary if I’m printing bigger than 16x24 (and even that may be questionable, depending on viewing distances, etc.). For my more typical print sizes, which are 8x12 or smaller, I spend more time looking at D800 images at 50% magnification than 100% magnification and judging sharpness from that.
All that said, there’s nothing wrong with striving to improve your photography technique--using a tripod, live-view focus, mirror lock-up / shutter release cable, etc. The mistake is in assuming it’s necessary all the time when using the D800.