www.scantips.com

Diffraction Limited Images ? Really?


More images. Again, these are FX D800 36 megapixels, and DX D300 12 megapixels. All are ISO 400. In this case, the FX actually has the smaller pixels (205 pixels/mm 4.88 micron pitch D800, and 182 pixels/mm 5.5 micron pitch D300). Nevertheless, to me, the smaller FX pixels sure seem to win (because of the larger sensor needing less enlargement to view it). Smaller pixels provide higher resolution, and it's hard to beat higher resolution. F/5.6 and f/8 are very nice, but I can't find any startling problem starting around f/11. There is always more diffraction as we stop down, but sometimes the increased depth of field can be a much more important help. So the subject is about f/22 being beneficial to increase depth of field, and that we should use our tools when they help us.

Next are the two main scenes here. In order to see details bigger, both FX and DX are selecting the crop to be the same 16.3% of the frame size (D800 7360 wide becomes 1204, D300 4288 wide becomes 702). Then both size crops are resampled to 600 pixels width, which for FX is about 50% size (of actual pixels), and DX is about 85% size (DX enlargement is made greater, in order to appear same size as FX). Saying again, as in real life, DX is enlarged half again more than FX.

Speaking of DSLR class lenses, certainly f/5.6 or f/8 are the best general purpose cases, for routine best sharpness. So certainly I am Not saying that f/22 should be the routine norm for everything. But I sure don't notice much problem with using it when it helps needed depth of field.

For the short lenses, f/22 is closer to focal length /1 instead of /4, which violates the rule of thumb given on previous page. It shows, especially if image object size is smaller.
With 50 mm lens, maybe the f/22 difference at focal length / 2 is debatable, at least for DX. Again, these are very large enlargements.
With longer lenses, even in these very large enlargements, any difference in sharpness at f/22 is pretty difficult to see (other than depth of field is better).
You should try this. Otherwise, when depth of field is needed, you're missing out on a good thing.

Samples of two sensors using the same lens on same subject

(DX begins here)


FX Camera, 205 pixels/mm,
0.00488 mm pitch

f/5.6  
  f/22

I don't notice much loss of sharpness at f/22. The DOF on the ground is sharper.
And these small ones are 100% crop from its full picture background, maybe 200 feet distant, immediately left, between the two signs. It is same image, still focused on this near blue sign.

Clear win for f/22 (but which does need four stops more exposure).

At 24 mm, f/22 was not as sharp (at least as seen in this huge enlargement). And DOF is not much improved, even at 200 feet.

Above is standing much closer, right on the near pole, to increase 14 mm size. Bigger than it would have been helps. f/22 is as good, and maybe has a slight edge now.

At 24 mm, the f/5.6 image was sharper, but would not be noticeable in a regular smaller enlargement of entire frame.

f/22 comes through for us at 50 mm.

No question, f/22 obviously wins.


DX Camera, 182 pixels/mm,
0.0055 mm pitch

f/5.6  
f/22

DX f/22 was very slightly less sharp. Unexpected, I hope it was not my error.
But DX is also enlarged more than FX here. And it is shown very large here, so I'd say no real problem, if it helps depth of field dramatically.

And it does, at left is the background at about 200 feet.

14 mm, f/22 was less sharp.

24 mm, the edge goes to f/5.6.

The clouds varied the light slightly, but I'd give the credit to f/22.

No question, f/22 obviously wins.

There certainly are times stopping down can help longer lenses (depth of field). When it does, go for it. You're missing out if not considering it.

Even more images on next page.


Copyright © 2014-2024 by Wayne Fulton - All rights are reserved.

Previous Menu Next