Why picture quality decreasing with very big apperture? [closed]

Question

Possible Duplicates:
What is a “diffraction limit”?
Why are my photos taken at f/11 less sharp than those taken at a wider aperture?

I have made 3 landscape pictures with f/12, f/22 and f/32. Shutter speed and ISO are almost same, but last one has a lot of blur and is very pale. Why this happen? My purpose was to get maximal DOF.

Details: Nikon d5100, 18-55 kit

Asked by yura

Answer

Without wishing to get into detail - the short answer to your question is diffraction. This happens at smaller apertures, usually when you get above f/16 to f/22 etc, and the extent to which it occurs depends on the optical quality of the lens used. In your case, the 18-55mm kit lens which is a mass-produced cheap kit lens will show this problem moreso than say a fixed focal length prime lens.

You can achieve good depth of field with a more modest aperture such as f/14. There is rarely if ever good cause to go all the way up to f/22.

Answered by Mike

Check more discussion of this question.

How much different are the Tamron 17-50mm f/2.8 stabilized (VC) and non-stabilized versions?

Question

I have seen these 2 lenses being suggested by many people across the site as general purpose zooms and an upgrade to the typical 18-55mm kit lenses. The price difference seems to be around $200 (~INR 10,000) between the two. The lenses are also considerably cheaper than their Canon\Nikon counterparts.

Having using the stabilized kit lens on Canon, I appreciate the importance of IS\VC for slow shutter speeds (not just low light, but also for effects). However, quite a few reviews seem to suggest that the non-VC version seems to be better optically overall. There seem to be more quality control issues with the VC version as well.

My question is two fold:

1. How do the 2 lenses compare against each other - is the premium for the VC version worthwhile for the quality trade-off?
2. How do the 2 compare to the Canon stabilized kit lens (EF-S 18-55mm IS) in terms of optical quality - are they worthwhile upgrades? (there is a similar question for Nikon)

Answer

I have the VC version (Nikon mount). I have tested it alongside my 50mm prime and found it to be just as sharp at f/3.2. It is very slightly softer at f/2.8, but only when looking at an artificial test pattern. That's one person's opinion based on a sample of one.

I have tested other non-professional zoom lenses (Nikon and Sigma) and the Tamron is far, far sharper than any of them. Unbelievably so.

A camera shop owner I know recommended the non-VC version as being very sharp and reliable (and inexpensive). He didn't say it was worth buying the non-VC version over the VC version. He just recommended it highly as a fantastic value. He did say he'd also sold a lot of the VC version and hadn't noticed any quality control issues.

Because the lens is sharp wide open, VC isn't really vital except in low light. I would say the VC is probably not worth the premium unless you intend to shoot a lot of low-light or indoors work without flash.

Can't help with the comparison to the Canon 18-55mm. I would imagine from what I've found compared to other Nikon equivalents that the Tamron will be noticeably sharper, especially wide open. You can use it at 2.8 or 3.2, whereas the Canon probably needs to be stopped down to f/5.6 or f/8 to be at its sharpest.

Check more discussion of this question.

What's the base f-stop when counting an ideal f-stop for lens?

Question

I've read a lot around this site that an ideal quality of the lens is achieved when using f-stop roughly 2 stops slower than the fastest value of the lens.

Firstly, I'm not exactly sure what "2 stops" mean when it comes to f-stop, but this isn't really important in what I'm about to ask. What I'd like to know is what is the base f-stop when counting this value.

Let's look at it this way: If I have a prime lens with f/1.8, then obviously I'll be getting the ideal value from f/1.8, let's say after 2 stops I get f/2.8?

But! What in case of zoom lenses, for example 18-55, which have f-stop f/3.5 - f/5.6. Should I take f/5.6 as my base value, or does the base value depends on the real fastest value at the exact focal length?

Therefore, if I set the lens to 18 mm, their fastest aperture is f/3.5 and thus the ideal quality is at f/5.6 and when I set it to 55 mm, their fastest aperture is f/5.6 and thus the ideal quality is at f/8 ? Or is it f/8 for the whole focal range ?

Answer

Real fastest value at the exact focal length.

But, this is just a rule of thumb — it's not necessarily exactly two stops in every case.

For the question of what two stops mean, see What does f-stop mean?. In short, each stop is approximately the-square-root-of-two times the previous one. That means half the light is allowed in (which is why the seemingly-weird series of numbers was chosen). Two stops is doing that twice, which conveniently works out to simply being doubling. Two stops from f/2 is f/4, and and two stops from that is f/8. Or, starting at f/1.8, two stops is f/3.6.

The test results comparison tool used by Digital Photography Review is interesting for looking at this. Try this test of Pentax's 18-55mm zoom, which like most kit zooms goes from f/3.5 at the short end to f/5.6 zoomed out.

You can see that at 18mm, the sharpness improves as you go from f/3.5 to f/4 to f/5.6. At f/8, it's more even from center to corners, but doesn't get any sharper. And beyond that, it drops off.

Then, change the focal length to 55mm. You'll notice that with this lens, overall sharpness goes down from 18mm at f/5.6. At f/8, it picks up a little bit, and a little more at f/11, and then back down again at f/16.

So, that fits with the two-stops guideline pretty well. Same for the Canon kit lens, and Nikon too.

With the Pentax DA 15mm f/4 Limited test, they've got test results in third-stops, and there you can see that f/7.1 is no better than f/6.3, and f/8 is worse — so that's only 1⅓ stops. On the other hand the DA★ 55mm f/1.4 shows peak sharpness at f/4.5 — closed down by 3⅓ stops. This doesn't reflect a problem with the lens, just different priorities in design. The Nikon 50mm f/1.4G is similar, with peak sharpness at 3⅔ or 4 stops down. However, the entry level Nikkor f/1.8G fits the "two stops" guide quite well, peaking at around f/3.5.

It's also worth noting that this test is primarily concerned with sharpness, because that's easy to measure. Other image quality and rendering characteristics are also affected. Vignetting (light falloff in the corners) gets better the more you stop down, and it's usually gone after two stops. And bokeh quality is usually improved by stopping down, too — generally, lens bokeh is nicer/smoother stopped down a bit, but of course it's less visible since you have greater depth of field. Plus, shape of the aperture blades will become visible in specular highlights, which is a side-effect unrelated to stopping down per se — some modern lenses have rounded aperture blades to make that not be a concern.

Check more discussion of this question.