7
\$\begingroup\$

I am sick and tired of getting photos that turn out just slightly blurry. I can't seem to figure out why this is or how to fix it. I like to shoot at events, indoors and outdoors, action shots and candid portraits. Mostly "in the moment" stuff.

My biggest guess at the problem would be because I use the 50mm f/1.8 STM. I love the low aperture bokeh, but I'm afraid that my photos might just be a touch out of focus. With such a shallow dof, it gets blurry with even a small focus error. When shooting in the moment, it's can be a hit or miss.

Another guess might be the glass. The lens is nothing special so I wouldn't expect an incredible image, but a sharp image should be possible with most lenses.

I don't think it is motion blur cause I shoot about 1/250 or faster. Back to my hunch about the AF, I sometimes shoot with auto focus point, instead of center focus point. I just talked to another photographer who told me that's a sin. I can switch to center focus point, but that would require all my photos have the subject in the center of the viewfinder. I'd bet there's a way around this but I am not currently aware of it. And would that even fix my problem?

I'd attach two photos as an example, but I think SE compresses the photos so it defeats the purpose, so I'll add link to the original. At a glance, they look fine, but a closer look shows that the subject is not as sharp as possible. I'd love to hear any advice you have.

https://flic.kr/p/NsLQ5k

https://flic.kr/p/MjRyMc

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The second photo is using a different lens. Does that not rule out your hunch about the 50mm? \$\endgroup\$ May 21, 2017 at 21:18

4 Answers 4

16
\$\begingroup\$

The first image was shot wide open with what is essentially the lowest cost Canon lens (50mm f/1.8), in demanding low light conditions, and at ISO 1600. I'd say your results are exactly what I would expect.

The second image, I don't see anything wrong with.

My best guess is that you are overly concerned with the quality of each pixel viewed at 100% to a degree that no camera/lens combo will likely please you. If you want to test this theory out, I'd recommend renting a much higher quality lens for a weekend. Its not very expensive to do, and will take your current entry level lenses out of the equation. Something like a Canon 35mm or 135mm would be what I would recommend. I'm going to warn you though, you might turn this relatively inexpensive hobby into something much more expensive :)

This site already has a great wealth of information on this topic I'd explore:

\$\endgroup\$
7
\$\begingroup\$

The first example is just a soft lens. You need to stop down to get sharp images with a cheap lens. A better lens you shoot wide-open and get sharper results. Even so, wide open rarely gets the maximum sharpness but it can be sufficiently sharp that it does not matter.

The second example is motion blur. At F/7.1 you have plenty of depth-of-field but if you look at the lady with the bow, her arm is sharp but her head is soft. Both should be in focus since they are about the same distance from the camera. Also the feathers on the other lade are sharp too. The arm that is in front is blurry though, so that may be just moving or out of the depth-of-field which would be surprising for such a small aperture.

\$\endgroup\$
4
\$\begingroup\$

AF systems are not perfect, with such a shallow DOF it's not uncommon to see slight misfocus.

Even if the AF system is perfect, humans aren't, both you and the subject may move slightly while taking the shot.

For 'spur of the moment' shots, you probably want to use a smaller aperture, you can still good Bokeh, but it will give you a little more error room with the DOF. Especially as most lenses are not their sharpest at their maximum aperture.

It's also possible that your AF is missing. Get a focus check card and try it out. Your camera may have an adjustment hidden in the menus to correct any slight errors.

\$\endgroup\$
2
  • \$\begingroup\$ The second photo is at f/7.1. Would you suggest using a narrower aperture than that, and if so, why? \$\endgroup\$
    – Philip Kendall
    May 21, 2017 at 21:12
  • \$\begingroup\$ I only looked at the first photo and the question text. If the issue presents at f7.1 as well, it's more likely to be AF error (although if shooting at f1.8 is normal then I still wouldn't rule out the narrow DOF being the main problem) \$\endgroup\$ May 21, 2017 at 21:15
4
\$\begingroup\$

Practical tests can be carried out to determine how good a result is achievable and what factors may be reducing the actual sharpnesses obtained.

Find a surface with a fine pattern, or horizontal lines or similar where the focus point can be easily discerned afterwards.
Angle the camera so you get a close to distant image across the surface so you can see how sharp the image is across a wide range of distances either side of the focus point - see example image below.
Mark the focus point (mentally or physically).
Take photos at various apertures with shutter speed usefully higher than 1/focal_length_in_mm. eg say 1/250 ms at 50 mm.
Note where the actual focus point falls, how sharp it is and the depth of focus.

Try varying aperture, shutter speed, user steadiness, ... .
Try the same settings on two different lenses and compare results. e.g f/5.6, 50mm, 1/250s on the 50mm prime and the zoom. Report.

I'd expect the 50mm, f/1.8 results to be reasonable despite the low cost. Such lenses are USUALLY low cost because they are simple, have been made in the millions for decades and are a means of seducing newer users to the high-$ Primes dark side. (Or were until some modern reasonably priced zooms got almost as good as cheap primes across some areas of their ranges).

The image below was taken at f/8, 500mm (fixed aperture and focal length mirror lens), 1/320s. Click the image for the full size version (3000 x 2000 - 'ye olde' 6 Mp 5D Minolta image) and note the relative sharpness of the seeds and the concrete texture around the central sparrow and closer and further behind it. From this you can see focus point, max sharpness and fall off of sharpness with distance. That gives you a good starting point for determining what may (or may not) be able to done to improve your results.

sample wide-field subject

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.