4
\$\begingroup\$

If I shoot at f1.4, I can easily separate my subject from background due to the shallow DoF.

In the case where I turn the focus to infinity at f1.4, I will have everything in focus, except the objects which are in the actual focus distance of my lens.

Is it better approach to use f1.4 for stars, city/mountains landscapes and other type of photography where object of interest are very far from me and I want to have them sharp?

My old approach was to shoot all these types of panoramas with f8 or so, to achieve deeper depth of field and have everything in focus, but the result was not always as sharp as I wanted.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ You should identify which camera, and which lens you are using. You should also rephrase "Incase I turn infinity at f1.4, I will have everything in focus, except of the objects which are in actual focus distance of my lens ?" as this question (or statement) is confusing. \$\endgroup\$ Commented Nov 13, 2019 at 18:44
  • \$\begingroup\$ I use mft Panasonic GX8 but the principe is the same for any other camera... lens 12-60mm 3.5-5.6 or 1.8 45mm \$\endgroup\$
    – JZK
    Commented Nov 13, 2019 at 18:48
  • \$\begingroup\$ The principles are not the same for all cameras. A 20MP Micro Four Thirds camera will suffer from diffraction at larger apertures than a 20MP Full Frame camera. If your f/8 photos are not sharp, it may be due to diffraction. \$\endgroup\$ Commented Nov 13, 2019 at 19:00

5 Answers 5

5
\$\begingroup\$

If I'm getting the question correctly, it is about whether at infinity focus you ought to use wider apertures like f/1.4 or narrower apertures like f/11 to get the sharpest results? The answer is, it depends... First let's look at what depth of field actually is at infinity focus.

The depth of field depends on three things:

  1. Focus distance,
  2. focal length,
  3. aperture (the f-number) and
  4. sensor or film frame size.

Focus distance is the distance between your camera (specifically the sensor or film) and the subject you are focusing on. That's usually expressed in feet or meters. Focal length is the distance between sensor/film and the convergence point of light in the lens. It's usually expressed in millimeters (50 mm prime lens, 28-135 mm zoom lens...) and determines the field of view. The f-number is the ratio of the focal length to the apparent diameter of the aperture iris as seen through the lens' front element. Usually you see this expressed as the inverse of that with "f/" in front. A 50 mm lens with an apparent iris diameter of 10 mm would have an f-number of 5 (50 mm / 10 mm), usually denoted as f/5. The same 50 mm lens with its iris open to 25 mm as seen from the front would have f-number 2, denoted as f/2.

Depth of field decreases with wider apertures. For a given focal length and focus distance you get a shallower depth of field at f/2 than you would at f/8. Depth of field decreases with greater focal length. For a given focus distance and f-number, a 50 mm focal length would give a greater depth of field than a 100 mm focal length. Depth of field increases with focus distance. For a given f-number and focal length, focusing on a subject at 20 meters away gives you a greater depth of field than a subject at 10 meters away. Details about why this is can be found on this site and many other resources. This is all due to how lenses and projections work, combined with physical limitations of materials like sensors and film.

The sharpest focus will be anything in the field of view at the focus distance. So if you focus your lens for 3 meters away, things at 3 meters will be sharpest regardless of focal length or aperture. In front of that imaginary spatial plane (between camera and focus distance) is an area of acceptably sharp focus, and also behind (between focus distance and infinity). The sum of those two distances is the depth of field. Important to know is that the "sharp" area between camera and subject is shallower than that behind the subject. Let's look at some examples. I'll be using an online DOF simulator. In case that link goes offline, you can easily find many more, or download an application for desktop or mobile devices.

Imagine we're shooting with a 50 mm prime lens. The subject is at 3 meters distance and we correctly focus. The aperture is set at f/3.2. The camera has a full-frame sensor (or 35 mm film). The total depth of field then comes out to 66.5 cm. 29.6 cm of that range is in front of the subject, 36.9 cm is behind the subject.

Move the subject to 5 meters and focus correctly, and you get a 190 cm depth of field of which 77.6 cm is in front and 113 cm behind.

Keep focusing further away and the depth of field will increase. The depth of field behind the subject will also increase at a greater pace than in front. With a 50 mm lens and f/3.2 focused at 20 meters, the depth of field becomes 65.59 m with 8.51 m in front of the subject but a big 57.08 m behind the subject.

At some point the part of the depth of field behind the subject becomes essentially infinite, even before you focused at infinity. This is known as the hyperfocal distance. For a 50 mm focal length at f/3.2 the hyperfocal distance is 26.94 m. This means that if you focus your lens at that distance or further, you're guaranteed that the "acceptable sharpness" stretches on to infinity behind the subject. You still get a good deal of depth of field in front of the subject too, but not all the way to the camera.

Hyperfocal distance is very useful for certain types of photography. For example in landscapes, if you know the hyperfocal distance for your chosen focal length and aperture then if the subject you focus on is at least that distance away, you know for sure that you're also getting everything behind it (like mountains, distance clouds, stars...) in focus.

When you focus to infinity the depth of view stretches from the hyperfocal distance to infinity. So that's a second useful aspect. Focusing to infinity and knowing the hyperfocal distance tells you how close something can get to you before it starts getting out of focus.

So that brings us to your question. If you are shooting things that are far enough away so that you need to focus to infinity or at least very close to that setting, you shouldn't need to worry about a wide aperture causing you to miss focus. Even with a 200 mm lens at f/1.8 (if such a thing exists) the hyperfocal distance on a full-frame sensor would be about 775 meters. Focus for infinity and everything further than 775 meters will be sharp. Focus at 775 meters and you'll get a depth of field from 387.3 meters away to infinity.

Assuming things are as far away as you think and you manage to focus at least to the hyperfocal distance you won't get pictures lacking sharpness due to a "too shallow" depth of field. So what could still cause you to get pictures that aren't as sharp as they should be?

First there's of course equipment. Quality lenses and a camera with a good sensor (or high quality film if you go analog) will tend to yield better results.

Assuming that's at acceptable levels you'll need to be able to keep the camera stable. Hand-held shots will seldom manage to be sharp at longer shutter times, especially as the focal length becomes longer. With a 50 mm lens and no other stabilization tech any shot with a shutter speed slower than 1/50 second is a gamble. If the camera is stable (on a solid tripod, not in anything heavier than a breeze) you can increase sharpness by using a remote to fire the shutter and using the lens lockup feature if the camera has it.

Narrower apertures lead to longer shutter times because you get less light onto the sensor/film. That is detrimental for hand-held shots. This might be why you found better results at f/1.4 than at f/8. But even if your camera is mounted on a tripod longer shutter times might reduce sharpness through motion blur. This could be due to vibrations from the mirror and/or shutter (more pronounced at specific speeds), slight movement due to wind or simply the scenery moving. When shooting stars or the moon it's easy to underestimate the rate at which the night sky changes. An exposure of even a few seconds would be enough to have stars start to streak and wash out details on the moon. So a narrow aperture might simply not leave a short enough shutter time to get a sharp shot. When it does you might have to crank up the ISO so much on a digital camera that a lot of noise is introduced.

But always going for the widest aperture isn't necessarily the best choice. Again due to the physics behind lens design, the sharpest area of the projection will be at the center if the image with the quality diminishing towards the edges and corners. That's where distortion, chromatic aberration and blur start to play a bigger role. Lenses tend to have a sweet spot in their aperture range where this is minimized. Typical numbers are f/5.6 and f/8. There's a point of diminishing returns in narrowing the aperture. At values like f/16 and higher you might start losing sharpness again due to diffraction. You can usually find information about which apertures a lens performs best at, make an educated guess or experiment to find out.

So in the end you need to consider things in this order if you want to shoot things at long distance.

  1. What's the subject? Is it only things in the distance or are there foreground elements you want sharp? Compose to find the suitable focal length.
  2. Take the closest thing that needs to be sharp and estimate the distance (or use assistance like auto-focus and a read-out of the resulting focus distance). Find the widest aperture you can use to focus on that thing and still have the DOF extend to infinity behind it, or focus behind it and still have it in the DOF before the focal plane. If it's further than the hyperfocal distance you can just focus at infinity.
  3. Now find out the proper shutter time at that aperture and for the given lighting conditions. If it's fast enough that you can narrow the aperture a bit more and sacrifice some shutter speed, do so if a narrower aperture puts you closer to the ideal value for sharpness.

For a digital camera you also want to take into account that you wish to keep the ISO as low as possible to reduce noise in the image.

\$\endgroup\$
4
  • 1
    \$\begingroup\$ DoF also depends on viewing conditions. The same exact image file displayed at two different sizes and viewed from the same distance will have two different depths of field. \$\endgroup\$
    – Michael C
    Commented Nov 16, 2019 at 1:03
  • 1
    \$\begingroup\$ Ultimately, DoF depends on two things: aperture and total magnification. Everything else (focal length, subject distance, sensor size, display size, etc.) are factors that affect magnification That is, they affect the angular size of the actual object as seen by the camera compared to the angular size of the object in the displayed photo as seen by the viewer from a specific distance. \$\endgroup\$
    – Michael C
    Commented Nov 16, 2019 at 1:05
  • \$\begingroup\$ @MichaelC Good point. I avoided going into the details regarding the circle of confusion and other factors to focus on what's relevant for the question but it's certainly useful to consider DOF as a function of aperture and final image or print. \$\endgroup\$
    – G_H
    Commented Nov 18, 2019 at 8:57
  • \$\begingroup\$ Quite well elaborated underlying physics with specific examples... \$\endgroup\$
    – NitinSingh
    Commented Nov 22, 2019 at 17:44
4
\$\begingroup\$

Assuming you are shooting from a tripod with image stabilisation off (and using self-timer or remote control so that camera vibration is not an issue) and with exposure times where noise does not start to accumulate, the sharpest images usually result for narrower apertures, except that you don't want to have diffraction (which occurs at the narrowest apertures) have a significant influence.

Usually one expects a lens' "sweet spot" where further narrowing of aperture does not significantly increase results to be half a stop to two stops narrower than maximum aperture. For objects at infinity, results will also depend on how good your focus/autofocus is at actually hitting infinity and how good your lens is at maintaining infinite focus over the field of view.

Usually one defines the "hyperfocal distance" to be the distance where infinity is considered to be at the far end of the depth of focus but its actual distance depends on just what kind of unsharpness one considers to be still in-focus.

With an aperture of 1.4, the hyperfocal distance will actually be quite long (meaning that only rather distant objects can be considered in-focus) but if you are aiming for actual objects close to infinity, you'll get slightly sharper results for actually aiming at infinity. Since lenses tend to have some minimal amount of variation (particularly when being detachable), it's usually most reliable to let the autofocus provide the aim, assuming that the target offers enough substance for the autofocus to work on.

Unless you are short of light, going faster than F2.8 on an F1.4 lens is probably not helpful for maximum sharpness.

\$\endgroup\$
3
\$\begingroup\$

As I understand the question, you're asking about lens sharpness/IQ when depth of field is not a concern.

Light itself is sharpest when it is not bent through an aperture restriction, but that can't happen. So light is sharpest when it is bent the least... i.e. the largest aperture opening/smallest f#, resulting in the least diffraction.

But for that to actually be the case requires a lens that is maximally optically corrected within its' aperture range. Such a lens is called "diffraction limited," and there are not many like that.

The Olympus 45/1.8 is not diffraction limited... it is sharper stopped down to f/4 because that removes optical errors; which is more of a benefit than the loss due to diffraction increasing is detrimental. https://www.opticallimits.com/olympus--four-thirds-lens-tests/704-oly45?start=1

But interestingly the Pansonic 12-60/3.5-5.6 is a diffraction limited lens... at each zoom position it is already as sharp as it can be when at its' widest aperture. And at each subsequent aperture it becomes less sharp due to diffraction increasing. https://www.opticallimits.com/m43/983_pana1260f3556?start=1

AFAIK, there are no perfectly corrected diffraction limited lenses. So one lens being maximally sharp at max aperture doesn't necessarily make it sharper than another lens.

The last part is how much resolution can you actually record? A lens that is perfectly sharp at f/4 can resolve 21MP in the red wavelengths on M4/3, and even more in the green and blue wavelengths. And the GX8 has a 20MP M4/3 sensor... so you can't actually benefit from a lens that is diffraction limited at an aperture wider than f/4. https://www2.uned.es/personal/rosuna/resources/photography/Diffraction/Do%20sensors%20outresolve.pdf

\$\endgroup\$
1
\$\begingroup\$
Is it better approach to use f1.4 for stars, city/mountains landscapes and other type of photography where object of interest are very far from me and I want to have them sharp?

It depends (stars and landscapes have different answer).

For stars, you want:

  • The largest sensor you can afford. For most serious astrophotographers, it'll probably be a full frame sensor as the benefits of full frame can be easily seen with astrophotography: full frame collects more light and therefore is less noisy.

  • The fastest and (if you want wide-angle milky way photographs) widest lens you can afford. It certainly won't be 45mm f/1.8 or 12-60mm f/3.5-5.6. (Unless your budget is really that small.)

For landscapes, you want:

  • A very sharp lens, typically wide angle. 45mm on a MFT sensor probably is too long. 12-60mm f/3.5-5.6 could be it @12mm, but then again it's a very cheap lens and this usually means not sharp. The maximum aperture does not matter, the important parameters are sharpness and focal length.

Additionally, for stars, you will be:

  • Using your lens wide open, at the maximum aperture. Stars are so far away that any depth of field issues will completely vanish. However, you will have focusing issues with stars. The way to focus is to focus manually on some bright star and make it as small a point of light as possible. Zoom on the LCD / viewfinder is your friend.

For landscapes, on the other hand, you will be:

  • Shooting at the diffraction limited aperture, because anything wider makes nailing focus harder and also the depth of field doesn't necessarily include all of the parts of the landscape photo if you are shooting at f/1.4 on a fast / not-very-short lens and there are some nearby objects. Anything narrower on the other hand means you will encounter diffraction that limits the sharpness of your photos. However, some advanced software (example: Canon's digital lens optimizer (DLO)) can sometimes counter the effects of diffraction. Your GX8 appears to be one of the cameras to have anti-diffraction processing.

For stars, you really have to use manual focus. If you encounter autofocus issues, for landscapes you might want to try manual focus as well.

Now, what is the diffraction limited aperture? For typical APS-C 24mpix, it's around f/6 - f/7. For your camera, it appears to be f/5.6. So, I would shoot at f/5.6.

\$\endgroup\$
0
\$\begingroup\$

Most camera lenses are preset with an index mark and a stop for infinity. In other words, turning the focus ring, retracts the optics and the mechanism halts this travel. The lens then exhibits optimum focus for infinity (as far as the eye can see ∞).

Not all setups provide such a “stop”. Given no stop, we fall back on the viewfinder / rangefinder method provided by the camera with lens. My technique has always been, rotate the focusing wheel/knob until it stops and shoot at this focus setting. No failures ever!

Further, when focusing, I set the lens aperture wide-open. At maximum aperture, depth-of-field is at its shallowest, this increases the likelihood that my focus technique will deliver the best infinity focus. Again, no failures ever!

Now the rest of the story: As light enters the camera lens, the amount of energy that enters is always restricted by the blades of an iris diaphragm. Some optical systems omit the iris, in this case, the restriction is the inside diameter of the lens barrel. As the imaging forming rays interact with the knife edge of the iris, twin demons are at work that reduce the acuity of the resulting image. Well studied by Lord John Rayleigh 1842 - 1919 Nobel prize physics 1904.

Studied at length by Lord John Rayleigh 1842 ~ 1919 British Nobel prize physics 1904 As these rays brush by the blades of a diaphragm, some are not blocked completely, they are shadowed and caused to bleed in intermingle with non-restricted image forming rays. The result is a drop in resolving power as the lens is stopped down. Table of resolving power for 589 millmircons lpm (lines resolved per millimeter)

f/1 = 1392 lpm

f/1 1392 lpm

f/2 696 lpm

f/2.8 487 lpm

f/4 320 lpm

f/5.6 249 lpm

f/8 184 lpm

f/11 127 lpm

f/16 87 lpm

f/22 63 lpm

f/32 44 lpm

Let me add, other lens defects are at work. Therefore I suggest, images of stars and distant objects are best imaged by stopping down the lens approximately 2 f-stops. This will likely be the setting that delivers the highest acuity.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ You completely ignore the dimness of stars, meaning collecting light at a fast aperture to have reduced noise may be more important priority than the maximal lens sharpness. \$\endgroup\$
    – juhist
    Commented Nov 15, 2019 at 17:24

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