[micromacro]

I wanted to ask it long time ago. Very often I see the statement online that the wider aperture lens focuses faster.

I thought that wider aperture makes lens faster because it allows to use faster speed with open aperture. How does it possibly affect faster focusing? Personally I did not notice the difference in focus speed, let's say, between Sigma f35mm f1.4, and Pentax DA*300mm f4.

Does wider aperture itself really affects focusing speed? I thought it's more about the quality of the lens, and the camera AF system. And again, searching today for the best focal range for tennis, I see that statement that f2.8 lens will focus faster than f4.
I don't understand that.

[skierd]

The answer is "it depends". Generally yes a faster lens should autofocus faster than a slower lens of the same type and build, depending on the situation, available light, and the cameras autofocus capabilities.

Autofocus systems use the light coming through the lens to determine whether or not the lens is focused on the object that the autofocus point is covering. More light (faster lens) means more light available to the AF processes to determine if focus is achieved and accurately, and therefore should be faster to lock focus. You'll notice the speed difference more indoors or in low light situations, or situations where there is low contrast.

Where things go haywire is when you start mixing autofocus drive systems, lens types and focusing group weights. In Pentax land, we have at least four autofocus drives now (screw, SDM, DC, and PLM) and each is different in terms of speed. Add in whether or not it's a zoom or a prime, and the weight of the glass being moved, and all of it gets a bit mucky in terms of focusing speed. In any event where talking fractions of a second here so except in a few situations like say sports, birds-in-flight, weddings, etc how fast the lens goes from completely out to in focus doesn't practically or drastically affect the image capture process.

As to why a 70-200 f2.8 might focus faster than a 70-200 f4, it's about the build and expected usage. The 2.8 is universally considered an essential workhorse lens for professional photographers and tend to be fairly expensive. They get stronger AF drives and other parts intended to make it focus faster and more accurately in addition to being faster. f4 lenses on the other hand tend to be aimed at serious enthusiasts and pros who want the same range, but are prioritizing the lower cost and/or the lower weight of a slower lens so it stands to reason that it will additionally have a lesser AF drive and other components that are more cost effective or lighter but slower to focus in addition to being a stop slower.

[pschlute]

"Fast" in reference to a lens refers to the maximum aperture. The wider the aperture, the faster the lens is. Your understanding of that is correct.

It is also true that certain lens/camera combinations can have a quicker autofocus with a faster lens. This is because more light gets to the AF points on the AF sensor. There is an optimum maximum aperture to get the best out of the AF system, perhaps others can explain in more detail. I believe certain Cannon cameras only enable all the AF points when a fast lens is on the camera. However the word "fast" is not a reference to AF speed when refering to a lens.

[micromacro]

Does it mean that if we take two close in features, or quality lenses, the same focal range, the only main difference is f2.8 and f4 , both for the same camera, will f2.8 lens focus faster than f4 lens when closed down to f4? Theoretically, if AF speed depends only on amount of light in this case, it should be the same AF speed for both lenses at f4?

---------- Post added 07-25-17 at 09:49 PM ----------

As to why a 70-200 f2.8 might focus faster than a 70-200 f4, it's about the build and expected usage. The 2.8 is universally considered an essential workhorse lens for professional photographers and tend to be fairly expensive. They get stronger AF drives and other parts intended to make it focus faster and more accurately in addition to being faster. f4 lenses on the other hand tend to be aimed at serious enthusiasts and pros who want the same range, but are prioritizing the lower cost and/or the lower weight of a slower lens so it stands to reason that it will additionally have a lesser AF drive and other components that are more cost effective or lighter but slower to focus in addition to being a stop slower.

Oh, ok. The more expensive lens f2.8 will cost more not only for the wider aperture construction, but for stronger AF drive.

[pschlute]

Theoretically, if AF speed depends only on amount of light in this case, it should be the same AF speed for both lenses at f4?

No, because all modern lenses AF at their widest aperture and the lens only stops down to the selected aperture (f4 in your example) after AF has locked on.

---------- Post added 07-26-2017 at 06:18 AM ----------

Oh, ok. The more expensive lens f2.8 will cost more not only for the wider aperture construction, but for stronger AF drive.

....and also maybe better lens coatings and better optical formular. The difference in price between two lenses is governed by a number of factors.

[mcgregni]

All the variables mentioned above will affect the focussing speed of any particular lens. There is no fixed relationship between aperture, amount of light entering the camera through that aperture, and the focusing speeds. In general more light reaching the AF sensors will improve focusing, so in general that means the wider aperture lenses have more chances of better AF performance than a lens with a smaller maximum aperture (eg F.4.5).

Just buying any F2.8 lens does not guarantee getting the best AF performance. My assumption is that other factors will play more significant parts in that ....especially aspects such as the camera focussing system, size and weight of the moving parts in the lens. I don't believe that the actual focussing technology (screw, DC etc) makes a big difference in itself. Big heavy DC lenses might not necessarily perform better autofocussing than small light screw drive lenses.

[pschlute]

An early riser too eh Nigel !

[stevebrot]

Does it mean that if we take two close in features, or quality lenses, the same focal range, the only main difference is f2.8 and f4 , both for the same camera, will f2.8 lens focus faster than f4 lens when closed down to f4? Theoretically, if AF speed depends only on amount of light in this case, it should be the same AF speed for both lenses at f4?

More light often means better AF performance, though wide maximum aperture lenses also tend to be somewhat soft with lower contrast wide open, both of which tend to degrade AF performance. What does that have to do with speed? Less than one might think. Consider a few points regarding PDAF...

• Before turning on the motor, the system knows what direction to go and a notion of how far to go
• Once the motor is engaged it will turn towards the presumptive point of focus at full speed until out-of-focus is no longer detected
• The more sophisticated systems may be able to predict the stop point to minimize overrun and jockeying near the point of focus

Note that if the system is able to detect a stop point, the only thing slowing the process down would be the distance to move the focus gear and time spent on jockeying and final check. Unless the sensor is abysmally insensitive (lots of hunting before signaling "not-out-of-focus"), motor speed is most important.


Steve

[mcgregni]

An early riser too eh Nigel !

On the train to work Peter! My earlies start at 7am!! ....... I've still got many years of commuting ahead before the days can be occupied by photography exclusively ..... If I'm lucky!

[beholder3]

For DSLR phase detect autofocus:

Very often I see the statement online that the wider aperture lens focuses faster.

That is simply wrong nonsense. It can not, simply due to physics.

Does wider aperture itself really affects focusing speed?

No. Never.

The PDAF systems used in DSLRs do not care about the aperture of a lens, because they are not impacted in any way with regards to speed.
The AF sensors see exactly the same image when a F5.6 superzoom is used and when a F1.2 lens is used.

The only single exception is that all modern systems use both F5.6 positioned sensors and F2.8 positioned sensors. The latter F2.8 ones are more precise, but not in any way faster. Beyond F2.8 (to F1.2) there isn't even any difference in precision. The F2.8 positioned sensors simply "shut off" with slower glass (they are completely blocked then and dont see any image).

Remember the AF sensors always see an image (actually only a row of pixels) at wide open aperture with a tiny, tiny sensor that works like a crop factor 10 (ballpark) ultra mini sensor (everything is sharp all the time).

[Adam]

I wanted to ask it long time ago. Very often I see the statement online that the wider aperture lens focuses faster.

I thought that wider aperture makes lens faster because it allows to use faster speed with open aperture. How does it possibly affect faster focusing? Personally I did not notice the difference in focus speed, let's say, between Sigma f35mm f1.4, and Pentax DA*300mm f4.

Does wider aperture itself really affects focusing speed? I thought it's more about the quality of the lens, and the camera AF system. And again, searching today for the best focal range for tennis, I see that statement that f2.8 lens will focus faster than f4.
I don't understand that.

It does, but only up to a certain point. If it's bright outside, slower lenses will allow for just as quick calculations, so most of the time you won't be able to see a difference. When there's something like 6EV or less ambient light, focusing performance starts slowing down, and a faster lens may give you a little more wiggle room before things get really slow.

The focusing technology is a bigger factor. The DA* 300mm has a slow motor and longish throw, while the Sigma has a faster motor and a shorter throw.

DC motors are generally faster than the old SDM in DA* lenses, while the latest PLM motor puts even those to shame. Screwdrive lenses are slower then PLM, but can compete with DC speeds if they have a short throw.

[Na Horuk]

AF is a complicated thing. A lens that is f8 will have trouble focusing in low light much sooner than an f2.8 lens. This can cause apparent "slowing down" of AF. But this does not mean an f1 lens would focus super quickly. AF has sensor points, which have a certain sensitivity (expressed in f number). So a lens that is faster than that number might have less trouble focusing in low light, but it might have problems with AF preciseness, and DoF being so shallow it actually seems off. f2.8 seems like a sweet spot where the AF system gets enough light to be fairly effective in low light, but it also gives enough DoF and contrasts to work well.

AF also depends on the focus throw and AF motor (and the weight of the moving parts), which means some lenses will achieve focus more quickly than others. DFA 100mm macro, for example, has a long focus throw because of the macro range. The speed of the screwdrive motor is still just as fast as always, but the lens has more ground to cover. Something like DA 40mm XS, with really short focus throw and no macro range, will achieve focus sooner, despite having same aperture and using screwdrive. One of the new lenses with DC motor might be even faster

The other correlation is that lenses with fast aperture are usually premium lenses. As such, they might have a better AF motor or better build optimization. This can lead to the impression that lenses with faster aperture have faster AF. It is not only due to the aperture, though.

So if you want fast AF, find a lens that fits:
a) Short focus throw (the amount you have to turn the focus ring to go from infinity to MFD)
b) Not macro (some macro lenses have "focus lock" button to prevent hunting in the wrong area)
c) Aperture between f2 and f4
d) Fast AF motor (screwdrive, SDM, DC, PLM in order; Third parties have motors like Piezo drive of Tamron or HSM of Sigma)
e) Modern lens coatings. Modern coatings allow good contrasts, good flare resistance, good colours, all of which can help the AF module.
f) Whether the AF assist light can help. In some cases the AF assist light can light up the subject so even a slower lens gets enough light. But with some lenses, the AF assist light is practically useless (particularly when focusing at super telephoto subjects or faraway landscapes)
e) There are reports that linear polarizing filters can confuse the AF. Circular polarizers are very common now, though. Just something to keep in mind if you use filters

Note that AF is only as fast as the weakest link. If you have no light, then even a fast AF motor wont help much. If you have fast aperture and good light, you might be limited by the cameras screwdrive engine.
Remember also that AF depends on the user. Are you using backbutton focusing? Are you using select AF point? AF.S, AF.C? Are you using Quickshift? Is the subject moving? Parallel to you or towards you? How is the light? AF is a tool to learn and then use to get what you need, not a tool to just blindly rely on.

Finally, some modern lenses are incredibly complicated. They even have programmable chips inside. In the case of some Sigma lenses, this seems to cause some AF problems on Kmount. Sigma 18-35mm, 30mm art, 35mm art all have really fast aperture and good focusing motors, but according to reports they seem to sometimes miss AF for unknown reasons. So what good is speed if its unreliable?

That is simply wrong nonsense. It can not, simply due to physics.

There we go. No need to wonder or think about it. Its "Due to physics" - all the answer anyone could ever want. When looking for a lens with fast AF, just ask yourself "which one is fastest, due to physics?"

[micromacro]

Guys, you are the best!

---------- Post added 07-26-17 at 08:33 AM ----------

Finally, some modern lenses are incredibly complicated. They even have programmable chips inside. In the case of some Sigma lenses, this seems to cause some AF problems on Kmount. Sigma 18-35mm, 30mm art, 35mm art all have really fast aperture and good focusing motors, but according to reports they seem to sometimes miss AF for unknown reasons. So what good is speed if its unreliable?

Yes, I noticed that. While having two same lenses for Canon and Pentax, sigma 31mm f1.4, I noticed that Canon version focuses precise and somehow faster that Pentax.

And Sigma 35mm f1.4 art for Pentax, which I bought brand new sometimes is hunting for focus when it's plenty of light. It's impossible to predict: at one moment focusing is immediate and precise, then it starts hunting. Not a big deal really when you have time to focus slowly, but still, it's almost $1k lens ...
In general, I decided to stay away from the modern 3rd party lenses after I buy Tamron 70-200mm, it will be the last non Pentax.

[beholder3]

There we go. No need to wonder or think about it. Its "Due to physics" - all the answer anyone could ever want. When looking for a lens with fast AF, just ask yourself "which one is fastest, due to physics?"

You don't seem to understand what I wrote. Read again. I just explained that PDAF focussing speed does not depend in any way on lens aperture. It can not. The AF sensors always get to see the exact same "image" exposed just the same (within its F5.6 or F2.8 or whatever limitations). That's a question of simple physics: the light rays to the AF sensors are totally not affected by lens aperture being F1.4 or F2.8.
The OP just asked a yes/no question: "Does wider aperture itself really affects focusing speed?" No. Never for typical DSLR PDAF.

[normhead]

Guys, you are the best!

---------- Post added 07-26-17 at 08:33 AM ----------



Yes, I noticed that. While having two same lenses for Canon and Pentax, sigma 31mm f1.4, I noticed that Canon version focuses precise and somehow faster that Pentax.

And Sigma 35mm f1.4 art for Pentax, which I bought brand new sometimes is hunting for focus when it's plenty of light. It's impossible to predict: at one moment focusing is immediate and precise, then it starts hunting. Not a big deal really when you have time to focus slowly, but still, it's almost $1k lens ...
In general, I decided to stay away from the modern 3rd party lenses after I buy Tamron 70-200mm, it will be the last non Pentax.

Interesting observation.

One thing that should be explored is the affect of lens design on AF. Is one 50 1.4 more compatible with your camera's system's AF than another? Photography has become people's attempts to make sense of systems they cannot possibly completely understand except through empiric evaluation. View cameras rock.

With all this technical knowledge bandied about, you still have to buy or rent the lens and try it to find out if you like it, or to know how it performs on your camera.