Is there a real benefit from having in-lens AF motors?

[squeamishossifrage]

I have half a dozen lenses with in-lens AF motors, but that has never been a consideration when choosing what lenses to take with me on an outing. The majority of my lenses rely on the in-camera motor. Also, motor placement has never been a factor in choosing lenses to buy, as my cameras all have in-bogy motors.

The one advantage I can see is that if the motor fails in the lens, then you are just one AF lens reduced to manual, but if the camera motor goes south, then all your lenses go MF. However, all seven of my in-camera bodies have never had a motor failure. In fact, when my original Minolta 7000 failed after 20 years it was not the AF that spelt its demise.

The industry push towards mirrorless in recent years has led to a simplification in the manufacturing process, as has the reliance on in-lens motors - but without the concomitant anticipated reduction in price. 

Are there any strong arguments or opinions out there in favour of in-lens AF motors, or is just a way of making more money for lens, and indirectly, body, manufacturers?

[petrochemist]

Jan 14, 2023 14:52:36 GMT squeamishossifrage said:

I have half a dozen lenses with in-lens AF motors, but that has never been a consideration when choosing what lenses to take with me on an outing. The majority of my lenses rely on the in-camera motor. Also, motor placement has never been a factor in choosing lenses to buy, as my cameras all have in-bogy motors.

The one advantage I can see is that if the motor fails in the lens, then you are just one AF lens reduced to manual, but if the camera motor goes south, then all your lenses go MF. However, all seven of my in-camera bodies have never had a motor failure. In fact, when my original Minolta 7000 failed after 20 years it was not the AF that spelt its demise.

The industry push towards mirrorless in recent years has led to a simplification in the manufacturing process, as has the reliance on in-lens motors - but without the concomitant anticipated reduction in price. 

Are there any strong arguments or opinions out there in favour of in-lens AF motors, or is just a way of making more money for lens, and indirectly, body, manufacturers?

I would think the reduced linkages make therm quieter & quicker, certainly my screw drive AF lenses are neither quiet or quick. Optimising the motor for the lens also becomes a possibility but that might mean they have less spare capacity & so fail easier...

[peterob]

How long ago was the last camera made that physically drove the AF? It seems a primitive way of doing it now.

[nickr]

In-lens motors can be tailored for the particular lens in use. For most, that's not all that critical, but for superteles etc it makes a huge difference to AF performance.

In-lens motors can be of the ultrasonic type, which are much quieter than screw drive lenses, or stepper motors, often better for video - in short, you can tailor the motor type as well as size to the required spec for the lens in question.

In lens motor lenses generally only have electrical connections rather than some mechanical ones, which can lead to less breakdowns.



30 years ago, when I was looking to buy my first AF SLR, I tried out the 3 contenders in the shop: Minolta Dynax 7xi, Nikon F601, and Canon EOS 100, each with their normal zoom. The Minolta focused quickly but fairly noisily, the Nikon focused slowly and very noisily, and the Canon focused quickly and silently. It seemed like magic compared to the others thanks to the USM lens, and that was that. I've never regretted going for an in-lens system since, it seems like a far better solution to me.

[squeamishossifrage]

Jan 14, 2023 15:12:23 GMT petrochemist said:

I would think the reduced linkages make therm quieter & quicker, certainly my screw drive AF lenses are neither quiet or quick. Optimising the motor for the lens also becomes a possibility but that might mean they have less spare capacity & so fail easier...

Certainly I'll give you quieter, but my two 24-70mm f:2.8 lenses show no difference in speed, although the in-body version has a wider angle to cycle through. On the optimizing front, the in-lens gearing can be just as optimized as the in-lens motor.

I remain unconvinced! 

[squeamishossifrage]

Jan 14, 2023 15:18:39 GMT peterob said:

... It seems a primitive way of doing it now.

That comes down to an engineering design philosophy - is it better to have  a short distance between the servo and the controller and a longer mechanical linkage. My only experience of this is from yacht autopilots, where it was definitely more reliable to have the two integrated, and a chain or rod mechanical drive to the rudder. 

[squeamishossifrage]

Jan 14, 2023 15:22:09 GMT nickr said:

In-lens motors can be tailored for the particular lens in use. For most, that's not all that critical, but for superteles etc it makes a huge difference to AF performance.

In-lens motors can be of the ultrasonic type, which are much quieter than screw drive lenses, or stepper motors, often better for video - in short, you can tailor the motor type as well as size to the required spec for the lens in question.

In lens motor lenses generally only have electrical connections rather than some mechanical ones, which can lead to less breakdowns.



30 years ago, when I was looking to buy my first AF SLR, I tried out the 3 contenders in the shop: Minolta Dynax 7xi, Nikon F601, and Canon EOS 100, each with their normal zoom. The Minolta focused quickly but fairly noisily, the Nikon focused slowly and very noisily, and the Canon focused quickly and silently. It seemed like magic compared to the others thanks to the USM lens, and that was that. I've never regretted going for an in-lens system since, it seems like a far better solution to me.

Regarding your first paragraph, yes, but the drive gearing can be just as tailored to the task in hand. and both philosophies rely on the power from the camera, be it electrical or mechanical. Depending on the lens, there's quite a lot of power that needs to go through that tiny pin-head to flat-plate spring connection.

On your final paragraph, if I were buying new today, I would go for an in-body motor for the camera, as although very few cameras are still being made with in-body motors, it does give you access to a whole range of older lenses. Any new lenses I buy, however, would be with in-lens motors, as these tend to be the newer ones with the best optical performance.

[andy]

Less mechanical linkages should mean less slop.

[squeamishossifrage]

Jan 14, 2023 16:38:04 GMT andy said:

Less mechanical linkages should mean less slop.

True enough, but it can actually be programmed out, although the older the lens gets, the sloppier it gets, and correspondingly slower and noisier. I have one lens, a 1987 Sigma telephoto, that when it focusses you can hear the motor going back and forth for about a second, but it eventually settles in the right spot.

[geoffr]

I have a Nikon 180 f2.8 AF lens, I bought it cheaply because it was damaged, as a result the drive mechanism is sticky at some points and mechanically the focus ring sticks slightly too. On the Nikon F4 it was fine but on later cameras, D2X, it wouldn't focus beyond the sticky patch. The reason being that the motor in the D2 didn't deliver sufficient torque for the lens. I haven't used it for a long time.

I actually have only three lenses with built-in motors and they are all much faster than their predecessors with camera driven focusing. It requires less energy because the motor doesn't have to drive the shaft between the motor and the focusing mechanism. Losing the shaft also removes the friction in the supports and the possibility of wear. There are no bearings to get contaminated, the shaft requiring a minimum of one at each end. Overall the camera and lens can be lighter because the motor drives the mechanism without the need for intermediate mechanical couplings. Wires are lighter and go round corners much more easily than drive shafts.

I can think of very few autopilots that have a motor driving a long linkage to the control actuator, wire is lighter and redundancy is much easier. True there are auto-throttle systems that drive the throttle levers and rely on the cables between throttle/s and engine/s but I can't think of any that do the same with flying controls. The 737-200 has power assisted controls and the a/p actuator acts on the assistance actuators rather than the ailerons, can't remember the elevator system well enough. Generally a/p actuators are located with the power control actuators and provide the input to them.

I still have a couple of cameras with in-body motors and they focus slowly even with in lens focusing motors leaving me unsure as to whether the reason for slow AF is entirely due to lens design. Assuming that the focusing speed is dictated by the speed at which the lens can be driven. I guess 30 years ago AF was generally slower than it is now. One other major consideration is that buy driving the lens focusing elements directly there is no/little backlash resulting in closer to deadbeat operation. A major consideration for a long lens.

[geoffr]

Jan 14, 2023 17:04:58 GMT squeamishossifrage said:

Jan 14, 2023 16:38:04 GMT andy said:

Less mechanical linkages should mean less slop.

True enough, but it can actually be programmed out, although the older the lens gets, the sloppier it gets, and correspondingly slower and noisier. I have one lens, a 1987 Sigma telephoto, that when it focusses you can hear the motor going back and forth for about a second, but it eventually settles in the right spot.

Would that be acceptable when focusing on a fast moving bird or when tracking an erratically moving subject? I can't use my 300 f4 AF (motor in body) to track birds but my 70-200 f2.8 AFS (motor in lens) with a 1.7x converter performs brilliantly in the same circumstances. Seems a good argument for in-lens motors to me.

[squeamishossifrage]

I agree, for that particular lens,  but on the other hand, my Sony 75-300mm in-body AF focusses just as fast as my Tamròn 70-300mm USD (though the camera is also a Sony, which may make a difference). 

[geoffr]

Quite honestly I think you are on a hiding to nothing. Lenses without focus motors are generally older and it is likely that the newer cameras that can drive them are limited to focusing them at an appropriate speed to ensure focus accuracy. Newer lenses tend not to use the camera's motor. Proving experimentally that a lens without a focusing motor can focus as quickly as one with is only going to be possible if the two lenses are of similar age and tested on the same body. I don't know about any other manufacturer but Nikon didn't make a 24-70 f2.8 without a motor, nor a 70-200 f2.8 so I can't begin to try. Even if they had done so would a D5 drive the mechanically coupled lens as fast as an electrically coupled lens?

There are just too many variables to allow a realistic comparison, not to mention that many other factors are in play, older lenses may have a shorter focus action than newer ones. Cameras may be designed to go easy on older lenses.

All I can say is that in my experience a lens with an internal motor out performs one without. As the lenses I use aren't available any other way I'll not concern myself with the question as it is irrelevant. Given that I am not a camera collector I am not about to step backwards and buy lenses that don't perform as well as the ones I have. Additionally, I only have teleconverters for AF-S lenses so lenses without motors won't work, even though all my cameras could drive them without the converters.

[Chester PB]

Jan 14, 2023 15:22:09 GMT nickr said:

[Chester PB]

The in-body focus motor works fine on my Pentax DSLR body: none of my lenses have built-in focus motors, all are autofocus and the youngest one is probably 12 or 13 years old.
The only lens with which I notice the slow (by current standards) autofocus is at the long end of a 20 year old Tokina 80-400, which is also the largest and heaviest lens I own, but I'm rarely in much of a hurry anyway.