49

The Canon EF 40mm f/2.8 has a designation of STM on the lens. What does this mean? What are the advantages of having it and does it replace an older technology?

We have a terminology thread that usually covers these questions but this is not yet addressed in it.

Community
  • 1
dpollitt
  • 46,525
  • 20
  • 136
  • 266
  • 3
    I was just heading over here to ask the same thing myself! Great minds think alike... :) – Mark Whitaker Jun 07 '12 at 20:46
  • I mostly shoot video, will the Sigma 35 1.8 HSM be as quiet as the STM kit lenses (or quiet enough for video?) I would rather get the Sigma and be able to shoot wide open rather than limited by higher apertures with the STM kit lens. –  Jul 26 '13 at 19:59
  • @damon James - that would fit better as a new question then a comment on this question. – dpollitt Jul 26 '13 at 21:57
  • With STM, per a sales person, a lens can auto focus in movie mode. So, I was told. –  Aug 30 '13 at 12:58
  • 1
    @user21907 You're sales person doesn't understand what they're talking about. (Auto) focus in movie mode is a function of the body, not of the lens. STM lenses are better at it, but all my lenses will auto focus in movie mode. – Philip Kendall Aug 30 '13 at 13:19
  • 2
    @PhilipKendall See this quesiton. When the STM lenses were first announced, the announcement seemed to say that hybrid AF would only work in video mode with STM lenses. This turned out to not be true, but I can see a salesperson being initially confused (or, still confused). – mattdm Nov 01 '13 at 16:23
  • There's a difference between "can auto focus in movie mode" (which is what the anonymous user was told) and "can auto focus using hybrid AF in movie mode" (which is what the marketing materials implied). Canon may not have made it particularly clear, but the sales person had over a year to get it sorted out :-) – Philip Kendall Nov 01 '13 at 16:42

3 Answers3

58

STM stands for Stepper Motor and is a applied to a new range of Canon lenses which feature a new design of focus motors which, along with a new iris mechanism are designed to eliminate (auditory) noise during video recording.

Canon haven't revealed any information about how the new design works but it's probably the same type of motor used in mirrorless camera lenses. It's a more precise version of a regular DC motor but still has the same direct connection to the lens focus group, which means manual focus has to be implemented using a focus-by-wire arrangement whereby moving the focus ring by hand sends a signal to the motor to move the focus group.

In comparison an ultrasonic motor (like Canon's USM) consists of a pair of concentric rings which vibrate at high frequency to rotate back and forth, an arrangement which permits the user to move the focus ring to directly move the lens element, achieving full time manual focus without damaging the motor.

Stepper motors are better at producing smooth, precise incremental movements, such as those required by contrast detect AF, and AF during video. Ultrasonic motors are better at jumping to the right focus point as part of a phase detection system. See What is the practical difference between phase-detect and contrast-based autofocus?

Community
  • 1
Matt Grum
  • 118,892
  • 5
  • 274
  • 436
  • 5
    Stepper motors are a specific type of AC servo motor, As such, they're really not a DC motor. Additionally, DC motors can be made as or more precise then a stepper. The functional precision of a servo system is merely a product of the specific implementation, so saying a stepper motor is a "more precise" version of a DC motor is an over-generalization, in addition to being incorrect about the DC part. – Fake Name Jul 27 '13 at 09:32
  • 2
    Stepper motors rotate in fixed increments (steps) say 1/8 of a turn or 1/16 of a turn in response to a pulse of current. Steppers can be driven very precisely, in a digital fashion. – Kaushik Ghose Nov 02 '13 at 02:28
  • @FakeName, Why would you think a stepper motor is AC? And why would you think cameras have DC-to-AC inverters? And why are you colluding servos with motors? Stepper motors ARE more precise motors. Typical motors only control direction and power of rotation. Stepper motors drive to a specific radial output (the increments and therefore precision can vary). BOTH of these must be coupled with a sensor feedback loop for dynamic control. The sensors are the same. The control circuits aren't fundamentally different. ONLY the motor is different. And it's a pretty massive effect. – Travis Watson Oct 03 '14 at 21:45
  • 1
    @Travis - 1. ALL motors are AC. "DC" motors just use brushes internally to convert the DC power to AC. Stepper motors however do not have brushes, so you have to externally commutate them. Colloquially, this type of drive is referred to as "AC", since it requires driving current in both directions through the field coils (which is basically AC). 2. I'm using the engineering definition of servo motor. 3. Precision is a function of the complete control loop. The motor type is irrelevant as long as the control loop can compensate properly. – Fake Name Oct 03 '14 at 22:06
  • 2
    TL;DR You can use stepper motors as a servo motor in a part of a larger servo system. The term "servo" motor has been polluted by RC model community to refer to one specific type of servo system that's commonly used for controlling models, but that's not the only type of servo motor, or servo system. – Fake Name Oct 03 '14 at 22:08
  • 1
    Here is an answer I wrote on the electrical engineering stack exchange site about what, exactly comprises a servo-motor/servo-system/etc... – Fake Name Oct 03 '14 at 22:12
  • 2
    Oh, also, in the great majority of applications, stepper motors are run open-loop, which means they do not have any position feedback mechanism. The only real reason to ever use stepper motors is to avoid the cost of said feedback mechanism (typically an encoder of some sort). Once you have the feedback mechanism and the corresponding control loop to make it work, a brushed servo motor will almost always outperform a stepper motor. – Fake Name Oct 03 '14 at 22:19
  • @FakeName, some DC motor configurations alternate current, not ALL; not even all with brushes. Wrong #1. Regardless of your definition of "servo", feedback is irrelevant to the discussion. Wrong #2. It's far more common to have discrete DC paths in a stepper motor controller than alternating paths. Wrong #3. The precision of a motor is NOT the same as the precision of a dynamic control system, open- or closed-loop. Wrong #4. Stepper motor benefit for AF is incremental precision. That word, from OP, is incredibly important in context. Wrong #5. Ugh, why did I even bother to respond... – Travis Watson Oct 04 '14 at 02:44
  • 1
  • A dc motor that does not alternate is either a homopolar motor, or a solenoid. 2. Unipole stepper motors are not generally used, because they cost twice the copper for the same torque, and it's still not DC, because the current is not continuous, and it's effectively using the motor as a autotransformer to reverse the field. 3. I never made any claims about the precision of the motor, but rather was only ever speaking to the whole control system. 4. You can't have a servo system without feedback. Feedback is literally part of the definition of a servo.
    • – Fake Name Oct 04 '14 at 03:15
  • 1
    Further, I'd nitpick about the OP's description a bit. Unless the USM is categorically less effective at focusing, both motors have to have the same degree of precision. I think the difference here is the ability to better control the velocity, which is needed for gradual zooming/focus changes. This is not a function of precision. I'd guess the velocity control in a USM may not be very good. – Fake Name Oct 04 '14 at 03:20
  • 2
    I cede the point about DC. There is one type of DC motor: the homopolar motor. However, it's largely a scientific oddity, and not used in many practical applications, and no lenses as far as I can tell. The rest are still AC, though, and that very much includes all types of stepper motors. – Fake Name Oct 04 '14 at 03:26