Sounds from IS

I've got no L lenses, but I do have a Canon EF-S 17-85mm IS (Image Stabilization) lens for my 20D. I'd never noticed any constant noise before (the focus and aperture motors make noise, but only intermittently). Putting my head right next to it and trying with IS on and off, the IS does indeed make a constant buzzing/whirring sound noise while it's active.

Tim

Thats because there is a small giro motor,wih a mirror attatched,compensating your movements.

In the diagram of one I saw it's not a mirror, but a lens assembly. But essentially, yes - there is a mechanical device moving about inside the lens to provide the IS and it's that which is making the noise. The other system in use (generally on compact cameras) involves taking a rapid sequence of images with the CCD and shifting and recombining them in software. This is cheaper to do than physically moving optics about, but increases the readout noise by orders of magnitude as you're reading the CCD numerous times per 'frame'.

Tim

I originally thought the sound was related to a gyro but there is widespread agreement in my camera newsgroup that motion detection for optical image stabilization in lenses is done using solid state devices not the kind of mechanical device you're talking about. There is speculation that the sound is made by motor control circuitry used to move the optics. What kind of motor controller makes a constant whir sound?

Thanks, Ron

It seems much more likely that the sound is coming from the actuators which move the lenses; that is the mechanical device I was referring to rather than a classic gyroscope in the form of a significant rotating mass. That is the bit that moves significantly, so that seems to be the most likely source of the noise. The stabilising action comes not from gyroscopic action directly, but from a feedback loop fed from gyroscopic sensors. The sensors I expect will be MEMS devices, which aren't really solid-state as they do have moving parts, albeit incredibly tiny moving parts which make no noise.

Tim

I see.

It is my understanding that the noise is non-intermittent and continues even when the camera is placed on a tripod. Presumably this is some kind of polling behavior. Why is it necessary and why does it make a noise?

Thanks for the time, Ron

I think you'd have to ask a Canon engineer or someone who's had one to pieces and run a bunch of tests - unfortunately I'm neither. Speculating, I don't find it surprising that the actuator would always be moving (and thus making noise) while the system is active. Any sensor will produce some level of noise in its signal (in the electrical, rather than accoustic sense), as will the electronics betwen the sensor and actuator. This noise can be translated to small movements of the actuators back and forth around the desired position. Damping that noise, electronically or mechanically, would degrade the response of the system to rapid changes. It's not unusual for control systems to oscillate a little around their desired operating point, particularly if they are optimised more for rapid response than minimum overshoot and oscillation.

Tim