Most Micro Four Thirds lenses have internal focusing, meaning that some smaller lens elements inside the lens move to achieve focus. This is generally fast and silent.
To illustrate the difference in focus noise, I made some measurements using an Iphone. Now, I don't believe the Iphone is very well calibrated. So the absolute noise measurements are probably not too accurate. But the relative noise levels can be studied this way.
To measure the noise, I put various lenses on the Panasonic GH2, and put the camera in continuous autofocus mode, AFC. In this mode, half pressing the shutter will make the focus job back and forth continually. I pressed the shutter 3-5 times for about one second each time, to capture the noise level.
Here is a video showing the study:
Lumix G 8mm fisheye
|50 46 52 51||50 dB|
|Lumix G 14mm||49 51 47 48||49 dB|
|Lumix G 20mm||61 60 58 59||60 dB|
|Leica Lumix DG 45mm macro||51 52 50 52||52 dB|
|Lumix G HD 14-140||55 62 65 62 61 60||61 dB|
|Lumix G 45-200||50 53 56 54||54 dB|
|Olympus ZD 4/3 50mm f/2||76 79 81||79 dB|
The outcomes are mostly as expected. The Lumix G 20mm f/1.7 pancake lens is the odd man out in this test. Since it is the only Micro Four Thirds lens that features a traditional focus mechanism, and not internal focus, it is also the most slowly focusing and noisy. The lack of internal focus means that the whole focus assembly moves during focus. You can see this in the video.
Perhaps the only surprise in the test is the relatively loud focusing noise from the Lumix G HD 14-140 superzoom lens. This lens is marketed as a video optimized lens, and features fast and silent operation. I would speculate that the noise picked up is not only from the focusing, but also from the OIS unit, which is powered when half pressing the shutter. Also, since this is the largest lens, the sound comes from a source closer to the mobile phone picking up the noise. Hence, the sound reading should be expected to be more audible. So I would take the measurements from the 14-140mm lens with a grain of salt.
The Olympus ZD Four Thirds 50mm f/2 macro lens is not a contrast detection autofocus (CDAF) optimized lens, and therefore it is rather slow to focus. It also generates a lot of noise, since it moves large lens groups back and forth. This lens is mostly useful for studio work on Micro Four Thirds, I'd say, due to the very slow autofocus.
The final word is that most Micro Four Thirds lenses, except the 20mm pancake, focus very fast and virtually noiseless.
Another aspect is the noise from the aperture change. I have not measured this, but my experience is that this is rather noisy with all lenses, except the Lumix G HD 14-140 superzoom lens. The 14-140mm lens has an almost inaudible change of aperture.
I would have liked to test the Lumix G 14-42mm f/3.5-5.6 kit lens. However, I took my lens back to the store due to a bad aperture diaphragm. After more than two months, they have still not fixed it.