Not only, if has an autofocus option (--autofocus flag). However, you cannot select the focus area (or focus points like in a pro camera) :(. But, the autofocus function is implemented and works nicely.
Manufacturers should stop calling it a day right after they get their stuff to "sort of" work. Especially if they make DIY products. Passing on the complexity of hardware interfaces to novice users is just cruel.
You mean using this hack: https://blog.robertelder.org/recording-660-fps-on-raspberry-... I didn't check that yet, but using the provided software and the flag (--framerate), I got the following timestamps:
# timecode format v2
0.000 8.329 16.657 24.987 33.316 41.644 49.972 58.302
Meaning that the camera is working at max. 120fps@640x480
I thought you were talking about my "autofocus" project (https://www.youtube.com/watch?v=PrbyPmq_Z7Q). The arducam-12mp-pan-tilt-zoom uses the same sensor (IMX477) as the HQ. Thus, the resolution claim should be the same ;).
I used two steppers to turn the aperture and the focal length rings of the lens independently. Then, you can set the aperture ring to 1.4, 2.8, 5.6, 8, or 16 and then focus the image with the other ring/belt. Both rings or belts move independently. In the beginning, I used a calibration routine to identify the maximal or minimal position of the rings. Both limits are detected by moving the rings in both directions and measuring the stepper currents. When the stepper reaches a limit, a current peak is measured, and that gives me the minimal o maximal angle.
That can be also done. The pins are there (as you can see on the board design). I just wanted to separate the threads. It was easier to control the steppers with the M5Stack.
