Yeah, my tests confirmed at least f/11 with override, but with a sensor this small (1/2.5") the diffraction limit is a serious problem at smaller apertures. Run the calculator
here with the 1/2.5" sensor from the A590 -- even at f/2.6, diffraction reduces effective resolution (slightly). At f/8.0, the airy disk is more than 5x the pixel size. Even if you resample the image down to 1/2 vertical and horizontal resolution (2 megapixels), it won't regain the sharpness lost to diffraction.
I strongly suspect that f/11 isn't going to be worth it for the extra depth of field, since you'll effectively be throwing a huge gaussian blur on everything, and DoF stacking is probably the best option. I'd like to offer a giant "F*** YOU!" to quantum mechanics for giving us diffraction limits.
Harware ISO Limits:Lowest ISO: ISO50Highest ISO: ~1600Test procedure: (Based on CHDK features suggested tests, tell me if this isn't sufficiently accurate)
* Rest camera on hard surface with view of variable-brightness scene. Make sure there's a good could good peaks in middle of histogram, by adjusting colors & Ev compensation
* Set Disable Overrides to OFF
* Set ISO to 80 in overrides, and Bracketing in Continuous mode to steps of -5 ISO.
* Set custom timer to 30 sec delay (to allow time for vibrations from pressing button to settle) and 10 shot sequence
* Shoot, making sure nothing moves in scene or no bumps to camera
* Sequence 2: set ISO to 60, bracket in steps of -2 ISO rating, and same timer settings
* For each sequence, find where histogram no longer changes
For high-ISO:
* Same, but started at 1600, increasing in steps of 200,300,50
I'm having a hard time evaluating the sequences for the maximum ISO value, because after the first photo, the histogram seems to smooth out somewhat, without actually moving to the left or right. It'd be awesome if someone could check me on that before I post this set of limits on the wiki.
