Mike K
Offline
• • • •
Upload & Sell: Off
|
 p.9 #16 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes | |
kevindar wrote:
How is this method any better than the computer patter interference method? for one, knowing you have achieved critical focus with the computer screen method is a lot more easier. second, why move all around when you can just repeatedly and reliably see the focus is not moving/changing, once you have achieved perfect phase detect focus?
Kevindar, see my slightly different AF calibration method at the beginning of this thread.
Both approaches use magnified Live View as the "standard" of accurate focus. I suggest an interference pattern only because it can be used on screen, without having to print it out. I like it as the pattern appears to change as the the focus changes, so its easier to assess AF comparisons. However, as the focal length gets much longer, the recommended 50X focal length will place you away from your computer, perhaps outside.
The main difference is that the dot tune uses the focus confirm "dots" as a measure of establishing the boundaries of PD AF. I simply press AF to see if magnified LV MF matches the PDAF setting. If not, change the micro adjust setting until Magnified LV matches PD AF. I repeat this comparison several times until I am satisfied I am within the normal imprecision of the PD AF method.
I believe Snapsy developed this AF calibration method on Nikon bodies, where the focus confirm lights provide more directional information than with the Canon set up.
As for the 50X focal length, this was suggested by Canon (yes Chuck Westfall) to minimize the impact of focus shift, a common phenomenon with wide aperture lenses (ie f1.2) at shorter (portrait or closer) distances. Obviously calibrating using the distance you are going to shoot at is the most accurate answer (but only for that specific distance).
Mike K
|
, I'll get it going when DotTune becomes available in the next non-beta release of ML.