Hi, given I got no feedback here (http://forum.openframeworks.cc/t/reaching-avfoundationvideograbber-from-ofvideograbber-api/10303/0) I started implementing the necessary changes to implement those features.
Now, the native grabber is wrapped under several layers inside ofVideoGrabber.
My approach is to expose it (right now it’s unreachable) through a couple of getter methods, one to be added at ofiPhoneVideoGrabber.mm and the other at AVFoundationVideoGrabber.mm, and then you can get a hold of it:
//getting myself a pointer to the real device grabber
ofPtr<ofBaseVideoGrabber> grabberP = grabber.getGrabber();
ofiPhoneVideoGrabber * iGrabber = (ofiPhoneVideoGrabber *)grabberP.get();
AVFoundationVideoGrabber * avGrabber = iGrabber->getAvGrabber();
iPhoneVideoGrabber * deviceGrabber = avGrabber->getGrabber(); //finally reach native grabber
and then the user can call methods onto it (at iPhoneVideoGrabber* level, which is Obj-C):
// focus
-(bool)setContinuousAutoFocus;
-(bool)focusOnce;
-(bool)lockFocus; //at curent focus
-(bool)touchFocusAt:(CGPoint)focusPoint; // (0,0) bottom left, (1,1) top right, calls focusOnce
// Exposure
-(bool)setContinuousAutoExposure;
-(bool)lockExposure;
-(bool)autoExposureOnce;
-(bool)touchExposeAt:(CGPoint)autoExposeHere; // (0,0) bottom left, (1,1) top right, calls autoExposreOnce
// Torch
-(float)currentTorchLevel; // [0..1]
-(bool)startTorch;
-(bool)setAutoTorch; //adjust torch intensity depending on scene darkness
-(bool)stopTorch;
//white balance
-(bool)setContinuousAutoWhiteBalance;
-(bool)lockWhiteBalance;
-(bool)autoWhiteBalanceOnce;
But maybe they should all be exposed at ofiPhoneVideoGrabber level… I’m just not a big fan of unnecessary wrapping. What do you think?
You can see what’s been done so far here: https://github.com/armadillu/openFrameworks/commit/7c6d4e28802a9bbfd47ed27651e7df10c1b78a1f
I also extended the iosVideoGrabber example to show all the features:
