Apple

Hi. After a three year break from CoreAudio, I am back. I am working on an iOS educational app for kids that will allow them to hook the microphone up to a number of different filters or processors. From what I see of AVAudioEngine, I am not sure if the different nodes can be connected in parallel (I don't mean exactly parallel), and that data seems to need to flow through things serially. Can I do this:

https://www.dropbox.com/s/hanpcy7dtdzbfcn/PastedGraphic-2.png?dl=0

Where the microphone sound data is connected to three or more units?

If not, Can I use the tap on a node to create my own dispatcher of the data, calling the different clients with data, as the data arrives? Am I better off just going back to AudioUnits?

My next question is, can I have multiple unrelated paths in one AVAudioEngine, or do I need to create multiple AudioEngines:

https://www.dropbox.com/s/v1hzql57omicrnq/PastedGraphic-4.png?dl=0

I'll be building a lot of simple generators and filters, like sine, square, sawtooth, lowpass, pitch shift, time shift, etc. Also a "backwards" unit. Except for a few, all others I would like to have work in realtime. Again, shall I stay with the old CoreAudio API or will the new AVAudio framework give me the flexibility I need? I think I can recall most of what I had learned before with CoreAudio, and do remember that things weren't easy. That's why I am giving the new framework a chance, if it'll make my life easier.

Thank you in advance for your help.

-mahboud