I completely agree with 0.5piRC. If you had come here with some ideas or partly formed patches then we could have had a discussion or offered some pointers, but simply posting the task verbatim???
Don't forget that lecturers also frequent forums like this one...
I have been slowly delving into reaktor over the past year or so and I am finding a ton of great resources here. adsrsounds.com
Also, I am reading Designing Sound by Andy Farnell. It is using Pure Data, but a lot of the same principles apply.
There are some good basics on youtube if I remember correctly.
The big thing really is remembering there are two types of data, numbers and audio stream. If you want to use audio to say... change the frequency of a filter you have to put a converter between the two.
Other than that if you know the principles of cv and gate (cv being pitch, gate being on....
In many cases OSC is used to encapsulate MIDI messages into UDP packets so they can be sent over a network (sometimes to multiple destinations).
However, the OSC protocol allows many more types of data to be transmitted. For example, you can send 32 bit numbers (float or integer) compared to 8 bit integers in MIDI, and symbolic messages, too.
Reaktor macros are implemented graphically and not programmatically as in via text. You would need to identify some sort of macro block that allows the text transform you are suggesting then route the result to a parameter pin.
I don't have time at the moment to download the patch, but looking at the video I think the "simple" implementation is using convolution where different "monster" impulse responses are chosen. The advanced engine then just adds additional processing. A bit of research of convolution should point you in the right direction.
BTW you can also create some ...
The quality of the convolution can depend on a lot of factors. IRs are meant to sample a space over time. With something like a synth, the change of frequency content over time would probably be less important (compared to a reverb). I have found that chopping the length of the sample to a few milliseconds (or tens or hundreds of milliseconds) can result in ...