ErikNatanael
commented
1 year ago Oversampling:
A GraphGen with oversampling will essentially run with a sample rate and block size multiplied by the oversampling factor. How will we handle one oversampling Graph within another oversampling Graph? The intuitive way seems to me to be to treat them as naturally compatible i.e. they don't need conversion. Another way would be to say that the inner Graph is oversampled compared to the parent, and if they should be compatible the inner Graph should have the extrapolated settings of the parent, i.e. sample rate oversampling and block size oversampling.
If a graph is within another graph and has a different sample rate and/or block size it should be converted to the parent's sample rate and block size. This will allow for small parts requiring sample-by-sample processing within a larger graph with a larger block size, and oversampling an entire or parts of a project.