I'm developing a Max external based on C++ code, in which I need to define the block size in a reliable way. The argument sampleframes (in perform64 routine) allows that, but it means transmitting the block size for every frame. On the other hand, the argument maxvectorsize (in dsp64 method), which is fired only once each time the DSP chain is being reconfigured/switched on, seem to provide the same information.
So my question: is there a case for which sampleframes (in perform64 routine) differs from maxvectorsize (in dsp64 method)? If not, is there a known drawback in using the latter instead of the former?
All the best,