I get the current blocksize from [bang~] [timer] and [samplerate].
(see attachment) But in upsampled patches the [samplerate~] outputs the "upsampled" samplerate, so my calculated blocksize is upsampling-factor * audio-samplerate.
hmm. this seems to be only the case when the logical time for bang~ happens to be below the duration for a 64 samples block (or perhaps below 1ms?). I found an old thread in the list from 2002 which seems to explain my problems:
http://lists.puredata.info/pipermail/pd-list/2002-04/006123.html
thanks anyway
cheers ; artin