Hi list.

If I want to delay a signal by 448 samples, I can either use a delread~-delwrite~ pair in connected subpatches (in order to force correct scheduling), or I can [inlet~] it into a subpatch with [block~ 512] and [outlet~] it back out to the parent (whose block size is 64). I haven't done the test yet, but I'm curious if anyone has -- which of these uses less CPU in general? Is reblocking expensive?

I know that using blocking for delay is not a good general approach, but in this case let's say I know that block delay would suit my needs exactly.