This is probably down to Pd (sometimes) reusing signal vectors for input and output, so clearing output also cleared input, as they are (sometimes) the same memory location.
Claude On 12 July 2026 21:15:48 CEST, Alexandre Torres Porres <[email protected]> wrote: >I guess I fixed it, sorry for the noise, it was working quite well at first >and help files still worked, but apparently having both > > dsp_add_zero(sp[1]->s_vec, x->x_n); > >and another perform for the input was a bad idea > >sorry for the noise > >Em dom., 12 de jul. de 2026 às 02:28, Alexandre Torres Porres < >[email protected]> escreveu: > >> geez, now I can't get it to work in any case, even though the help file >> still works, so it's something more obscure, my patch is basically empty >> and just with a read/write object... >> >> Em dom., 12 de jul. de 2026 às 02:15, Alexandre Torres Porres < >> [email protected]> escreveu: >> >>> Hi there, I've never seen a pd object that can't handle two connected >>> signal cords and simply sum them together, but my [del.in~] object is >>> doing that for some reason. >>> >>> see >>> https://github.com/porres/pd-else/blob/master/Source/Audio/del.in~.c >>> >>> it's quite similar to delwrite~ and I'm clueless why this could be >>> happening. I need to use [+~] otherwise the delay line gets silence. >>> >>> see also >>> >>> https://github.com/porres/pd-else/blob/master/Source/Audio/del.out~.c >>> >>> and https://github.com/porres/pd-else/blob/master/Source/Shared/del.h >>> >>> help anyone? >>> >>> thanks in advance >>> >>> Alex >>> >>> >>> https://mathr.co.uk --- [email protected] - the Pd developers' mailinglist https://lists.iem.at/hyperkitty/list/[email protected]/message/SC2WFOD7HQJJGVJGZ4CQZ6F4XY5RLKH3/
