[PD] pd~ fifo latency

2014-01-27 Thread puredata
hi all, in [pd~] documentation: -fifo n sets round-trip delay in blocks when trying to use -fifo 0 or -fifo 1 (i want the lowest delay possible) = pd hangs. when not specifying -fifo i get 256 samples of delay (adding 5ms to my 13ms round-trip latency). how to have just 1 sample delay

Re: [PD] pd~ fifo latency

2014-01-27 Thread puredata
Hi Miller, Linux with 0.43.4-extended (ppa): Using -fifo 2 did the trick (1 block delay). Cheers ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list

Re: [PD] pd~ fifo latency

2014-01-27 Thread Miller Puckette
I just tried this (on Linux) -- I get a delay of 64 samples with -fifo 1 (as I was hoping for), but Pd hangs if I start a pd~ subprocess with -fifo 0. (I think it's reasonable to enforce a minimum fifo of 1 block since 0 block delay woud make pd~ less efficient than simply putting everything in

Re: [PD] pd~ fifo latency

2014-01-27 Thread Miller Puckette
Cool... I did some tuning on pd~ for 0.45 which might explain the -fifo discrepancy - so once you're up to 0.45 you'll want to adjust the fifo size down one block to get the same result. cheers M On Mon, Jan 27, 2014 at 08:49:34PM -0500, pured...@11h11.com wrote: Hi Miller, Linux with