hi brett, that track is still up on my site. for some reason the link comes out as a .pd file not an .mp3

http://www.m-pi.com/this-is-serious-mum.mp3

just cut and paste that and it will work.

also heaps of stuff here: http://www.m-pi.com/remixes

>It's weird that many ppl seem to be using pd but that the output~ page in the forum still has threads in it from 2004 in the top page!! <

it took me a few months of solid patching (a few hours every day) to get a workable setup for actually making tracks. it's certainly no small undertaking.

>I'm pretty new to pd and just working my way through tutorials at the moment, but do you have any tips with regard to actually going about customising your own setup?<

you are on the right track going through the tutorials. the way i did it was first to build stuff to cut up and effect samples, and then secondly make a system to control those processes live. mine was all based on the [key] command, and i just triggered everythign from my laptop's qwerty keyboard. this was nice when i was travellign as it meant i didn't need to cart any gear around. also good for playing live cos i could pick my computer up and jam on the dancefloor. there are a few options though, especially triggering stuff with sensors and such. but i'm sticking with the bare bones keyboard approach cos it works for me well enough.

> like whether to keep lots of separate instruments or try to keep everything under one roof...<

i try to keep my stuff in one patch as much as possible. a couple of reasons for that, but the main one for me was that i kept modifying abstractions and then other patches that relied on those abstractions would stop working. generally much easier just to have one or two or a few patches to do everythign you need. even if you incorporate everything you make into one patch it doesn't get too big. usually well under 1 meg.

>I think I will tend to mainly use samplers and control structures for controlling my external Midi gear, but in a live setup, not sure how to integrate it into Logic Pro?<

my thinking on this is that if you have a guitar it has 4 or 5 strings, and you manipulate those strings in a variety of ways to make most of the sounds you need. if you listen to my audio..all of that is just 2 or at most 3 channels! so i always have only 2 or 3 samples playing at once. my stuff from back then was a bit light..not really hard hitting on a dancefloor (which is what i'm interested in) ..but i think you do what to keep everythign as minimal as possible. as far as live performance goes, i wouldn't go anywhere near something like logic audio.

if you have midi gear, then def work on triggering that with pd. i'm working on synthesis within pd now, rather than the sample based stuff...but it's a constant battle to keep cpu usage to a minimum. triggering external devices will be no problem for pd and will leave you heaps of cpu for doing sample mashing.

can't stress enough though. KEEP IT AS SIMPLE AS POSSIBLE. for live music, traditional musicians only play one instrument at once. if you want to make whole songs live, then you are going to have to do the beats and bass and interesting stuff all at one time, so you want to keep it as simple as possible so that you can inject a lot of liveness into it. generally, the more channels of audio you have going at once, the less room there is for jamming out in an impromptu fashion....unless you have magic fingers.

>Look forward to hearing your stuff if possible.<

cool, thanks. quick background on my stuff..."this_is_serious_mum" is a live jam recorded in one take. just 2 channels of audio driving all the sounds from small sample loops being cut up in realtime by me pressing keys on the keyboard. it's a super simple setup, but i think the reason why it works ok is that i spent more time actually playing and practicing than i spent on coding the bastard. i toured across europe and japan and australia playing this stuff and it was generally well recieved. at really good gigs it was the biggest rush ever.

so yeah. good luck. grab the bull by the horns and just go for it.

Cheers,

matt

I have an Edirol FA-101, and I'm trying to get it to work on my PowerBook with Gentoo Linux installed. The issue is that sound just doesn't work for me.

I do this first:
jackd -d freebob

and I get
jackd 0.102.20
Copyright 2001-2005 [...]

JACK compiled with System V SHM support.
loading driver ..
Freebob using Firewire port 0, node -1
libiec61883 warning: Established connection on channel 0.
You may need to manually set the channel on the receiving node.
libiec61883 warning: Established connection on channel 1.
You may need to manually set the channel on the transmitting node.

This is normal and is what I get on my amd64 system when I use this setup. When I open Pd, however, with -jack as a command line option, the Media menu is empty and Pd doesn't respond to anything. Also, if I don't use -jack, I can go to the Media menu and select jack, but after that everything stops responding.

Also, if I start jackd with -R -P 70, I get the same output from jackd, except after a couple of seconds I see "Aborted" and jackd stops running:
... etc, etc ...
libiec61883 warning: Established connection on channel 1.
You may need to manually set the channel on the transmitting node.
Aborted

I'm using pd-0.39_p1 from the Gentoo portage tree, same as on my amd64 machine.

Thanks.

I think i'm understanding, thank you for your reply hardoff

my knowledge of PD is extremely limited as of this stage

so you're saying, as i understand it anyway, instead of having two random no. generators to randomise what channel im outputting to as well as the volume of the output, i just have one which controls all 5 volume outputs on the channels, then when one is selected, it mutes the others?

so like Polo! is piped through all channels, but PD tells the other four to be quiet after the number gen picks a channel

to add a bit of clarity....

You'd go in, say Marco
then, "Polo!" would be randomly played back to you in a channel at a random volume (for arguments sake lets just say out of 100)
e.g. Polo comes in at volume 63 in the center channel
then you sya Marco again
then Polo responds at volume 99 in the front right channel

Thanks everyone again for looking

[trigger] / [t] objects are necessary to give a defined order to the operations of a patch. For example:

in
| &#92;
[+ ]
|
out

At first glance this would double the input float, but without a trigger object the order of the passing of data from in to [+] is undefined, there is no way of telling whether it will be sent to the right or the left first. If it sends to the right first, then out = in + in. But if it sends to the left first, then out = in + the previous in.

To avoid this non-determinism, you need trigger objects. If you want the first case:

in
|
[t f f]
| |
[+]
|
out

Or if you want the second case:

in
|
[t f f]
&#92; /
X
/ &#92;
[+ 0]
|
out

[trigger] objects send their outputs in right to left order. In the second example, I initialised the [+] with a creation argument because the first time the object is triggered no value has been received at its right inlet.

Another use for [trigger] is to convert between types, useful in many situations, most often where you don't care what a message is and you just want a triggering bang.