Another forgotten site

i forgot i was uploading to another site....

all ship-shape here now

new track unveiled for the end of time

first mix and a psychedelic video....no name yet....

Raspberry Pi usb sound into pure data- the Griffin Imic

the griffin imic - get it here

After testing a few usb sound cards i settled on this one giving the best results for the bat-box project (and the set up described here is the one i used for this specific project)

 i used a 50 sec buffer size and the sound set at 2250hz. This seemed to work really well, i didn't really investigate getting the sound up to 44100hz,(i spent a day trying it with not much success) because 22.5kz is  stable, and good enough sound quality, obviously this is something to investigate more when not in the middle of getting a project going.

The imic is really stable when running, although there are a few random issues. 

the imic doesn't need setting through the terminal alsamixer each time when using pure-data extended, but the first time you use it (and if something goes wrong) do this…..

-open terminal

- and type in  (stuff between quotes, don't include quote marks when typing)

"alsamixer"

-f6 key lets you choose the card

-f5 lets you change the mic in , sound out ,treble and bass setting

press escape to leave the alsamixer

Pure-data extend generally flashes a red  warning as it boots up warning of a problem  with the audio,  "audio I/O error" -  ignore this as it doesn't stop or crash the audio.

the imic really likes the top usb-port, moving it from here can sometimes cause problems. If you're using more than two usb items (which you would do as you need a mouse and keyboard at least) invest in a usb powered hub -the best one i found was from the mod my pi website, remember as in all periphals working with the pi,  not all powered hubs are equal - some work better than others). 

The imic can go down if you're using too much power, for instance when using with an arduino, but once the program is optimised it will run with no problem each time. (-run without a gui , take bits out of the program to make it more efficient).A good indicator is the green cpu monitor in the corner of the pi, you want to leave a bit of a gap at the top so its not totally green, when it goes totally green thats when the sound card is in danger of going down.

Also with the imic you get one shot per boot, if it goes down it wont work again, if you close pure-data extended and don't reboot the next time you open pd-extended the imic wont work, so each time you close pd-extended you have to reboot in order to use it again.

Sometimes after p-d extended is started but before the pure-data patch is opened the imic decides to not work, the easy fix solution is just to restart the pi, and generally it works. If it carries on, re open alsamixer and reset the sound card…-this might help….

...... also turning the pi off for a day then coming back to it has weirdly worked a few times.

Getting pd to work well on a raspberry pi 2

A few tips to get pd working well....

- open terminal

- then type (stuff between quotes, don't include quote marks when typing)

"sudo nano /etc/security/limits.conf"

-and then at the bottom of the document add (type in)

"* - rtprio 99

 * - memlock 1000000000"

-then save.


The sound settings here work realy well with the Griffin Imic

-start pd
-on title bar go to media > preference > startup
-add (type in) the following flag in the startup flag field 

"-rt -alsa -audiobuf 50"

- the bring up sound settings and change

"audio buffer 50ms

sampling 22050kz"

Miller Puckette tells it like it is


Miller Puckette - Intro to Raspberry Pi from Alexander Matthews on Vimeo.

Miller Puckette - Raspberry Pi and the SD Card Storage for Pure Data from Alexander Matthews on Vimeo.

Miller Puckette - Pure Data on the Raspberry Pi from Alexander Matthews on Vimeo.

Refining the Raspberry Pi - batteries

the batteries used in the bat box are these...

Anker Astro 5600mAh

They are pretty good, they can power an arduino, sound card and a raspberry pi (overclocked to turbo) for over four hours....

Raspberry Pi - Vocal Migrations

Working with the artist Kathy Hinde, I have been developing the software for the bat box element of her project Vocal Migration. The bat-box is a stand-alone vocal sampler run on a raspberry pi, using a ultra-sonic sensor to modulate the sample. If you want to find out more about the project check here. 

The box is made up of a raspberry pi, run from a battery with a arduino board giving controls from knobs and buttons, and also modulating the sound output from a ultra-sonic sensor. Sound in is provided by a sound card, and a powered vocal mic and sound out is via a small powers speaker.

Running PD Extended On A Raspberry Pi

very excited by this.....

I got the information on how to do it from (mainly) theses two sites

http://arduinosynth.blogspot.co.uk/2012/07/compiling-pd-extended-on-armhf-armel.html

http://log.liminastudio.com/writing/tutorials/how-to-build-pd-extended-on-the-raspberry-pi

plus information gleaned from here

http://www.mail-archive.com/pd-list@iem.at/msg52749.html

Rather than complie myself, i used a package already prepped by Mr Audio Synth. Tis PD extended 0.43.1

avalible here

https://docs.google.com/file/d/0B_UT435wUJeTVTh3VEFIODNrOU0/edit#

also the version of raspbian wheezy i used is the latest available here

http://downloads.raspberrypi.org/images/raspbian/2012-08-16-wheezy-raspbian/2012-08-16-wheezy-raspbian.zip


ok...you got everything, ?

>>>>>>>>>>>>>>>open LXTerminal, then type in

sudo nano /etc/apt/sources.list

>>>>>>>>>>>>>>> select all, delete, then paste this in

deb http://mirrordirector.raspbian.org/raspbian/ wheezy main contrib non-free rpi
deb-src http://ftp.uk.debian.org/debian/ squeeze main
deb http://backports.debian.org/debian-backports squeeze-backports main
deb-src http://archive.raspbian.org/raspbian wheezy main contrib non-free rpi

>>>>>>>>>>>>>>>> then exit, remember to save (y , then return when name to save comes up )

sudo apt-get update

sudo apt-get install rsync autoconf libfftw3-dev liblua5.1-0-dev swig libvorbis-dev ladspa-sdk libspeex-dev libmp3lame-dev lua5.1 tcl-dev

sudo apt-get install libfftw3-dev

sudo apt-get -f install

sudo dpkg -i Pd-0.43.1-extended-20120723.deb

sudo apt-get -f install

>>>>>>>>>>hey hey its installed!!!!

now open pd

go to audio settings and change delay time to 100 ms

should be working cool….

Die or get rich tryin'

another lp done and dusted.

download it here

Nananananalord

New series of e.p's download the first here

Marks got a new lp out

find out more about it here

Slash Stroke Magazine

read the article here

benefit benefits 2

get it here

The Experiment

I've been putting together a new site just for program stuff here. No ones realy gone there yet, so I'm putting the stuff on here too....that'll show em.....

from 31 October 2011.....

The Experiment Live

""For one night only, Thrill Laboratory present an experiment like no other: real scientists, amateur paranormal investigators, and the live exploration of a haunted building. University researchers have developed cutting-edge medical technology to monitor the effect of paranormal beliefs on the bodies and minds of those susceptible. Dr Brendan Dare brings to you the unpredictable results of their scientific research from the remote basement of a reputed Broad Street haunting. The Experiment will be captured by a film crew and transmitted real-time to a live cinema audience."

website here

Project summary

I was asked produce a horror soundtrack that could be shaped by biological-data, The bio data was captured by electronically by measuring pulse,breath and sweat (GRS) responses and transmitted  wirelessly to a hub computer. The composition made up of individual music tracks which can be modified in various ways, creating a changing horror sound track reacting to the highs and lows of emotions felt by five people during a filmed paranormal investigation.

For this i created two interlocking compositions who's different parts that could be highlighted and modified. These composition can be played singularly or an overture can be created from different parts of each. 

For the composition I pulled out every horror cliche i could find. The general feel for the compositions over all was a pastiche ofJohn Carpenter horror sound tracks. The main piano theme itself is  a derivative of the Halloween soundtrack.

The tritone, also know the devil chords (diabolus in musica) is a well know technique for horror sound tracks,as well as apeggiating the notes up and down a semi tone. Low bass sounds, discordancy, creepy violins, sound effects such as thunder and creaking doors, as well as sound design using reverb, tape delays and stereo placement.

Each track was 46 seconds long, there were 16 tracks in total.

an example of the compositions functionality
n.b this recording does glitch!(due to the recording it off thing....)

ingredients .….the kit used

data stream from the p.c. hub

PD to understand the data from the hub and sending to  Max/MSP via Osc, 

Max/MSP run midi/control notes ,to receive bio data via an OSC link from PD 

Logic,running 16 synthesisers/samplers ,effected and sound designed

Purpose of Max/MSP

Max/MSP was used to control the midi data of the project,to play the composition.

The patch reads a 16 track midi file, each track of the midi file can be modified in  (for instance each track of the midi file has the option to be appegiated) "live" before being sent to logic, via inter application midi, to trigger logics powerful synths,drum machines and samplers. The volumes of the synths,the note it plays, its perimeters and effects can all be  modulated from Max/Msp.

These controls are modified by the bio data, or can be controlled in real time via a midi controller.

A separate patch (to be explained later) was also developed in Max/MSP

Purpose of PD

A program needed to be developed to communicate with the hub computer, receiving the biometric data. The hub computer communicates with the "music" computer by sending its data via TCP, either through ethernet or a wireless network. After experimenting with TCP protocols in  Max/Msp, I quickly discovered its wasn't ideal for the task  and used the programming environment PD (pure data),  The patch received the raw data, packaged it, and then sent it to a separated program i wrote in Max/MSP via OSC. I made this choice so a separate laptop could be used to receive the data, and then send it via midi to the "music" machine. A separate laptop sending out the data meant that a crash wouldn't nesicerly bring stop the music. This I felt would shore up a weak link, if the program crashed (and it didn't!) it wouldn't also take down the music program. The max patch also translated the data and turned it into a midi control value between the scale of 0 to 127, which made making the program bio reactive easier, and more adaptable for other data in the future.

Purposes of Logic

Logic was used as an audio engine. Logics on board sampler, the albino synth plus its various effects units where all put to good use. These all were controlled through midi sent by Max/Msp. One disadvantage of this system was the Tempo cannot be easily controlled via midi in Logic, so this was not used as an element in the composition, tempo needed to be controlled to make sure syncopated elements of the effect units were kept in time with the standard tempo.

Change Factor- how the composition can change

All tracks can be appegiated (in the case of the sample and drum machine different samples/patterns are chosen) and the base volume and wether the track is muted can be changed.

The two compositions were made up of the following elements

track 1- low drone-resonant filter can be modulated

track 2- bass note- filter cut off frequency can be modulated

track 3- low string- filter cut off frequency can be modulated

track 4 piano chords- can be raised a fifth

track 5- piano riff- can be raised a fifth

track 6- high string- filter cut off frequency can be modulated

track 7- drum machine- use hi hat sound or drum pattern

track 8- sampler- choice of 16 different spot effects (choose at random or repeated)

track 9 -drum machine2- start stop drum pattern

track 10- morrocan pipe- resonant filter can be modulated

track 11- glacial synth- filter cut off frequency can be modulated

track 12- discordant organ- filter cut off frequency can be modulated

track 13-ghostly voices-resonant filter can be modulated

track14- creepy music box-can be raised a fifth

track15-appegiator- filter cut off frequency can be modulated

track16-spare

The Total Randomness Index

with these perimeters we can roughly work out we have…….

 at least 2 different ways of modifying the sound of the 16 basic tracks

that gives us 512 different combinations of the basic tracks, each track is 46 seconds long

that gives us a playing time of around 6.5 hours

On top  ,each volume level that can be adapted, and blend smoothly in or out. or be mute so it cuts in and out abruptly.

additionally each individual track can be arppegiated which give us another at least 256 different ways the tracks can work together.

Although there maybe many different combinations the tracks will always retain a flavour of the original composition.

How it deals with the data

Due to the experimental nature of the performance and the fact the system hadn't been used live, I used a very simple algorithm to trigger the composition. Each participant was given a fear index, based upon the total readings of their bio data results. This index was ranged from 0 (for base level) to 3 for maximum stress displayed. All the participants fear index was added together to give an average result, the number of tracks that could be triggered (heard) was based on this result. As the fear index fluctuated around every 20 or so seconds this made an ideal trigger to change the playback order of the tracks and their volumes, giving a varied composition. the relative volumes between the track were altered with a volume slope between the two values giving a smooth transition.

Also the GRS index gate spikes,rising quickly and settling back down again, this made it perfect to open and close resonate filers and cutoff, giving a distinctive "womb" sound.

the GRS index in some case was also used to trigger sound effects,  when the level reached a certain level a random horror sample was played. Hopefuly the effect of this was to audibly highlight any spikes in the data

The compositions key change (arppegiated note) trajectory was in this case operated manually. The mechanics were there to connect it to a bio generative behaviour but in light of the performance it seemed beneficial to change the dynamic manually to build and disapate tension.

At the same time all changes could be operated manually,as previously mention the system hadn't been tested and manual control via midi enable me to make sure the compositions trajectory moved slowly.

Not included…...

With more observation of actual real life examples of biodata, it will be easier to identify spikes, trends and conditions in the data which will influence the mapping of more sound behaviours.

Tempo at this time is not included in the change perimeters due to logics inability to map this behaviour.

Mark Offenhand takes on Dvorak

Marks take on Humoresque.........

Mark Offenhand

Mark originaly came to life during Making Friends And Enemies.
He's gone large. Heres how he works.......

The Olden Hinde does the Rochdale ring

The Olden Hinde does the Rochdale ring

The Olden Hinde

This one i done last year with kathy....tried to tell you about it at the time....

find out more here

What year is it?

I have been away for many months, i have been living in other spaces....time to catch up......