This page holds Sam Freeman's hackpact documentation
Any feedback via Twitter I guess, because there's no comments feature here…
I had arranged to do this site today before learning of hackpact.
Mike (Pempi) wanted help with a basic site and I had five hours of sitting on trains to do…
As it happens, I feel this is an appropriate start to the month because html was the first that language I taught myself c.a. 11 years ago.
The code-creates-stuff process had me hooked from the start → I like stuff and I like making stuff thus I like programming
Back in the day I was on Windows98 writing code in Notepad, after picking up the basics from some demo web development software.
Today on Mac OS X (version 10.4.11) I wrote in jEdit (version 4.2final (there must be a newer version out there as i've had this more than two years(maybe i'll follow that link now that i've made it…)), and used Safari (Version 4.0 (4530.17)) to preview the pages because Firefox hangs so often since its last update.
I did use Firefox (version 3.5.2) to refer to CSS reference and for other browsing because Safari on my machine doesn't know how to find the internet (the list of thing I should update on my machine would be very long if i wrote it out).
Fetch was used for ftp'ing
Today's project began in a bar near Leicester train port at lunch time.
Over a drink and a pizza, Mike showed me the sites of a few similar artists, and the html that he had hacked together, and we discussed the design specification for http://pempi.info
Then between LEI and DBY, DBY to LDS, and LDS to HUD (plus some at home) I wrote the css and html, made a gif (which i'm not really happy with) and hacked and old img for a background which wasn't called for, but, you know…
Finally I wrote this documentation and next, to end the day, I will link to this from the hackpact home at http://toplap.org/index.php/Hackpact
it's not finished but it is made → a new version of 'GetBuffers'
max patch online here
The basic idea is that a circular buffer is recording input audio and sections can be grabbed by the user for use elsewhere in MSP.
the theoretical problem is how to grab a section that starts near the end of the buffer and stops somewhere past the beginning of the buffer…
my solution is to have two circular buffers of equal duration recording the same input but to have one of the buffers offset in time by 50%
in this way the data at the end of one buffer occurs in the middle of the other.
the patch that I build during my MA (c.a two years ago) was based on [record~] and had [waveform~] objects for overview and closerview UI.
in this new version [poke~] is used to record the input and a Jitter solution is prototyped for the overview.
Today I made this page having completed and compiled sgii2p011.app
the page speaks for itself in documenting the software, this is the brief documentation of that documentation.
I had hoped to actually use said software to capture some live coded sounds, but the day just didn't go that way, so instead I present that there page on the i i wiki as the what I made today.
This whole hackpact thing is, to me, just an exercise in getting things done, a daily deadline. So far, I've posted three just before midnight and this one the next morning - despite having done the screen-cast at 8pm…
today I had an idea for a 5min Max patch; a plan which I followed. It didn't quite take 5 min, but rather than drag it out, I implemented a fade out.
Also today I have had an idea for a new type of screen-casting for “live code” ← quotation marks inserted there because the result will not be a full real-time documentation of a performance → the output will not be a live code set, but it will be the product of a live coding session…
The performer will be able to start and stop the recording such that the finished video will consist of some selected bits while skipping out other periods of time (perhaps where something complex is being set up).
Anyone who grew up with a tape cassette recorder knows of this type of linear editing - you cannot go back to change previous transitions, but you can leave it on pause for as long as you wish before continuing…
The next thought, then, is that such start/stop edits are basically rubbish, to which I propose a method for preemptive cross-fades:
_• whilst recording, the player decides how long the cross fade from the current passage to the next will be. [Let's say 30 sec for this example].
_• the player marks the point in time where the cross fade will begin, at which the system begins to buffer the screen and audio data for the next [30 sec].
_• recording is then on pause.
_• when they are ready to record the passage, the buffered data streams will be cross faded with the live data.
_• repeat to end.
What do you think??
I can see (in my mind) how to code this up with Jitter, but will not have time for that this month…
The only thing that I made on day 6 was a list of photos.
The day was taken up diving to the Honister Pass, Cumbria, climbing the 'zip wire' via ferrata route with my eldest brother and Dad (and nine other people)
At various points along the route the guide type dude took photographs of us, the punters, which were offered on disc for sale at the end of the climb for an extortionate fee.
So, the list that I made that night is of the 30 out of 71 photos that were 'applicable' to us three out of the twelve people in the group.
… and of those, only handful have I uploaded :
monthly inclusive improv session
live coded this towards the end.
(.aif write of buffer final contents included in zip ← mainly featuring cello by Scott McLaughlin)
Also played my 2007 laptop junkie system, and a version of QWERTYpartials.
and a plan for a monome based experimental performance was also hatched . . .
algorithmic composition using first twenty-five prime numbers, modulo, and midi.
groupings of modulo outputs summed are as they are so as to ensure that the pitch and velocity values remain within midi range
→ with retrospect, it might have been a good idea to attach a [closebang] object to [flush]
was thinking about the monome thing
about messages to it and from it
the format is same as [matrixcontrol]
let's take an on state for a walk…
thinking about a monome from Samuel Freeman on Vimeo.
no sound by the way... video is as much about testing the selection feature of sgii2.n and compression settings as it is about showing the patch...
the [r-incdec] sub patch (which aught to be an abstraction) was made a few months ago, but i did add the int input management today
Pempi video draft edit, various soups, sponge cakes, some photography, rearranged most of the furniture down stars.
Booked to attend seeing sound
finished Pempi video
but only managed to get a copy with over compressed audio online → internet connection at home sucks
will have to wait til I'm next in uni