hrmmm…
I’ve mixed feelings about this track.
the link for tess remixes goes to soft spectrum !
apparently I am terrible at copy and pasting. >_<
fix’d now
sci - 2 by bored
from around noon on the 10th of april, 2007. made in one hour.
all sounds were taken from image files.
I am not sure why this isn’t on sketches from math class, really. I’m quite fond of it.
it is now 4:40am and I am legit going to sleep. I only saw this because I compulsively check tumblr before going to sleep, like any normal person.
also I just sent in the minor-fix’d final version of my song for R&R. on that thought, can you like judge it super quickly and just tell me “yeah this is not what we want on the album” so I can move on? the suspense is killing me >_<
also oh hey look at that. 111 and 11. neat.
ah, the problem here is the header.
you should save your sounds as *.RAW
but…it’s also not quite that easy. you have to take a few things into consideration.
how to start: for everything I’ve done, I played a tone at the note of A (~440 Hz, or some multiple) and exported the sound as a WAV at 44.1 kHz (which gives the resulting image a bit of a tilt, but whatever)
this makes one period about 401 pixels wide, so set 401 (or some multiple. if you see “scanlines”, change the width. 802 seems to work for me in most cases.) as the width for image import, and pick a random number for the height.
important: if you export the sound as an 8 bit WAV, open the file as an image with a bit depth of 8 bits per pixel. (8 and 24 are the best options to work with. I use irfanview for this, no idea what import options other programs have)
of course, you could ignore that last bit, but you’ll get a very noisy image.
so what are you looking at? (assuming 8bit sound opened as an 8bit image), black means low point, white means high point. a saw wave would look like a perfect gradient going from left to right. a square wave would look like a straight line division: black on the left, white on the right, and no grey.
for a 24bit image, I haven’t quite figured that out. each channel seems to be something different.
ok I made some changes, now what? first, you should note this:
those first few, not-gradient-looking pixels? that’s the header. if you keep these intact, it makes things just a little bit easier. nothing required though
what you need to do: save the resulting image as RAW data if you can.
if not, a BMP will suffice. (just change the extension. it’s basically the same thing, just backwards.)
open up your modified RAW file, chop off the first 0.1 seconds or so (if needed), and you should be done.
hopefully this was helpful :)
reblog
1. for awake-at-a-decent-hour people
2. because this probably fits pretty well on my main blog too