Home Music Articles Forums Blog Chat More...      

Images


by

SmokeyVW

 Genre: Experimental

Get Flash to see this player.

Description
album: Waldo
track 1: 4:58 Conflict Of Interest
track 2: 5:56 To The Feynman Point
track 3: 4:27 Climax Pop
track 4: 5:02 Banished From Pandora
track 5: 4:20 Quellek
track 6: 5:28 FungibleSonicConsciousness
track 7: 2:33 Extraordinarily Corrosive
track 8: 3:07 253
track 9: 7:31 Images
track 10: 7:35 Disturbed
track 11: 4:10 CFL
...more albums...

Song Description

birds on wires
tree silhouette
bubbles
chain link fence
rings
bridge
tree with roots



Audacity -- Spectrum
Leave a Comment
You must be registered and logged-in to comment.
Song Stats
Hits: 2302
Comments: 11
Fans: 8
Plays: 133
Downloads: 76

Uploaded: Aug 20, 2010 - 11:06:03 PM
Last Updated: Mar 11, 2012 - 05:12:08 PM Last Played: Aug 11, 2017 - 09:44:29 AM
Song License
Creative Commons License:
Attribution-NoDerivs

Creative Commons

Song Actions
share on facebook share on twitter download song
Comments
bud said 2668 days ago (August 21st, 2010)
This is cool
This is very cool
Check out my latest song called Oh Coast
particledots said 2668 days ago (August 22nd, 2010)
.
i agree
Check out my latest song called Ghosts of Spur Cross
jiguma said 2667 days ago (August 22nd, 2010)
Wow!
Bill, the video is just amazing. Would love to hear what it is you've done here (although I suspect it might be a little beyond my understanding.
Check out my latest song called We End Up Being All The Same (42)
SmokeyVW said 2667 days ago (August 22nd, 2010)
.
i wrote a little program that reads in an image and interprets it as a spectrogram. it's sort of the opposite of what Audacity is shown doing in the video

basically the picture is cut up into rows of dots. the bottom row is scanned from left to right to control a sine wave synthesizer (the brightness of each dot is the loudness). the frequency for that row/synth is very low (like 50Hz or so). then the next row is scanned to control a synth with a slightly higher frequency. the process continues on each row of dots, slowly rising in pitch. the top of the image is the highest pitch.

(actually the program does this by using an inverse Fast Fourier Transform on vertical slices of the image, but the effect is the same)

i converted and listened to several hundred images found off the internet and picked the most interesting ones. along the way i started to be able to see an image and guess what it might sound like - so i directed what images i searched for later in the process.

then i mixed and cross-faded them in GarageBand. the last step was to record playing it on Audacity with the spectrum display...
Check out my latest song called Tru Th
SmokeyVW said 2667 days ago (August 22nd, 2010)
.
i should have also mentioned the two events that were the inspiration for this project:

someone at MacJams posted about a Justin Beiber song played 800% slower and published to the web as an Ambient track... it got me thinking again about sound processing algorithms

i saw an episode of the Nova tv show on PBS that showed a blind woman riding a taxi in London and aiming a device out the window - the device converted the images into sounds and the woman has learned to interpret the sounds as images - they drove past the Millenium Wheel, and she something like - it's large and round with little knobby things all around it.
Check out my latest song called Tru Th
mike_d said 2667 days ago (August 22nd, 2010)
Yea-
that is very neat! I thought the "bubbles" sounded the creepiest of them all. The bridge and tree with roots were tied for second most interesting.

Two questions:
Can you scan the pictures at different speeds?
Can you assign different "voices"?

I wonder what a sound wave from a popular song would sound like when scanned like this, like a Beatles tune or something.

Very neat stuff Bill!
Check out my latest song called Bleed out (redo)
SmokeyVW said 2667 days ago (August 22nd, 2010)
.
well, no, the program is very basic. you can rescale the photo first (and rotate them like i did on a couple of photos here). shrink it horizontally to make it go faster - stretch it vertically to add more high frequency sounds. it might be fun to do an entire piece on the photographic side: cut and paste images to form a single big picture that becomes the song.

i'm willing to give away the program as is - it runs on Intel iMac. but it's a little finicky to use - haven't packaged it up properly. let me know.

the program only does sine waves. that's a result of the iFFT algorithm. i'm wondering about how to change the program to avoid artifacts (like low frequency warbling and a weird high frequency oscillation). i'm missing something fundamental. there seems to be something like DC bias messing it up. it's also a mystery to me right now about why it seems to do "edge detection". large areas of a single color disappear and just have an outline. i dunno why. i'd also like to create cleaner sound files if possible.

oh, and it's only mono right now...

to create "voices" you could manipulate the images first - do things like putting parallel horizontal white lines across your picture. or apply blurring. blending copies of the image upwards (like visual echoes) would form a series of harmonics on the resulting sound.

you can use Audacity (it's free) to view songs such as Beatles tunes as spectrograms.

i liked the bridge the most myself

thanks for your interest
Check out my latest song called Tru Th
perceptualvortex said 2667 days ago (August 23rd, 2010)
Interesting
The bubbles sound great... like haunted souls. I also think the chain link fence is really good. Still listening... The rings are great too. I'd like to hear the fence and rings at 2x and 10x speed, both with and without the pitch change. Interesting work. My favorite is the bubbles; It's a rich, creepy sound.
Check out my latest song called Mercury (RPM 6)
DWL said 2667 days ago (August 23rd, 2010)
Just plain weird
Very clever and very impressive though!

Cheers

Dick
Check out my latest song called This Empire's State
paul f. page said 2664 days ago (August 25th, 2010)
Freaky stuff...
I didn't know what to expect when I hit PLAY, so this was a complete surprise. Reading about your process is very interesting. There are some seriously strange things going on here that are just "crazy unique." To me, this is quite a cool composition that tempts the listener to think about what it all means. Of course, it probably means nothing at all, but you keep me wondering. (Um, not a track for a headset. I feel like I have scorpions crawling through my brain...and that is, I believe, a high compliment.)
Check out my latest song called 185
sschedra said 2663 days ago (August 27th, 2010)
Bill,
Great innovation. I have played around with the Sound generators at Wolfram Tones (http://tones.wolfram.com/). I recently used it as the starting point of a piece I submitted to the 60X60 competition.

I love various ideas for synthesizing music. This is a neat experiment. I wonder if we could produce a sonic representation of a picture that would be experienced in a similar way as viewing a picture. The concept here would be that as you "view" the sonic image, the viewer could use the volume to introduce more material and the panning would allow focusing on specific detail. Less luminosity would mean darker themed sounds, Cooler colors, cooler sounds, etc . And the sonic image would last only a few seconds so your would view it in a loop to derive the different elements.
Check out my latest song called Cheap Tricks (LSP2)
Artist Info
artist photo
Name: Bill Grundmann
Location:
send a note
[see more] click here to check out my albums click on the blue table headings below (e.g. Last Comment) to sort the song list ... [see more]

Check out some of my other songs:

Genre Info
Experimental music is any music that challenges the commonly accepted notions of what music is. There is an overlap with avant-garde music. John Cage was a pioneer in experimental music and defined and gave credibility to the form. As with other edge

More songs in Experimental:

Chat
Who's Online
Invisible Members: 0
Visitors: 101