Info for Making Music via databending for a project


right so i decided to host a databending competition…why because free time and there are some users interested in this…
there will be no prize this is just for fun…
and who knows you might learn something new and add a new dimension to your music making right…so on with it

how to databend:
forgive me im a noob at this maybe someone has better info on how to implement databending into art but as for me honestly I just randomly cut and paste/copypasta some hexadecimal/ascii code in a text editor or hexadecimal editor…messing with the sequence of the code…or change some of the hexadecimal values (i dont actually know hexadecimal i just put in something randomly to change the code)
and then I use the import raw data feature in audacity or screenshot an image and resample it with harmor

also this:

how to capture the databent image:
after databending and image screenshot the result…then .convert image to bitmap using ms paint or whatever, import raw data into audacity encoding u-law, big-endian to get something that isnt white noise…feel free to play around with the file format or the encoding to get a variety of sound

how to capture databent audio:
i’d say record the output of your computer’s soundcard using whatever…

a more in depth article on how to databend to get people started off

also how to convert image files…the quick way ms paint…as pointed out in this article…

different file types will yield different results when converting image to audio or audio to image

you can either use an image source or an audio source to databend and resample it

so to start off with
here is a source image that you can use…

and an audio source to databend (i enabled downloads):

and a free hexadecimal editor i found

and a free online photoshop site

what is hexadecimal:

if you decide to databend an image:

  1. post both the original source image(you can use any image) and the resulting databent image.

  2. convert the databent image into audio… and post the resulting audio… keep in mind when converting image to audio that different algorithms will yield different results so experiment and go with the one you like the best when converting… (if you want you can use what i provided to start off with but I highly encourage those that want to do this to make their own databent image)…

  3. resample the audio using whatever software you have at your disposal…( you can also use photoshop prior to resampling to make the image sound more interesting…)

  4. incorporate your captured databent image, having now turned it into sound and resample it and turn it into a full cohesive track

if you start off databending audio

  1. post the original sound, and then the resulting databent sound. get the sound by either importing raw data into audacity or by recording your computers output using whatever means.

  2. convert databent audio result into an image and post it…(the quick way to do this is to use the spectral viewing feature in your daw and screenshot it)

  3. (optional) databend the image or photoshop it or both

  4. convert image back to audio, fsu resample it

  5. incorporate the resampled sound in a cohesive track.

you can also databend both audio and an image.
if you decided to do both just follow both sets of rules for Images and Audio…

you can also use synths and other samples but it has to complement the databent sound you created and resampled.

And you could databend other file types and convert into sound to create more samples that you can use to create a track…

all submissions will be decided via poll in a seperate thread…DEADLINE MAY 15th

cheers have fun guys :smiley: looking forward to seeing what you come up with databending stuff


I’m into this, I’ve actually been pissing about with this stuff since you reminded me of it.
Is this more of an open ended see what people come up with sort of thing, or is there a deadline?


Deadline will be decided among those that decide to participate


BTW, images will be displayed differently depending what you’re viewing it in after it’s been databended. It’ll look completely different in Window’s default image viewer compared to Photoshop for example.


another challenge centered around resampling

yo count me in


Of course I’m in


Thank you for sharing info about databending. I was previously unacknowledged with this possibility, im glad my paintings/graphics have a further use.
Count my participation, good luck to anyone.


this is the format for your entry into the competition if you decide to both databend audio and an image

also this is the piece of music that incorporated my resampling of that sound created from databending

^not a submission…not participating cause im hosting this contest…it would be a wierd thing to be apart the competition that i created…im treating this competition like an unofficial beat battle


if you wanna chit chat about the challenge there is now a discord channel…in the idmf discord.

also either post your submissions in this thread or dm your submission in discord its just easier for me organization wise


updated the rules…for clarity.


what is a good deadline for everyone???

a few weeks??

how does a few weeks after the submission deadline for idmf’s idm comp sound?

may 15th good ?


I have the databent sound from databent picture. It will take a week for me to incorporate it in a track somewhere cohesionally.

@bfk Have you tried databending other sources into sound? Random extensions like system ddl files, shortcuts… It works as with pictures, some samples can be made into drums, other into synths.

May 15th sounds ok. Some people are working on idmf comp, so if deadline is sooner there could be more stress.


I haven’t done data bending of other file types in a while…I totally forgot about it…tbh

also ppl can also use other file types to databend in addition to data bending an image and/or audio…to create more samples that they can use for this competition.


I’ve got some very interesting results with blender files too, some percussive and tonal sounds


found a tut for converting image files to sound, fsuing the sound, and converting the sound back to an image


Weirdly I keep getting corrupted files with this method unless I do very specific effects. I’m avoiding the header, but it seems like using a few different effects breaks the entire image no matter what I do. At one point paulstretching even worked, but if I added anything else into the mix it all fell apart. I only got the very bottom of it to go, and that alone wasn’t really much of a glitch at all.

Same shit happened in notepad, I avoided the top half and went to town. No dice on reopening :frowning:

I guess I have no choice but to use the site that glitches things for you


Somehow a quick paulstretch worked after like 5 attempts :smiley:


are you using vcv rack??? pretty cool…interesting sound :slight_smile:


Yeah, I have it bridging into Live (off screen). I found that module and got inspired for this :smiley:


I am getting in on this, sounds fun!