Freesound Forums
Dare 26 - unusual sound creation techniques
Started March 8th, 2014 · 135 replies · Latest reply by minagarmehei 9 years, 2 months ago
Dear friends
This dare is specifically aimed at the Collaborators Collective (but not restricted, anyone can join).
The net is full of tutorials about subtractive synths - the typical analog synth.
The topic is covered in any book about sound synthesis and any half decent analog synth (hardware or virtual) will give at least a basic explanation in the manual.
But there are many, many, many more ways to generate sounds. We explore a couple of really unnusual ones in this dare... should you dare to follow me into this uncharted (but fun) territory.
The idea here is to explore sound synthesis - as a group. Sharing ideas and results. The focus of this dare will really be the discussion thread, where all those ideas should be shared and discussed. Here:
http://www.freesound.org/forum/dare-the-community/34720/
In this here thread I invite you to post links to your works.
- If you create sounds unsing the techniques below, please upload them to Freesound and share a link here. Please tag your sounds with the keyword "dare-26"
- If you create a track using these sounds, please upload elsewhere (Freesound does not accept music) and post a link on this thread.
The Techniques:
1) Databending. Check this thread and the link there to an article by Glitchedtones.
Tells you everything you need to know.
http://www.freesound.org/forum/articles/34698/?page=1#post72999
2) Somewhat a variation of the above, and one of my personal favourites: image-to-sound.
You can read about it here:
http://www.freesound.org/forum/off-topic/5192/
A few programs suggested by Bram himself on that thread:
http://www.nicolasfournel.com/audiopaint.htm
http://hem.passagen.se/rasmuse/Coagula.htm
http://sourceforge.net/projects/opensonify/
(audiopaint is very good and I use it frequently 
)
You can also try this VSTi:
http://www.kvraudio.com/developer_challenge_2009.php
EDIT: Please note the plugin I mean is Harsh Digital Nose (by insertpizhere). There is a lot of other stuff in this link.
and this new standalone program just in:
http://www.kvraudio.com/product/paint2sound-by-docnashsynths
My logo has been run through AudioPaint 3.0 (nice one, AlienXXX )
http://www.freesound.org/people/copyc4t/sounds/222065
and is now available after moderation
Hi copyc4at
Please be sure to try the logo also as white on black - usually produces more interesting sounds
(sounds with a lot of white tend to sound 'noisy')
While you are at it, also try the logo (white on black) rotated by 90 degrees.
Too tired tonight to upload any sounds myself, but have a few ideas bubbling...
@AlienXXX
You're right, it's just that I liked the cat-aclysmic sound of the white background 
Anyway I've just uploaded 6 new sounds with the white on black logo (normal, rotated left and right) run through AudioPaint and Virtual ANS:
http://www.freesound.org/people/copyc4t/sounds/222565/
http://www.freesound.org/people/copyc4t/sounds/222564/
http://www.freesound.org/people/copyc4t/sounds/222563/
http://www.freesound.org/people/copyc4t/sounds/222562/
http://www.freesound.org/people/copyc4t/sounds/222561/
http://www.freesound.org/people/copyc4t/sounds/222560/
On a side note, Audio Paint works nicely on Linux via Wine (as do other aforementioned tools), Virtual ANS has a native Linux version.
DFX Buffer Override as Raw Data:
Buffer Override VST by DFX opened as Raw Data in Audacity. Removed some bits of white noise from the beginning and end of the file. No processing.
Hi everyone
I have so far just played around with sounds others have submitted
http://www.freesound.org/people/AlienXXX/sounds/223008/
http://www.freesound.org/people/AlienXXX/sounds/223009/
http://www.freesound.org/people/AlienXXX/sounds/223111/
http://www.freesound.org/people/AlienXXX/sounds/223112/
http://www.freesound.org/people/AlienXXX/sounds/223113/
http://www.freesound.org/people/AlienXXX/sounds/223114/
No master pieces in the lot, just simple examples to illustrate what can be done with the apparently crazy sounds from image-to-sound conversion.
I have used only tiny fragments from only some of the sounds on copyc4t's pack. Even from such a limited set it is possible to obtain a large range of sounds via use of pitching, filtering and other effects.
The file obtained from the conversion of one single image to sound (or databending) will likely contain many independent sounds that can be trimmed, isolated and processed.
I didn't realise what a fun toy/tool audiopaint is! I also messed around with some different file types as raw data in audacity, the results I've uploaded in a pack.
http://www.freesound.org/people/luckylittleraven/packs/14100/
@copycat it's really interesting how the rotation affected the sound so much
@alienXXX nice! you really took it to the next level
Thanks for sharing these articles and this dare, it's really fascinating
@luckylittleraven:
Interesting indeed, a consequence of how geometry is turned to sound, and once you notice the relationship between the picture elements being scanned and the sound being produced, you can draw something from scratch in a graphical program and already have a rough idea of how it will sound
Awesome sound pack !!
@AlienXXX:
Thank you for making so many cool sounds and loops!
I am now trying to build a drum kit using copyc4t sounds.
http://www.freesound.org/people/AlienXXX/sounds/223133/
http://www.freesound.org/people/AlienXXX/sounds/223132/
http://www.freesound.org/people/AlienXXX/sounds/223243/
http://www.freesound.org/people/AlienXXX/sounds/223244/
http://www.freesound.org/people/AlienXXX/sounds/223245/
http://www.freesound.org/people/AlienXXX/sounds/223246/
Would be great to see others creating more sounds using the methods and software suggested do far, or taking some of the sounds already submitted and doing something with them.
luckylittleraven wrote:
I didn't realise what a fun toy/tool audiopaint is! I also messed around with some different file types as raw data in audacity, the results I've uploaded in a pack.http://www.freesound.org/people/luckylittleraven/packs/14100/
@copycat it's really interesting how the rotation affected the sound so much
@alienXXX nice! you really took it to the next level
Thanks for sharing these articles and this dare, it's really fascinating
Cool sounds!
I can tell you have been playing with different tools. - Also make sure to experiment with the options on file import into audacity (or other editor) when importing a file as raw data. Selecting between 16 or 32 bit, for example can make a big difference to the resulting sound.
Building a 'drum kit' from sounds derived from copyc4t's image-to-sound samples.
Today I added a few more bassdrums and started adding some snaredums to the pack.
http://www.freesound.org/people/AlienXXX/packs/14106/
A drumkit! I love it, what a great idea.
Also thanks copycat, I had a lot of fun creating that pack, but probably more fun writing the descriptions and imagining
32 bit made a big difference, much cleaner looking and sounding (it seems funny to decribe a sound as cleaner looking haha and it's still really just a load of intense noise) It's such a weird concept but so cool. My computer finally has a voice and it does not sound happy. Some of those raw data sounds actually make me feel weird like it's messing with my energy man
Well, since AlienXXX is honouring me with a... "drum kat"
How about getting some of the goodies from luckylittleraven's pack?
There are many synth leads and FX in those sounds, but for starters here's the base of the LuckyLittleDrumkit: (update: all these sounds have been moderated and are now available! )
http://www.freesound.org/people/copyc4t/sounds/223458/
http://www.freesound.org/people/copyc4t/sounds/223459/
http://www.freesound.org/people/copyc4t/sounds/223460/
http://www.freesound.org/people/copyc4t/sounds/223461/
http://www.freesound.org/people/copyc4t/sounds/223462/
http://www.freesound.org/people/copyc4t/sounds/223463/
and a 120bpm loop showcasing it:
http://www.freesound.org/people/copyc4t/sounds/223457/
luckylittleraven wrote:
A drumkit! I love it, what a great idea.
Lets see if I can actually put together some fairly decent sounds for all the drum pieces.
luckylittleraven wrote:
32 bit made a big difference, much cleaner looking and sounding (it seems funny to decribe a sound as cleaner looking haha and it's still really just a load of intense noise)
Well, having done this before, in some cases I know what to expect or can tell what is going on (although there are still plenty of surprises to keep me entertained and interested).
The "vrrrrr" kind of sound all your samples had and the vertical light-dark-light-dark lines on the waveform display were a telltale sign: there was a very marked difference between parts of the waveform when the image was converted to sound. You can see this if you import the sound into a sample editor and really zoom in on the waveform.
This gave it what I will describe as a 'fast rate chopped-up sound' quality.
Your second sound still sounds wierd, but when you look at it zoomed up in sound editor you see that the waveform is more 'normal' - mostly triangular and trapeze shape waves.
(you will also notice the sound has shrunk to 1/4 the size, since you are now taking 4 times more bits to create each 'sound point' from the same file)
(while we are on this subject... if you want to make a sound completely horrible... take a normal 16 bit sound, export in Audacity as raw data and re-import it as 8-bit. Or vice-versa.
Play at low volume... You have been warned!)
luckylittleraven wrote:
It's such a weird concept but so cool. My computer finally has a voice and it does not sound happy. Some of those raw data sounds actually make me feel weird like it's messing with my energy man
Well, one of the characteristics of these synthesis techniques, image-to-sound and databending is that you get really strange and unnatural sounds.
(both techniques are basically both based on converting a different type of data into sound. With the direct raw data conversion to sound being by far the most brutal)
To begin with, most natural sounds have a higher contents of lower frequencies, and energy contents of the sound is progressively lower at higher frequencies. Even synths follow this principle, and pretty much every synth uses a lowpass filter to achieve precisely this effect.
Sounds without this gradual decrease of intensity for increasing frequencies sound very harsh.
Even on less classical kinds of synthesis, such as FM (frequency modulation), sounds retain a higher contents of lower frequencies (except if the modulation intensity is very large or if the carrier and modulator are non-related frequencies - but since sounds produced this way are not musically pleasant, they are rarely used and many synths are not even designed to generate them)
The sound produced from data is not white noise either. There are all sorts of patterns in that date which prevent the sound from being white noise (think of text and, for example, how spaces are always position between each word or how certain letters occur more frequently than others - those are 'patterns', in the sense that the data is not entirely random=white noise).
When sounds are not white noise, our brain expects the frequencies to relate to each other, in particular we expect harmonics to be present (a simplified explanation: higher frequencies in the signal are multiples of the lower frequencies). This relationship is also absent, since whatever patterns exist in the data that was transformed to sound, they do not relate at all to the normal way sound is produced (by vibrations in physical bodies, or even by oscillation of electric circuits)
There is nothing 'spooky' with the sounds themselves, but our brain struggles with the fact that they are so unnatural. (shivers down your spine when you listen to these?) They tap into the basic reaction of fear when faced with something that is unknown or unnexplicable.
This theory is easy to test - put the sound through a low pass filter with a gentle slope. Does it sound more natural and less menacing now? - it is because it now obeys the rule of decreasing intensities with increasing frequency. Probably still sounds unpleasand, but no longer sounds (so) un-natural.
Appologies for such long post... I will be quiet now...
3 of the many faces of this sound:
http://www.freesound.org/people/luckylittleraven/sounds/223069/
For the GM/GS aficionados, a helicopter!
http://www.freesound.org/people/copyc4t/sounds/223478/
A synth bass,
http://www.freesound.org/people/copyc4t/sounds/223479/
A distorted lead,
http://www.freesound.org/people/copyc4t/sounds/223480/
As much as I'd like to participate, I just can't get my head into sound stuff at the moment. But I have been searching for images that might be useful for sonification. I've tried a few of these in Bitmaps & Waves and they seem to produce some interesting sounds which show potential for further processing, but presently I can't get to the next level.
I've noticed also from using the B&W; software that abstract geometric type images usually work best, at least according to my tastes. Try stuff by Victor Vasarely, Piet Mondrian or Bridget Riley for example, stuff like that.
Anyway, here's the link to some interesting images I found recently. The other links resides (hopefully, if I actually bookmarked them) on my Linux partition, so I'll dig them out later. It's a Windows evening.
http://www.imcce.fr/langues/en/ephemerides/phenomenes/meteor/CABERNET/index.php
Oh, and this guy has some great tutorials on image data bending and other stuff. http://blog.animalswithinanimals.com/
Hi toiletrolltube Thanks for participating in this thread, man!
Good suggestions are just as valuable as good sounds. So thanks.
And especially since you are covering a subject we actually (strangely) had not yet touched on: what pictures (likely) make good sounds?
toiletrolltube wrote:
I've noticed also from using the B&W; software that abstract geometric type images usually work best, at least according to my tastes. Try stuff by Victor Vasarely, Piet Mondrian or Bridget Riley for example, stuff like that.
This is VERY true.
I have had a look at few images by Victor Vasarely on a google search and agree with your assessment.
Also, suggest you all try 'sonifying' (is that a word?... It is now!) simple line drawings (white on black, for best results).
Search for some 3D wireframe pictures and try those. You will be surprised!
Well it's nice to be here Mr. Alien - despite some of the things I say. Actually, this thread is starting to become like kindling for the almost dying embers in my audio hearth. I just need to blow on it a little. Maybe.
And I'll follow up on your suggestions about sonifying line drawings, thanks. Sonifying is a great word
Oh, and here is another link to recent finds: http://www.rabbitholeprojects.com/images/03-nov-11-phillip-stearns-subliminal-machines
I normally don't like copying over to ntfs, but sometimes it's worth the risk.
Something from the alien menacing side of luckylittleraven's pack:
http://www.freesound.org/people/copyc4t/sounds/223508/
A Sci-fi alarm, and since we're about Dare 26, it's at 120bpm and loopable
It's a simple cut, slightly down-pitched, of this sound:
http://www.freesound.org/people/luckylittleraven/sounds/223071/