/musicprod/ - Music Production

this is completely new to me so forgive my lack of knowledge

i tried it out and converted this image to tif. then i added to audacity and got the following sound. now, how did you convert it back?

https://instaud.io/cd4

again, forgive my ignorance for it is there.

>>1261

>>1262

thanks for the interest

to convert the file back to image, you export it, select "other uncompressed types" as file type, then you go to the options and select "raw" in header option.

If you for instance imported the picture with "U-law" encoding, then you must export it likewise. Same goes for all the other encoding options.

It's crucial, you don't change the first few samples inside audacity at all. You must also export the file as the same type as it started out to be.

I've exported 30 or so seconds of "Sunday Morning" by Velvet Underground to a .bmp file, which is first the picture.

The second has been resized to 200% on width and height. It turned out slower than the original when imported, who would have thought.

At the last example, I've copied the entire image, and put it right next to the other, and then stretched the copy to 50% of original width.

It's hard to describe, it sounds like the track is running on two different pitches, while stuttering and keeping same tempo.

You should really listen to it, though.

I've taken the liberty of using the click removal on the results, because the quality gets mangled for each transformation. It will appear some places in the audio.

(Default)

http://vocaroo.com/i/s1C14O82ssBV

(2*Width and Height)

http://vocaroo.com/i/s0XMggoQd7jG

(Half-length copy besides original)

http://vocaroo.com/i/s1LSqPKrGtow

It's not that fancy, but in consideration to, how simple editing it was, I'm not disappointed at all.

>>1266

yep same, no matter what i do, not working. could be i'm missing some information

>>1269

oh boy, forgot all about embedding and uploading doesn't go at once.

Here's the .tif, and the "Noise Median filter"

Another sweet as fuck thing I've learned about for using sound to mess up images is to convert to color-supporting SSTV audio signals, save them as waves, Distort them in different ways or add sounds to it (like playing music faintly on top of it) and then reading the now distorted SSTV signal (not all distortion methods work, and you have to leave the start of the sound intact to be readable)

i've had some.... interesting results with this picture.

>>1271

I got the "server is busy but your post was probably accepted" message with 3 of my posts, but this one is visible, the 3 before it are not. hopefully they'll show up later.

Here's another thing I did.

I took an image (of neon genesis) and opened Gimp and split it in to 3 monochrome images (RGB) which can then later be recombined into a color image.

I then loaded those images in 3 different instances of Harmor (image gain, but not image frequency) experimented with some subtle in-harmor effects for contrast differences, added some drums to different parts to see how that would show up, and played each channel in sequence, screencapping the monochrome spectragram, cropped them into individual images, re-aligned them, imported into gim, and compose color image from RGB images.

This is the result, after some minor color correction because it was really really washed out because I the effects I used basically lowered the contrast.

So it stroke me, you can choose the picture dimensions, of which you'd input an image on.

It could be valuable to know, if I was to "sphererize" my image horizontally. To stretch the middle so it's larger than the sides.

In the meantime, I've changed track to work with, now it's "Happy Cycling" by Boards of Canada. Mainly because there's more percussion to it, while still having pure tones.

First image is of this sound, now put to a 1:10 format rather than 1:1.

In the second picture, I tried to "scale" it 1% in ms and then move a tiny part of it from left to right side, so it forms a rectangle rather than a parallelogram. Pretty interesting outcome. The picture is just an illustration, though.

I tried this on a 10:1.2 version as well, but got barely any changes.

I even tried with 40% scale, but it only made some clicking noises.

Third sound is like the first, but moved flipped horizontally. It sounds a bit familiar to picture 2.

At the last picture, the original and horizon variant are side by side, so they are like a mirror to each other.

>>1281

I couldn't see if I was uploading more than one picture but anyways. Here's the fourth picture. And the sound for all of them.

(Default Happy Cycling)

http://vocaroo.com/i/s1S0Pyl73Spe

(1%Scala Rectangle)

http://vocaroo.com/i/s0tWrO9c7SDf

(Horizontflip)

http://vocaroo.com/i/s0Zd5v4yqdXa

(Horizontflip+Default (or mirror))

http://vocaroo.com/i/s0Sa9hrhP6CW

Still pretty fascinated.

>>1282

I downloaded the first of those (the default) to use as comparison, and I did some basic things to it with Harmor.

Gain and frequency image resynthesis compressed to 20% vertically 100% horizontally in MS Paint, then stretched to original size, resulting in massive data loss.

http://vocaroo.com/i/s1V8BoZZEWBA

Gain image only compressed down to 10% of original and stretched back out

http://vocaroo.com/i/s0z6GrkCDpvZ

Frequency compressed to 10% verticaly only.

http://vocaroo.com/i/s1BEOmhiJKA1

Compressed Gain image, loaded as frequency image (intensity as pitch shifting rather than volume)

http://vocaroo.com/i/s0JUGpIgCvaF

Frequency image loaded as gain. (intensity as volume rather than pitch shifting) surprisingly still has those 2 bell sounds

http://vocaroo.com/i/s1b8TFYXhr7F

Same as above, but I turned scale UP, because doing so NORMALLY results in higher clarity IE quieter sounds get even quieter and loud sounds get louder. However, all the frequencies are loud with this method so it came out... very weird.

http://vocaroo.com/i/s1r9NU1mW7uj

also here's a gif i made out of some databending I did a long time ago back when I also used /x/. here's hoping 5.56 mb isn't too big.

>>1287

I think some of my posts has sunk into the vast ocean grunge of this board as well.

Anyways, these are pretty good results, you've come up with, I'm just not understanding quite what you did to get them.

I've been playing around with some photoshop filters recently. "Diffuse" was one of the few that yielded any noteworthy outputs. First picture has been "Normal diffused". The second is "Dark Diffuse", the third "Light Diffuse", and the last is a combination of all the sounds, but there's only the sound, because I'm too lazy to import that mess to a picture again.

Although it sounds like nothing but dataloss, there's some effect to it anyways. It reminds me of the sound of snaredrums. It could probably get interdasting on lone instruments, like a single kick or hihat note.

(Normal Diffuse)

http://vocaroo.com/i/s1ssOjKTGAzh

(Dark Diffuse)

http://vocaroo.com/i/s0KLXamRNrFT

(Light Diffuse) - Reminds a lot about the dark diffuse, but there's some minor changes anyways.

http://vocaroo.com/i/s1PhgK4WMBK5

(Combo)

http://vocaroo.com/i/s1LnH3TZT1U8

So I tried to edit with another program this time, some pretty nice alternative to photoshop, called "Photo Filter".

One of the major upsides is the ease to use, which came very handy for the inexperienced and lazy bum, that I am.

I quickly learned how to apply some filters, I didn't find on Photoshop, such as "Lake Reflection", which is the first picture. Possibly one of the most interesting things so far.

Second pic is another filter, called "Glass". If you acutally do zoom in, you'll see the picture being completely rearranged,

which I found pretty interesting, compared to learning, how the sound actually is exported unto the image,

and how to manipulate it.

Third picture has been applied with "Vertical Ripple", with as low changes as possible (3). Sounds similar to "Glass".

In the fourth and fifth, I changed the color channels, first from RGB to BGR and then RGB to RBG. I don't know how it works at all, but you would think the changes are unimaginable minor? It seems not. I even tried to turn it B/W, which resulted in static distortion, so colors are a pretty important deal in this science. The Vocaroo might munch up the quality enough to be inaudible, but you can try this yourself, if you are skeptic.

(Lake reflection)

http://vocaroo.com/i/s0Z6cK0s2Sot

(Glass)

http://vocaroo.com/i/s1foGaoSNu4n

(Vertical Ripple 3)

http://vocaroo.com/i/s0eYcKYFJ8M6

(RBG (Green Blue Switch))

http://vocaroo.com/i/s07xZWPftdMK

It DID turn out to sound too much like default, due to vocaroo's quality.

(BGR (Blue Red Switch)

Same here. It's such a shame really. I might get to upload the difference elsewhere, because it's definitely noticeable.

>>1304

Fair enough

I'm very familiar with the beepmap, so understanding the harmor wasn't a deal at all.

Thanks for the very detailed explanation nevertheless.

>>1366

Update.

So the most interesting thing about those distortions I did by sending beepmap images through SSTV audio and distorting them that way, is that it makes the image "jagged" and shifts color, but since color is meaningless in harmor for "beepmaps" the only thing relevent was jaggedness. SOO I simulated jaggedness by using stretch/skew in MSPAINT as screencapped here.

Unfortunately it didn't sound all that interesting. at all. It sounded like a sloshy sort of phaser.

Normally those higher harmonics are extra texture for the fundamentals, but when they're shifted away they just sound like so much white noise, so when they're swept over like that, they sound like white noise with some random tone, and a phaser.

The main body of the sound, was a bit jarbled and warped in a semi-interesting waybut the only way it was clear was at the very start of the sound before the high frequencies caught up, at which point it's like it was also low-passed.

So I thought "I'd have to split this into bands, and warp them each different amounts and ONLY warp mid to low frequencies and leave the high frequencies unchanged.

This will sound sort of like a mix of a phaser and a really high feedback flanger, I can already tell just looking at the spectrum that that will happen.

For this to be interesting I need to be able to do high intensity edits of the bottom 1/10th of the image, which is only a few dozen pixels, and that's... well... that's not going to work, not only that but I'd have to be able to do it consistently and with a curve rather than linear (because audio doesn't scale linearly) which means I'd need to really understand some advanced photoshop tools, which I simply don't have any other use for.

I didn't save it because it just sounded like some kinda floppy phaser and flanger combination with way too much feedback where the flanger went way out of sync. might be interesting for reggae dub, but I don't into reggae dub so...

I've been trying for a while to import .tif images to audacity, and realized it is only possible, if the color channels are "non-interleaved".

So then I began a long search on programs to export tif files like this, but "ifranview" was the only program I found, and I am not even sure, if it can directly import or export tif files. So I'm just going to keep using photoshop, and if you want to experiement with tif files as well, pirate that shit, m8.

Importing a tif in audacity wasn't a big deal, it just needs to be imported by "Unsigned 8 bit PCM" encoding. Then, like for bmp, you drag a sound onto the audio track of the tif file and mix them together, and then you export as "Raw (Headerless)" and "Unsigned 8 bit PCM" again. For the file name, just put in the .tif extension to make it a picture file at once.

So this is the result. "Happy cycling", but now as a tif. As it is pretty obvios, non-interleaved tif-files are much more colorful. Even if you'd ignore the red and blue sections and only look at the middle. It sounds pretty much the same when exported to wav, but the file has for some reason another more messed version of the same sound coming right after. I'm mostly interested in the first part of the file, though.

Since the picture of the sound now is more colorful, I thought about adding changing the hue to it. The second picture, I did so by 10 degrees.

The third has been changed by 90 degrees, and the fourth as been changed by 180. It seems, the more degrees, the more it messes up the rhythm of the track. It sounds pretty glitchy alright. I thought tif and bmp would work somewhat the same, and changing the color or hue to a bmp usually just kills the sound quality, but these were some really unexpected results. I'm going to do more things related to colors from now on.

TIF Cycling (Default)

http://vocaroo.com/i/s1dxKgcrqkkq

TIF Cycling (+10 Hue)

http://vocaroo.com/i/s1eBJP5vIl3l

TIF Cycling (+90 Hue)

http://vocaroo.com/i/s0bRhXn5ft3G

TIF Cycling (+180 Hue)

http://vocaroo.com/i/s12GALA2bqVa

>>1367

I would actually really love to hear, what you've made.

For me, it's all about creating and deforming sounds into something, you wouldn't be able to by any other means.

great, I didn't upload +180 hue.

peace oot