This what I have for you guys today. I spent the bulk of my free time today finding classic modular synth records that I used to have. I used to order weird records from this guy "Jack Diamond" who I can only find the slightest traces of on google, so I guess he's either dead or out of the online record selling game. Anyways, I found some of the old Jean-Jaques Perry and Gershon Kingsley albums an awesome and a few soundtracks that I had. And it cut into my random noise making.

In it's place I offer something perhaps even more exciting?

This is a synth loop, doubled on piano and clavinet with a little drum loop to go with it. It's kind of like the birth of a song. This isn't really what "makenewsounds" is all about, but I think it's okay because it is a pretty cool sound ultimately.

The only difference is that I am not providing the Wav file this time, since this is a bit of noise that will someday be part of a song, and I'm not releasing it yet. It's more of a peek at the process. Main synth loop done with massive, piano and clav with kontakt, and the drums are from a sample pack I downloaded for free from defeq.com which are unfortunately not up anymore, sequenced in ableton live using impulse.

Today I started out by working with this free granular synth program, my intention being to explain how granular synthesis works, it being a pretty cool way of making noise.

Unfortunately the program crashed about 80 times so I gave up on it for now. It runs fine on it's own but when I route audio to any type of recording software it spazzes out.

So instead I decided to take the simple little melody I based it around and just mess around with it in the amazing native instruments synth "massive". I won't bore you with all the details of the routings, but suffice it to say that when I make a video showing off this synth in my tutorials section (soon to come) it's going to be very fun.

This is also the longest sound on MNS so far, by a wide margin I think.

Here's the wav file.

This is a simple concept that I am really glad worked out. The last couple days I have done sounds that I was simply not happy with, and I only have a certain chunk of time to work on this each day so I was pretty much out of luck. Here I used supercolider to basically make sine waves (the simplest sound possible) I start with each one at a totally random pitch, and over 2 minutes they slowly drift from that random pitch to a place in a just tuned chord.

Simple, but effective.

Here's the wav file.

So for #5 here, I started out with a short snippet of a Busta Rhymes acapella. I basically opened it in Ableton live, and zoomed up super close and took tiny little sections of it and made them loop, usually between 5 and 10 times. I tried to be pretty random with things, and to switch up how long the tiny loops were as I went. The results are pretty varied. You get some quick repeating things, and some more textural stuff. I tried to do a lot of variation in the chops, and put a lot of effort in. Some of the things I spent a ton of time on don't really show it. For example, I would spend 20 minutes doing these very small loops that change in mathematically proportionate ways (like loop one thing 7 times then the next one would be a 5/7 the length and 5 times, then 8/5 the length and 8 times, etc..) and the end result is a tenth of a second of "bleep!". I also got a bit "wacky" towards the end and tried a couple weird juxtapositions. I did, for example, one loop that got progressively shorter every other loop, and longer on the alternate loops, you can hear it when he is sayig "duty yo" over and over. Weird stuff like that. What you have is a couple lines of Busta extended to over a minute. Turns out his lyrics are deep after all.

Here's the Wav file.

Talk about this here.

This is the simplest sound so far in terms of process. I simply created a 0.1 second burst of white noise in supercollider, then I recorded it and brought it into ableton live and applied a delay to it, recorded it, then reversed the recording (made it backwards) then added a different delay, then reversed that and added another and so forth. The result has a nice quality to it that goes beyond how easy it was to get.

There also seems to be some weird phase canceling going on that makes it drop out in weird ways.

Here's the wav.

talk about it here.

When I started this one, I was thinking I would do something with a vocoder. For anyone who doesn't know, a vocoder is a machine that takes one sound (traditionally a synth) and uses another sound to give it an envelope (traditionally a voice). This is usually the effect you're hearing when it sounds like a computer is talking. Think about daft punk's "around the world" or really almost any of their songs.

I don't own a hardware vocoder but there's a decent one on the nord G2. What I often like to do is to use something else in place of the voice, so instead of making a synth "talk" I give it the envelope properties of another synth or sound.

What I did here, was to create a sort of evolving noise patch. I used a whole bunch of LFOs (a repeating wave that you can use to turn things on/off/up/down) to make it sort of random sounding. Once I had my weird noise patch, which I was going to use as the "voice" part of the equation, I made a fairly simple chord out of different saw waves to run as the "synth" part.

By accident, I routed the noise patch into both parts of the vocoder. This has no effect on the sound since it's just putting it's own envelope onto itself, but I thought it might be cool to double the effect, which is possible. That's a way to bring out the words more clearly in the synth, but it turned out not to be too interesting here. I was about to get back on my planned path when I remembered another thing the vocoder can do, which is reverse the frequency bands. What this means is that the envelope from the highest pitched part of the "voice" signal is applied to the lowest pitched part of the "synth" one, the lowest to the highest and so on down the line. I flipped that switch and was pleasantly surprised by how much more interesting the patch got. I ended up using this reversed vocoder as the "voice" part of another vocoder, routed the chord into it, and recorded the output, which is today's sound.

here's the wav file

Why not go here and chat about it?

Here's part 2. So I decided to start off with some sampling. I found a video on youtube of one of my favorite drummers, Hamid Drake, playing a sort of funky little pattern. You can see the video here I used a free program called soundflower to route the audio from the clip into ableton live. In live, I isolated a drum loop from the recording. I then slowed it down to about 1/4 the original speed. In live, you can slow down a sound without changing the pitch, and when you change the speed radically, it introduces a lot of noise to the sound. There is a setting where it isolates what it thinks are the hits in a sound, and stretches around them, which creates a rhythmic sort of noise when taken to extremes. I had a nice sounding little loop but I wanted some more variation so I ran it through a comb filter. A comb filter is a sort of complicated system whereby a signal is sent back to itself delayed by a small amount which creates interference between the waves, in practical terms, it means that you can carve out parts of a sound and leave only certain frequencies much stronger. I did this several times with several different frequencies so that I had 4 different versions of the same loop, but all with different frequencies accentuated. I then went through and cut out a few hits and made the basic "beat" you hear in the sample. In live, you can take a small section of a sample, and "scrub" through it, which is to say, have it start wherever you like in the sample. This is really cool because you can take a sample and drop it onto the 1st beat of the song, then you can move the little marker around and decide what part of the sample you want the first beat to be. I did this a bunch of times and ended up with what you hear. No further editing or processing.

Here's the wav file.

This is the first one.

I started this off in supercollider. It's a free program that is like the rabbit hole from Alice in wonderland. Once you get in you realize that nothing is what you thought it was. You basically have to write out your sounds and sequences and everything in object-oriented computer code, like C+. This sound started off with a simple sin wave and I repeatedly added in modulations into every place I could. I added a couple of filters which I also modulated and then exported the sound and brought it into ableton live. I filtered the high end and sent that to another track which I applied some spectral filtering (ie: I used a plugin that applies a convoluted process to separate the sound into a tons of thin frequency bands and applies different effects to the different bands). I then filtered the low end and sent that to a track and put a tesla coil plug in on it. Those of you who read my blog saw the awesome musical tesla coil video I posted. This plugin basically turns the audio you send to it into a digital equivalent of what happens in a musical tesla coil. I mixed those three things (the original, the spectral filtered and the teslafied) into one, and applied a bit of compression and reverb and voila! You have a very weird sound.

Here's the wav file for anyone who wants to mess with this at all:

http://www.mediafire.com/?zwmtmzzwdoy


Stay tuned for more in depth analysis of the upcoming sounds.

This is a new blog I'm starting where I make sounds and samples and things and explain what I made. All sounds posted here will be 100% original, all available for anyone to use 100% royalty free.

Peace.