Posted
If you have any sound bites of work you've done in Synfire/Omnisphere, I would love to hear it!
Here is a little excerpt I threw together while watching some football on Sunday. :)
(http://hackintosh.solar-sound.com/~phil/synfire/synfire_test-3.m4a)
(I couldn't attach the sound file, so this is just a straight link until I can create a suitable format.)
This was a bit of a special approach/challenge in that all of the MIDI data was generated randomly in a program called Noatikl (by way of a script from Earthmuse), and the result was then imported into Synfire to be chopped up into 4 or 8 bar phrases, etc.
I applied harmony, a very crude arrangement, and then the whole thing was dumped out into Ableton in one fell swoop.
There isn't much in the way of proper mixing, EQ, or arranging, and the drum loop is just a plain placeholder for now (it is not a final product yet.)
I plan to do some more work on it later.
"12.5" Excerpt (Work in Progress) (12.5__excerpt__WIP.mp3)
125_excerpt_wip_169.mp3
125_excerpt_wip_169.mp3
Mi., 16.12.2009 - 02:41 Permalink
Keith:
I really love the texture and possibilities. I've never heard of this before, but I had never heard of Synfire before either.
So how do you envision the process? Generate random midi through Noatikl and import into Synfire where you apply harmonic content and do some phrasing and rhythmn?
I guess I should read the user manual and download a demo version to try. The only problem is (as you can tell from many of my previous posts) is that this looks like a good vehicle for someone with a lot of technical expertise, as it's not just a simple plugin to Synfire.
I would be interested in hearing your idea of how to utilize this tool in composition. Good stuff!
John
Mi., 16.12.2009 - 04:42 Permalink
So how do you envision the process? Generate random midi through Noatikl and import into Synfire where you apply harmonic content and do some phrasing and rhythmn?
That is exactly how I envision the process (and is what I did to generate this piece primarily.)
I hadn't experimented with Noatikl->Synfire before, so this is all in a sort of test phase.
I would be interested in hearing your idea of how to utilize this tool in composition. Good stuff!
I'll post more as it comes down the pipe. Still have a lot of things to play with. :)
Mi., 16.12.2009 - 09:18 Permalink
I like the example and the general idea of using Synfire to process random material generated by other software.
On our long term agenda, we have an idea for an XML import/export feature that will read and write phrases in XML format. This will allow external software to synthesize music for further processing in Synfire and vice versa.
Do., 17.12.2009 - 18:06 Permalink
On our long term agenda, we have an idea for an XML import/export feature that will read and write phrases in XML format. This will allow external software to synthesize music for further processing in Synfire and vice versa.
I think this will be very interesting to explore once it is implemented. I can't wait. :)
Sa., 19.12.2009 - 06:29 Permalink
Re: Noatikl.
I've read the entire manual, and played around with it a little bit.
It's pretty cool!
But it is kind of overpriced for what it does, IMO.
The main strike against it is that it is not deterministic... i.e. as far as I can see it doesn't store a seed (for the random number generator) for each song. Every time you press play the result is different.
The thing I really do like about it is the "follows" voice... which is something I'd like to see in Synfire. :D
Sa., 19.12.2009 - 12:37 Permalink
As a principle of design, Synfire stores random seeds where ever random streams are involved. Nothing will change unless the user deliberatey requests a new seed.
I love generative music. It's relaxing and refreshing. However, the human brain quickly recognizes the patterns behind the algorithms, even if a generator noodles random tunes for hours. Therefore the perceived musical expression is rather constant and meditative.
You find some examples of generative music attached that I created about 12-14 years ago, using a very early precursor of Synfire - "Leviathan".
Lingerie: Generative expression rendered on a Waldorf MicroWave. Pure, unedited output.
Alien Rain: Ditto., using a Korg Wavestation A/D
Sparx: Algorithmic percussion, generated phrases, arrangement manually optimized (copy/paste)
Lingerie (lingerie.mp3)
lingerie_209.mp3
Alien Rain (alien_rain.mp3)
alien_rain_127.mp3
Sparx (sparx.mp3)
sparx_146.mp3
lingerie_209.mp3
alien_rain_127.mp3
sparx_146.mp3
So., 20.12.2009 - 08:44 Permalink
I was noodling around tonight with an artificial learning based improvisation program written in Lisp and Max/MSP. I soloed the random piano portion of the above "12.5" excerpt into the AI program for 16 bars, and re-recorded the output back into an armed recording track in Synfire.
I harmonized the recorded result and reproduced a new figure which I copied to a new document and assigned a gated reverse Ebowed guitar sample to. The original piano part was reassigned to a CS-80 self oscillating filter patch, and I added a quick drum loop and bass drone to the part. The exact same E minor progression from "12.5" was used.
A bit of recycling, yet a very different result. :D
Maybe I can meld the two parts together in an interesting fashion...
(http://hackintosh.solar-sound.com/~phil/synfire/synfire_test-4.m4a)
"12.6" (Clip) (12.6__excerpt__WIP.mp3)
126_excerpt_wip_202.mp3
126_excerpt_wip_202.mp3
So., 20.12.2009 - 18:07 Permalink
I love generative music. It's relaxing and refreshing. However, the human brain quickly recognizes the patterns behind the algorithms, even if a generator noodles random tunes for hours. Therefore the perceived musical expression is rather constant and meditative.
To a large extent I'd have to agree with you. That's why generative software is so well-suited to ambient output, and why at the same time it is often so difficult to manifest the simplest of chord progressions it seems . . .
It can be done however, and to me it seems that as always, the depth of the result is in large part a function of the depth of the input.
This example, although not harmonically very complex, does afford some nice and less-predictable output (solo violin especially). So it can be done, but one reaches a point where they have to say 'hmm, i could really do this myself' (but mostly i'm just too damn lazy) . . .
So., 20.12.2009 - 19:24 Permalink
I like the track. Are you using Max for Live? That marriage is pretty enticing, though I am weary of getting back into the MAX world with some much left to learn on Synfire.Your description reminded me of this...
&feature=related
:)
Sadly, no. And I had an awful of a time getting anything to work, mainly because I had the Max 5 runtime. This thing only worked properly in 4.6.. It would be nice to see it ported to Max for Live though, would be one hell of a tool for that program.
No crabs were hurt in this process, but I can't say that much about my own braincells...WAAY too much beer. :sick:
Edit: there are two other versions of the program that do work with Max5, but they seem to only work with audio... :(