New Gadget Madly In Hope

24 May 2015

The Lost Chord

Droneo and Srutibox users: while you wait for the astounding Droneo 2.0 to come out ... me too actually ... I think I've discovered the Lost Chord, and you can play it with Srutibox and Droneo,and also online at
The Lost Chord

for reference: ,

Actually lots of people have discovered the Lost Chord, but didn't know it or call it that. It's actually a whole class of lost chords. Coming from me, I bet you can figure out that you can't play them in equal temperament.

If you recall Harry Partch's "One Footed Bride" and other measures of interval consonance, you'll see a big lump - the foot if you will - near the octave (2/1):

The Bride's missing left foot (1/1) is actually where the Lost Chords reside. That is to say, the mind altering, fascinating, hypnotic chords of song and story are the extremely close intervals near 1/1.
Anyone with a comb filter or chorus/flange effect could have told you that, but here's the thing: as the interference beats get slower, the more this timbral fusion passes from a perceived effect to a composition itself. Also, you need a really stable and pure bunch of sound sources, which these two apps amply provide.

Sir Arthur Sullivan - or rather Adelaide Anne Proctor - appropriately provides a clue in that the Lost Chord is played on an organ, which has the feature of rather stable tuning and lack of modulation - which means a pair of really closely tuned reeds or pipes could actually produce "Lost Chords". In fact, it's hard to imagine any meaningful practical microtonal research without the stability of reed and pipe organ "oscillators". Strings - except for Ellen Fullman's Long String Instrument - have too many unstabilizing influences!

So... how does one experiment with the Lost Chords using SrutiBox or Droneo?

Pick a base frequency that you like - your'e going to hear it for a while, like 200 (an A) or 432 if you are one of those 432 people. You actually might like the lower octaves better.

Set up the first reed to play it (setting the interval to 1/1 or 0 cents).

Set up a second reed to just a tiny bit sharp or flat - so tiny that its cycle is 8 or so seconds or more. an easy way to do that is to set it to a ratio like:

(f+(1/cycle time))/base Freq, so, 432.125/432 should have a beat of once every 8 seconds.
Because these are Fractions, which you all learned about in New Math, you could multiply these numbers up by a few hundreds to make them easier to type, e.g. 432001/432000 would beat once in 1,000 seconds.

Obviously, if you move the base frequency down, say, an octave, the beat will also correspondingly move down. The important thing is to keep that interval below, say, 101/100 or 1.01 cents. There's actually a big perceptual difference the slower you go.

Now you can change the timbre to something other than a sine and hear a really slow comb filter effect. Add a few more harmonically (rhythmically) related close intervals and bask in the Lostness. You can also fiddle with a mix of these intervals using the Churn and chorus features of these two apps.

Droneo/Srutibox are pretty accurate in the frequency department, but they do have some limits since they are basically phase incremented wavetables. Droneo 2.0 will have a floating point interpolated phase that probably would be accurate enough for years.

Happy music making!
13:13:08 - jhhlnet - No comments yet, click here to add one

06 May 2015

What's doing?

Hi Folks, I haven't checked in for a while!
Here's the story:
I have a pretty good version of Droneo 1.4 in January... but it had a number of bad issues in it:
For years, the control rate calculations like evolution and patters, were done at the beginning of the buffer computation cycle. But changing the engine to AudioBus meant that the size of the buffer - and the speed at which it updates, was not under Droneo's control. So that meant that evolution would proceed at an undetermined pace.
Also, the conversion to AB also made my good old iPhone 4 hiccup with more than about 4 reeds, since it wasn't efficient enough and also maybe because the audio IO is now floating point based, which doesn't seem to bother the newer devices but might be slower on the old ones. Why do I care? because the iPhone 4 is my home entertainment system device.
Droneo 1.4 had a number of innovations that you will never see. But that won't matter becasue I'm working on
Droneo 2.0 (and abandoning Dronica, which was to be basically what Droneo 2.0 will be).
I've been going back and forth about whether to make a new app or just replace the old one since about 2011!

Anyway Droneo 2.0 is going to be pretty modern. It'll support all the iOS device sizes in portrait and landscape.
It will deliver the promise of Dronica, that is, it simplifies and empowers the features of Droneo: no limits on number of reeds, Churning, Consorts, and Mirrors are replaced with evolution between states (scaled beats), and the pace of that evolution is governed by a heartbeat. Voice banks are being replaced with a tag-based system, and old banks and patches will be legible. There'll be a tighter connection to the Droneo Voices web service, so you can post and retrieve Droneo voices to the public. I'm not sure, but I think you'll be able to stuff them in iCloud too, and pick up your own Drones on different devices.

The Tone Spiral is gaining power as it includes thousands of Scala scales as Guides. It may also be able to import and export scales from Gestrument.

The interval Spec is also expanding, adding ideas that are also in PolyHarp: for example, I have a notation that expresses a number as its prime factors, so you won't have to figure out what 3^24th is, etc. (in that case, it'll be "24;" more on that later). This makes a lot of sense if you want to quickly suss out the consonance between various intervals - you can just subtract the powers of the primes. Of course, negatives and other real numbers are allowed, ;.5 is the square root of two, for example. Polyharps' parade of roman numeral intervals (I,IV,VI#...) are also supported.

It'll be localizable, (I pity the translators...)

The wavetables (timbres) that are the heart of the synthesis are getting updated: they'll have better resolution, may contain several cycles of waves, and may also allow generating them and importing them from audio copy.

I mentioned the beats in passing, basically, a beat is a bundle of volume, pan, timbre and a passel of evolution parameters.

It'll save voices to Audiobus and also act as an IAA generator.

The instructions won't be one huge page anymore.. I'll see if I can make it so that it exists both as a web page and PDF doc. I may need to write some scripts to help me with that.

So, a lot of work! I'm about 1/4 done with it. So that'd Droneo 2.0

PolyHarp is the other main thing obsessing me. There are some corners of it that need some work still. PolyHarp builds scales out of chords, can act as a guitar, piano, multi chord strummed device with intervals plucked from hell, in 200 voice polyphony, or act as a MIDI controller.AB, IIA, etc. It is about the most powerful autoharp fantasy ever.
I want to see if I can make the internal synthesizer more efficient and powerful and sound better. It's OK so far, but I think it could be better.

Then there are fixes to AUMI, which I want to experiment with some more in terms of a motion flow tracker. fast movement is hard to track right now, I want to make the frame rate something fierce!

So that's what's going on!
-- Hery
00:17:51 - jhhlnet - No comments yet, click here to add one