New sound application

Post Reply
Posts: 243
Joined: Thu Dec 05, 2013 4:56 am

New sound application

Post by Koekepan » Thu Dec 04, 2014 11:59 am

I have determined that there are very few good tools for composers to plan music.

For those who are familiar with musical notation, that is an option, but even so there are some weaknesses to that scheme and it is very opaque to people who do not understand it.

There are no good tools for people who want to use xenharmonics. Or if there are, I haven't found them in well over a year of looking.

After many months of thinking hard about how such a tool would look, I recently had a breakthrough and discovered that a lot of people have been asking for it, mostly without being able to explain why. Here I'll describe what it looks like, and even give some ideas about how a current tool could achieve that.

The basic core approach is the pianoroll. Why? Because it has certain key valuable elements:
  • Every note's relation to the height of every other note is visible at a glance.
  • Every note's duration and position on the timeline is similarly visible at a glance.
  • With the use of intensity/saturation/contrast, every note's volume is visible at a glance.
What's more, with a few slight modifications from the bog standard pianoroll, it can be more powerful:
  • Allow lines to bend up or down, and you have portamento/slides.
  • Allow lines to fade in and out, and you have envelopes, volume automation, crescendo and decrescendo control.
  • Allow lines to snap to grid for quantisation, but then allow those grids to change along a timeline, and you have tempo control for your song, as well as time signature control.
  • Allow the vertical grid to be modified, perhaps even during a song, and you can change from snapping to the regular western 12TET scale to renaissance era just intonation, or 17TET classical arabic, or a strict diatonic or pentatonic or any other scale. Xenharmonics become practical, even approachable.
All this may sound really crazy, but it strikes me that it should be feasible as a fork of VirtualANS. It practically is a pianoroll already, give or take some details around quantisation.

The idea is not to make a full scale synth tool out of it (although some prepared sounds would be convenient) so much as a sketchbook for the kinds of composers who plan their work in detail. The grid lets composers plan intervals and even complex chords and progressions. Letting the grid's structure change within the piece allows for complex time signatures, key signatures or even complete scale shifts. A composer would use it to hammer out a structure, and then turn to some other tool (Sunvox, or a full scale DAW) for production. It's like the electronic equivalent of a music notebook, or a guitar for hammering out chords and harmonies.

And, NightRadio, if you did it for real scale changes? You'd basically have the total worship of the xenharmonic community. You would undergo apotheosis. There is no credible competition. Or at least, if it exists, it's the world's best kept secret.

So, I propose PixiRoll. A scale line on the left, which is modifiable either by the song or ideally within the song. Timeline along the top or bottom. A small sound selection as per pixitracker, or the VirtualANS palette. Different instruments have different icons/colours. Snap to grid, but allow for bending. Also allow for volume shifts. I wouldn't necessarily bother with stereo, because this is more of a planning tool than a true production tool.

Maybe I'm completely underestimating how difficult all this is, but I think it's well within the kind of thing you've already done. One thing I know I'm not underestimating is how fervently the insane xenharmonic underground wants this. And it would be convenient for regular old composers who just want to plan their 12TET pieces, or switch between 12TET and just intonation.

Posts: 243
Joined: Thu Dec 05, 2013 4:56 am

Re: New sound application

Post by Koekepan » Fri Dec 05, 2014 12:24 am

Today I took another very careful look at VirtualANS, and I think that I understand how close it is to the best answer.

Modify VirtualANS to allow for:
  • Snap to grid for quantised output
  • Modifiable grid for timing
  • Modifiable grid for tuning
  • Don't wipe earlier materials when modifying the grid, so you can switch between timings and tunings in one sketch
  • Allow users to define, or load predefined scales
  • Allow the grid to be extended as necessary for a song or snippet
  • Allow parallel grids for multiple parts, which can then be merged into a single work, like a mixdown
None of that should take anything away from VirtualANS, and the rich capabilities of the VirtualANS would stay effective.

One could write a symphony in it, using renaissance tunings, and suddenly make the third movement change to classic raga tunings. And back. Why not?

I know I'm crazy, and I know I represent a minority of lunatic composers, but on behalf of the lunatic fringe I'm making my pitch.

Posts: 243
Joined: Thu Dec 05, 2013 4:56 am

Re: New sound application

Post by Koekepan » Sat Dec 06, 2014 11:57 pm

OK,I know I keep replying to my own thread, but bear with me...

Try the following experiment in VirtualANS:

Play a conventional minor trichord on your favourite MIDI keyboard:
acF (A sharp, C sharp, F natural).
Sounds OK, doesn't it? I did mine around octave 5, so my A sharp was around 584.5Hz. My C sharp was around 693.2Hz, and my F at 873.8Hz.

Now add the fourth to the tonic, minor third and fifth: acdF.
It sounds much less great. The fourth pulls against the third and the fifth unpleasantly, leading to a sound which is not merely harmonically unbalanced, but actively discordant. The fourth is at 783Hz.

Now advance the experiment a little: create a minor chord based on particular ratios, or as close as virtualANS will let you get:
600 Hz - 700 Hz - 900Hz. I get 598.9Hz, 701.7Hz and 895.3Hz.
Not too bad, really. Add a fourth, at 802.4Hz, and now you have the same chord I mentioned above.

Now play the just intonation (approximated) chord with the added fourth alongside the 12TET chord (from your MIDI keyboard). The just intonation chord is substantially less discordant than the 12TET chord.

In both cases the fifth is very close to true. The difference lies in the minor third and the fourth. The minor third in just intonation is flatter with respect to the tonic than the 12TET third is, meaning that it is further from the fourth (which is very close in both tunings). In other words, harmony meaningfully changes with seemingly minor changes in tuning.

And THIS is why tunings are important!

El Nino
Posts: 298
Joined: Thu Mar 07, 2013 11:54 am

Re: New sound application

Post by El Nino » Sun Jan 04, 2015 10:57 pm

maybe you would be interested in using this instrument for midi input:

Posts: 243
Joined: Thu Dec 05, 2013 4:56 am

Re: New sound application

Post by Koekepan » Mon Jan 05, 2015 9:58 pm

Very impressive tool.

Still, I think that (feel free to disagree) the software approach is the more flexible one here.

Especially for precomposition.

Posts: 34
Joined: Tue May 12, 2015 11:31 pm
Location: Ferguson, Missouri, USA

Re: New sound application

Post by rainydayshirts » Mon Jan 25, 2016 3:06 am

YESSS! This propposed tool would be revelutionary!
Jesse R. Schilling | rainydayshirts

| Bandcamp | Deviantart | Patreon |

Post Reply