Predominantly Erroneous (Exohedron nonsense blog)

From now on I'm going to interpret the phrase "Socratic method" as meaning drinking hemlock.

 

At some point this winter I should probably actually turn the heating on, if only to find out how well it works.

 

"Deal with the stuff you actually care about first" also applies to programming, as it turns out.

 

Supposedly we want to learn the laws of physics but really we're only interested in the ones that actually get enforced.

 

When trying to find an entry for the Yankee swap, I realized that my tastes in stuff haven't really changed much since I was ten; the only difference is that now my budget is bigger.

 

The age spread in my office, coupled with the generally irreverent personalities and the overall willingness to split hairs on technical definitions, definitely makes for some fun conversations about what counts as "old".

 

It's good to know that when the server crashes and I can't do work my brain sees fit to supply me with an unending rendition of Strong Bad techno for the entire day.

In unrelated news, today at work the system was down.

 

Did you throw a light switch rave?

 

Unfortunately, that was not what the light switches were installed for.

 

I guess I wouldn't be surprised if, due to some sort of psychological/social conditioning, 0 C and 32 F somehow manage to feel different.

 

I think I would be much happier if I could spend like 10% more time just explaining things to my colleagues at a chalkboard instead of writing stuff down in papers and hoping that they eventually get around to reading it. I mean, yeah, I get much more out of reading papers than listening to people talk, but, um...hmm.

I like the process of giving lectures or discussing things I know about in real time. Even posting on a forum doesn't give quite the same feeling.

 

Would "aphoria" be the right term?

 

I think the main drive behind the design of the conlang I'm sort-of working in is that I was disappointed by how much the grammar of Lojban was tied in to the semantics, in tyat you couldn't actually know if a given sentence was grammatical from looking at the conjugative particles alone; you had to know some information about the roots as well.
And that sort of design isn't terribly strange; we are used to verbs that reauire various numbers of nouns to form complete ideas, or functions whose arities or type need to be declared.

But I wanted a firmer separation between semantics and grammar. So I tried to make a system where every clause had the same structure, at the cost of tree-like sentences and a rather un-natlang-esque semantics, in particular removing the distinction between roots-that-are-verbs and roots-that-are-nouns, so that, for instance, "advise" and "advice" are now "advi"+[verb particle] and "advi+[noun particle], with a single semantic object "advi" that doesn't have a specific part-of-speech built into its declaration.
So now the process of determining geammaticality only requires locating the particles required for each clause and checking them against the implied tree structure; the roots can be safely ignored.

I still have issues with how Lojban handles sentence parsing, but that's a different, if related, issue.

 

The fact that actually running any experiments to test the hypothesis would be highly unethical doesn't provide evidence for or against the hypothesis, nor does it strengthen any other evidence for or against the hypothesis.
Even if the ethical status of the experiment depends on the truth or falsity of the hypothesis, that does not provide evidence for or against nor strengthen any other evidence for or against.

 

Probably not a good sign when your harddrive starts purring.

 

And now to never touch this game again.

 

And yet more tumblr discourse that makes me glad that my tumblr is inactive, because I am fully cognizant that my opinions on identity and labels and all that fun stuff are pretty much asking for Strongly Worded Letters from a variety of Concerned Citizens. I mean, sure, there is very little chance that I would gain enough of a following to come to such attention, given that other than said opinions my blog would probably be mostly mathematics, but I can't imagine that the possibility is low enough to justify the hassle.

On the other hand, I should revive my blogspot, because I know that I can do LaTeX directly into that.

 

Something something new year.

Anyway, more importantly I want to talk about electronic music making in the context of talking to my dad about it.

For background, my dad mostly listens to classical music, and understands music composition and production in terms of that. He also listens to the Beatles and some jazz, but he's mostly about classical (read: baroque, romantic, whatever the hell neo-classical might mean, that whole branch of the taxonomic tree, whether or not it's actually cladistic or not)

In contrast, I make electronic music as a hobby. A very slow, not-terribly productive hobby in the sense that I don't really finish much, I just tinker until I get sick of the track or I lose the file.

A little while ago my dad wanted to know about electronic music production, because he likes understanding what I do. So I tried to explain about how a lot of electronic music production isn't just about creating melodies and harmonies and such (and indeed, a lot of the basic criticism of electronic music is the lack of or simplicity of the melodic and harmonic structures of electronic pieces, which is not a subject I particularly care to debate because I don't actually care about melody or harmony) but also about controlling a lot of other variables that aren't really variables for acoustic instruments.
In particular I explained about timbre, the qualities of sound that make, for instance, a flute sound different from a violin even when they're ostensibly playing the same note. And while arrangers certainly would consider the merits of playing piece on a violin versus a flute, rarely are they called upon to actually design the instrument to use (note: this is not to imply that all electronic music composers design their instruments from scratch; there is a lot of preset usage even among "good" composers; the point is that this freedom is available, and also for other reasons I might get into later). And so there is another layer of things for the electronic music composer to consider beyond just when to play what notes.

But today I realized that there is yet another thing that electronic music composition does that classical composition does not, or at least doesn't do much: interpretation.
A human playing an instrument usually sounds different from a robot playing the same instrument; we often use words like "warmth" for humans, and "cold" or "mechanical" for robots. And some of this is probably bullshit, but quite a bit of it comes from the fact that humans don't play exactly what's written; they interpret it, they play with some level of irregularity, slightly off time, slightly off pitch, some tremolo or vibrato, not every note as loud or as accented as the others. It is this non-uniformity that creates that sense of organic-ness to human playing, or at least good human playing. It isn't necessarily sloppiness so much as it is understanding that absolute precision is less interesting and less attractive to most listeners.
But classical composers don't usually write in the music "play this note slightly late, play this note slightly sharp." They'll certainly say "play this C note sharp by half a tone, as in C#" but not in terms of cents. There are dynamic markings, sure, but only like six standard ones, and anything outside of that is generally considered at least faintly ridiculous; moreover, they are certainly not precise in any sense (how many decibels is ff? And then how many is fff? fffff?)
In contrast, when you make an electronic composition, you are getting a robot (well, a computer) to play the notes for you, and it will do what you tell it to do. So if you want anything that resembles "interpretation", you have to put it in yourself. And so you do specify, in cents, if your tone is off by some amount, or at least the computer views it that way, even if you're just wobbling the pitch bend with your shaky, analog human fingers.
I've definitely had to spend time adjusting the velocity of notes up and down by gradations much finer than ff versus fff, whatever that may mean, because whatever the instrument is feeding in to reacts very nonlinearly. The worst bit is when I'm dealing with a sidechain, because the thing that's triggering the sidechaining probably isn't even audible as an instrument; its only purpose is to yank on the sidechain to cause the volume of other instruments to go up and down.

Anyway, the whole point of this blather is that I wanted to write this down before I forget so I can tell my dad. But I might actually start writing a series of posts about electronic music composition, and possibly resurrect/reconstitute my dead G+ post about impact of Vocaloid technology from a producer standpoint and the mirroring cultural phenomenon.

 

Messages hidden as Morse code in the metrical patterns of poetry.

 

Musings On Electronic Music Production: Part 0

I think I want to actually gather my thoughts on electronic music production, both the ideas that I've managed to accumulate over the years on how to produce it, and how it differs from what one might consider "traditional" musical composition and production.

Partly this grew out of discussions with my father, who likes classical (Baroque and Romantic and neo-Classical) music but doesn't understand or appreciate most electronic music, and probably doesn't really recognize his appreciation for the electronic pieces he does appreciate. Partly this grew out of discussions I had about a decade ago on a forum devoted to electronic music, wherein I and several other hobbyist electronic music producers tried to understand what the hell we were doing. Partly this grew out of discussions I had on the now defunct Google Plus with people who were getting into electronic music from a variety of viewpoints, some of them uninterested in the music itself but rather in the production techniques thereof.

Regardless, I want to talk about electronic music.

Having learned a bit from the mathematics thread, I am making this post here so that I will have a table of contents that I can update without actually having to displace any actual content. I am hoping to cover the following topics:

Sound design
-Intro
-Oscillators versus samples
-Effects
--Volume, Filters, and Envelopes
--Delay, Reverb, Flangers, and Phasers
--Distortion
--Vocoders
--Misc?
-Timbre as an arrangement choice
-Timbre as a variable
--How to play the same note over and over for three minutes and still make something interesting

Mechanistic sound
-"It sounds so cold and lifeless"
--The horror and sorrow of existing in Detroit
-Interpretation, and the faking thereof
-Repetition
-Repetition
-Repetition
--How to play the same note over and over for thirty minutes and still make something interesting

Modular Synthesis
-Yes, this deserves its own section
-I think I'm going to stick FM synthesis here as well

Performance
-What is up with those button pads
-Is record player an instrument?
-Live PA

Hatsune Miku
-The problems of simulating the human voice
-Vocaloid technology as vocal-talent-analogue
-Inverted popstar performances

If other topics come to mind I might talk about them as well.

To note: this will not be a guide to actually making electronic music. There are plenty of people who can tell you which dials to press and which buttons to turn, but I have largely given up on that endeavor after realizing that after all these years I still don't really know what I'm doing, or for that matter what any of these settings are doing. I'm here for theory and observations and brainspew; if you want practicals, I can but won't help you.
That's also why I'm not making this into a thread of its own: I'm here to chew gum and ramble about and around but not actually through anything useful, and I'm not terribly fond of gum.