Crimes of the Near Future






Coming Soon...

I've finished the songs.

Which is to say, I've

(1) Come up with backing tracks, and remixed them...quite a lot. Usually this involves changing one thing at a time until I realise the riff or sound I started with is the one which doesn't belong.

(2) Written a load of (pseudo)lyrics. Eroswings suggested I try doing scat singing - in fact that's more-or-less what I'd planned from the beginning. I come up with a basic vocal melody, and I've written a little program to generate nonsense words, which I can fit to it.

(3) Recorded them. This involves packing my laptop and a crappy old microphone to the garage, where I can make strange vocal noise without anyone overhearing and thinking I'm having a nervous breakdown in Farsi.

(4) Processed them within and inch of their lives. Now, the thing is, I'm a pretty terrible singer. On a good day, I can harmonise to a monotone - which is what I do in recording.

I then run the recording through a compressor, a gate, a denoiser and a pitch corrector - this last to wrench my wandering pitch to 110Hz, aka concert pitch A2. I can then use an offline pitch-shifter to move the pitch of individual syllables around, making a little melody.

Honestly, it would be easier to use a vocoder.

(5) The difficult bit. Arranged my vocals over the music, EQing them so everything's clear, and adding reverb, chorus, tape saturation and whatever effects seem like a good idea at the time.

(6) Taken my five - yes, five - songs, and mixed them into a continuous, er, mix.

Stage (7) will be: Upload the result to youtube.

I'm going to wait a few hours before (7). So I can listen with a fresh pair of ears, do any last minute tweaking, and not rush into deciding this is the final, finished product

After all that, I'm not entirely happy - I made a lot of mistakes, and hopefully learned from them. This EP (11 or 12 minutes) is a practice run, and I reckon I can do the next one much better.

So, at midnight on March 1st, as promised...

Slow


I have five pieces of backing music composed.

I have some lyrics (of a sort) to go over them.

I have the time and equipment to record and mix them.

Unfortunately I'm

(1) Very good at inventing distractions
(2) Dealing with a lot of bureaucratic shit
(3) Suffering from a six-week cold
(4) Thus unable to sing
(5) Far more nerishly excited by the prospect of redesigning my virtual studio for the fifth time in a month than actually using it
(6) Going through my annual period of depression, and
(7) Really, very good at inventing distractions

However, I'm hereby making a solomn promise to my blog that I'll start recording within the next 24 hours.

That's the trouble with having no god to be ashamed in front of, no social circle to be pressured by, and no neurotic need to make my parents proud. The only person pushing me is me.

And I'm not very pushy.


Legs

Two things we tend to believe about computers:
1) Every year they'll get faster, more powerful, and maybe even easier to use. The same for software.
2) Most people will be up to date.

Which is odd, because neither has been true for well over a decade. In fact, they may never have been true.

In the 1970s it was a rule of thumb that every year there would be a new generation of hardware on the market, and it would be ten times the speed of the previous year.

Estimating computation speed has never been an exact science. You can chose CPU-cycles, gigaflops, or MIPS as your yardstick, all giving different results. And in any case, two systems which run at the same speed for one task might have very different results for another.

But as a rule of thumb, it worked. Every year, add a zero to your chosen measure of speed.

In the 1990s, the rule was: Every year, speed doubles. And the unspoken implication was: Speed will continue to double for evermore.

Which in retrospect made no sense at all.

If you imagine a nation where every 30 years the population doubles, then to feed everyone you'd need the production of food to double in the same time.

But if it takes X amount of ingenuity to multiply food productivity by two...it will take X times 2 ingenuity to multiply it by four. Which means every 30 years, you have to be twice as clever. And then twice again.

Like the Red Queen said, you have to run twice as fast just to stay where you are.

Even if you're dumb enough to think you can double scientific progress by doubling the number of scientists - or doubling their wages - it's still not sustainable.

In any case, the implicit assumption would be that scientific progress is a never-ending road, which itself assumes that the universe is infinitely malleable, if only we can get smart enough. Infinitely smart in fact - which I'm pretty sure is a meaningless phrase.

Odd how we can believe obviously false things just by not quite stating them explicitly. Clarity is the enemy of the ideologue.

The world of computing hit the brick wall of unmalleable reality around the year 2000. Unless someone found a way to increase the speed of light, or shrink atoms, electricity couldn't be made to flow faster through wires.

We got hyperthreading - a way to shoehorn two short instructions into the space of one long instruction.

We got the botched implementation of 64-bit processing, which had the big selling point that arithmetic with literally astronomical figures was now slightly easier.

And we got multiple cores - a tacit admission that CPUs couldn't be made faster...and the best we could do was run several in parallel, hoping that those tasks which didn't have to be run in sequence could run concurrently...and the results reintegrated by clever shuffling.

My 64-bit laptop has eight cores. Which is to say it has four hyperthreaded CPUs.

Which leaves only the problem that most software is still 32-bit, and most either can't handle multicore processing, or runs tasks that can't be split into several parallel streams.

The focus in computer development is now away from vainly trying to squeeze ever decreasing efficiencies out of silicon and copper.

The USB3 standard is ten times the speed of USB2.1 - but the new faster data streams still have to be funnelled through the same CPU.

Screens are wider - mine is 17.5 inches. Good for watching widescreen TV - in fact it's rather difficult to buy a small flatscreen plasma monitor for your PC. They can't make them much better, so the new notion of an upgrade is to make them bigger.

Mouses, MIDI keyboards and headsets can now be wireless. Which means you get the slight convenience that wires which occasionally got tangled...now don't. And you can now sit on the other side of the room, listening to your MP3s through bluetooth.

Old, obsolete, slow processors are reincarnated in netbooks and iPhones - where you can do what you did 20 years ago, at the speed of 10 years ago, but mobile, and on a very small screen. And pay for the experience.

It's a similar story with software, and operating systems.

Windows 8 was essentially Windows 7 with a new interface. An incredibly annoying, hard to use, badly thought-out interface, marketed on the bizarre assumptions that (a) everyone had touchscreen technology, (b) everyone wanted to use touchscreen technology, and (c) using a touchscreen to operate software designed to use a mouse...was somehow better.

It also pointlessly rearranged the configuration options, on the grounds that moving around the items in your shelves is the next best thing to getting better items.

Windows 8.1 made the further advance of bringing back some of the useful features which 8 had taken out.

"Bloatware" is the name we give to small, useful programs which acquire large, useless extra features. The alternative to bloating is to repackage the same program with the exact same functions, but with snazzier graphics and call it an upgrade.

Which is probably why our second belief is false. People aren't up to date.

Why use MS Office 2014 when MS Office 2003 does everything you could possibly want? Actually, MS Office '97 was all you needed, and it almost never crashed - but it won't install on your new system.

Why use Reason 7 when Reason 5 runs without a freaking dongle, and everything Reason 7 can do that Reason 5 can't do...you do anyway in Reaper?

More to the point, why install new versions of old Firefox plugins, knowing the new versions cause crashes with your other plugins?

I grudgingly moved from Windows XP to Windows 7, mainly becuase my new 8-core widescren laptop cannot run XP. The BIOS simply doesn't have the option to use the old IDE filing system - it can only use the admittedly better AHCI.

So I'm now running Windows 7 made up to look like XP.

Except half of my XP programs won't run under 7 - in spite of Microsoft's insistence that they can, in 7's "compatibility modes'.

Which means, for those XP programs which don't have a 7 replacement, I'm running them in a virtual machine.

And finally, being up to date is expensive. And most people haven't got much money.

Out of date software that people still use is called "legacy" software, and the dirty open secret is that pretty much everyone is a legacy user.

Like in Blade Runner, the future is old. In the future, we'll all live in the past.


FAWM 2014 Day 5

Five days of rain.

One change of bank account, one bureaucratic screw-up, five telephone calls to fix it, three assurances from three bureaucrats that it's all the fault of a different bureaucrat.

One rhinovirus, four types of pills, three days with a head full of cotton wool and a nose full of concrete.

One laptop virus, one complete reinstall of Windows.

Two backing tracks composed, no lyrics written.

One slight change of plan: I reckon I can do an EP in a month? Five songs?

FAWM 2014 Day 1

Things that can get in the way of your first day of composing:
  • Staying up till 4am the night before, watching consecutive episodes of "Castle" - even though you're coming to hate all the characters
  • Having a cold, a sore throat, and generally feeling a bit meh...when you wake up at midday
  • Headphones developing a dry joint and not working -
  • Realising the TR808-style drums you made yesterday are a bit rubbish -
  • Coming up with a backing track loosely inspired by Depeche Mode's "Just Can't Get Enough"...and then realising you've spent two hours making a bad pastiche of, um,  Depeche Mode's "Just Can't Get Enough".
Nevermind. As someone said, "An expert is someone who's made all the mistakes that can be made in a very narrow field". So I'm several points closer to being an expert. Making mistakes is easy. Recognising them is easy for anyone who isn't an idiot. The hard part is not doing them again.