For those of you who have been reluctant to give Spoke a spin, a limited free version is now out on the App Store. Give it a whirl, give us some feedback!
We have some new features in the works as well as some optimizations coming up. Apparently Spoke actually runs slower on OS3 than on OS2.2.1, and this is something we are working hard to diagnose and address.
UPDATE: We tracked down the issue with OS3, and Spoke will soon be running much faster on an iDevice near you.
For the developers out there: using NSTimer to frequently trigger a UI update on the main thread is a bad idea. For some reason detaching a new thread that sleeps often yet still performs the UI update on the main thread leads to huge performance gains. Anyone know what’s going on under the hood that makes this so?
Strange?
Foremost, a most hearfelt thanks to Parvin Panahi for turning lead into gold. Platitudes and platforms aside, design gives shape to notion.
Prison was hard, but we’ve moved on and dragged along with us just the darnedest little beat sketchpad we could have ever imagined. No grids, no top, bottomless, faithful: Spoke groove machine delivers.
Our thesis:
The drum machine began with analog sounds, but its brain was always digital; the step-sequencer places voltage potentials on a grid.
Analog synthesis was, in truth, merely a convenience; it did the job adequately. Analog is great for warmth and richness, but analog is shit for timing. So-called analog sequencers are, at their best, digital with regard to the time axis. True analog was never feasible for timing.
Contemporary sequencers have extremely high resolution; Logic has a resolution of 960 ppqn. Theoretically, Spoke places beats at sampling resolution. At 120 bpm that is 22050 ppqn. A frequency of 22050 Hz is sufficient to reproduce audio at the threshold of our hearing range. What is the threshold of our timing range?
At 22050 ppqn we are surely beyond the limit of human timing sensitivity. At 22050 ppqn position becomes effectively analog.
This is what Spoke seeks: analog positioning on a digital platform.
Of course we are limited by resolution. Of course this limit is arbitrary. Of course this limit is transitory. Of course this limit need not go very high before we supercede human timing accuracy.
Nonetheless this is the spirit: do away with the position grid, do away with the quantize. Place the beats where they need to be. Place the beats where they sound most true.
Strange?
