audioeffects.airspace

AirSpace - a tape echo sent into a chamber, with the dry beside it.

You send your signal to a tape echo whose repeats feed an echo chamber, and your dry signal comes back exactly as it went in: the filters, the drive and the tape act only on the send.

Time is the gap between repeats and Feedback how much of each repeat goes round again. Loop Top and Loop Bass filter the loop, so each repeat is darker than the last, and Drive saturates it. Top Cut darkens everything sent to the chamber, once. Low Cut, Decay and Damping set the chamber. Mix is audiodelays.Echo’s: the dry at unity up to 1, the chamber alone at 2. The Effect controls page lists what the knobs span and the patches.

More Feedback makes the chamber’s tail longer: at Time 500 ms, Feedback 0.85 at least triples it. At the defaults you hear the repeats as a longer chamber, not as a rhythm of echoes.

Move Time while it plays and the repeats bend in pitch, the way tape does: down while it lengthens, up while it shortens. Glide is how long a move across the whole Time range takes, so it sets how far they bend: 12 % at the default of 4 s.

Keep your input at or under -17 dBFS: at Mix 1 the dry and the chamber are both at unity and add, so a hotter input can reach the rail. That holds on every patch and with any one macro at either end. Open two or more of Feedback, Decay and Time together and the chamber builds further, by up to 25 dB: keep your input at or under -26 dBFS there. Latency is zero. reset() empties the tape and the chamber and restores patch 0, whose Time of 111 ms lands at once instead of gliding.

Limits

A control that jumps makes the output step: move it in small steps from the host if you need it smooth. Time is the exception: it glides.

A patch that changes Time glides there too, and the repeats bend while it travels.

At the defaults the repeats add only a little length to the tail; the stretch comes with Feedback above about 0.6 or a long Time.

At a short Time with high Feedback the repeats are heard as a flutter, not a longer chamber.

There is always at least 20 ms between the dry and the chamber.

The loop’s filters are gentle, 6 dB per octave, and the Top Cut’s is steeper: there is no loop bell, no presence EQ and no wow. Drive’s level is not compensated. The chamber’s size is fixed, there is no tempo sync, and a mono source gets a mono rack.

At Mix 0 the output is your input, byte for byte, and the tail that was ringing when Mix went to 0 is gone when it comes back.

When your source ends, the tail rings out as it would on silence.

audioeffects.airspace.BLOCK = 256

the palette’s block.

Type:

Frames per pull

audioeffects.airspace.LINE_MS = 505.0

Time’s top plus headroom for the node’s two-frame clamp (audiodsp_feedback_delay.c:156-159).

Type:

The tape’s line

audioeffects.airspace.FULL_RANGE_MS = 480.0

Glide is the time it takes, so delay_slew is this over Glide.

Type:

How far a move across the whole Time range runs, in ms

audioeffects.airspace.LOOP_BASS_LAP_FACTOR = 2

tail_samples counts this many times laps_to_zero’s laps, as DigitalDelay does with Repeat Cut in (CUT_LAP_FACTOR). Loop Bass is a high-pass in the loop, always in circuit: it sends a lap’s value less the mean the loop has been carrying, which can be larger than any value on the line, so the per-lap argument (each lap’s peak below the last) does not hold, and the factor is measured. On the tape’s own output, at Feedback just under each step of laps_to_zero’s count from 0.004 to 0.85 and in steps of 0.01 from 0.40, Loop Bass 20 to 1000 Hz, Time 20 to 500 ms, 48 and 22.05 kHz, on full-scale noise, random full-scale signs and clicks, the longest tail took 1.49 times the count with each lap lengthened by both one-poles’ memory (Feedback 0.0039, Loop Bass 1 kHz, Loop Top 16 kHz, Time 500 ms: three laps against two), 1.11 at Feedback 0.4999 (17 laps against 15) and 1.02 at 0.5312; the more laps the count holds, the less the high-pass adds. Mono, Drive and a lower Loop Top measured no longer. Twice the count leaves a third again over the worst.

audioeffects.airspace.DEFAULTS = (111.0, 0.375, 4000.0, 0.2, 6000.0, 100.0, 5000.0, 100.0, 1.8, 4000.0, 0.35)

The constructor defaults, in macro order; NaN falls back to these.

audioeffects.airspace.DEFAULT_STEPS = (68, 56, 76, 25, 82, 52, 74, 64, 65, 76, 22)

patch 0. While a macro that arrived as a whole step (the seed, a patch, a whole set_macro) sits on its default’s step, the default itself plays, so a fresh instance, patch 0 and a host echoing get_macro all play exactly 111 ms, not the step’s 112.08. A constructor value that is not the default is not a step and plays as given, even where it lands on one.

Type:

Each default’s step on the 0-127 grid (_component.macro_of)

audioeffects.airspace.STEP_TOLERANCE = 1e-06

How close a 0..1 position must come to a default’s step to play the default: far under one grid step (1/127), far over a float’s rounding.

audioeffects.airspace.REVERB_DECAY_I = 1

The chamber’s indices in Reverb’s macro table that this rack drives.

audioeffects.airspace.CHAMBER_PATCH = (85, 65, 64, 0, 99, 76, 127, 64, 19, 71, 102, 64, 127)

Character 85 (chamber), Decay 65, Size 64 (1.004), Predelay 0 (the tape is the predelay), Diffusion 99 (0.70), Damping 76, Bandwidth 127 (Top Cut is the band), Low Cut 64, Mod Depth 19 (0.30 ms), Mod Rate 71 (0.89 Hz), Width 102 (0.80), Tone 64 (0 dB), Mix 127 (the tail alone).

Type:

The chamber at this rack’s defaults on Reverb’s grid

audioeffects.airspace.slew_of(glide_ms)[source]

delay_slew for a Glide: the whole Time range over Glide.

class audioeffects.airspace.AirSpace(source, *options, **keywords)[source]

Bases: Component

A tape echo sent into a chamber. The module docstring is the player’s page.

NAME = 'AirSpace'

The stable provider name audioeffects.create() discovers.

TIER = 'audiodsp'

STOCK or AUDIODSP, with REQUIRES naming the audiodsp-own modules an AUDIODSP class needs.

Type:

roadmap §3

CAPABILITIES = ()

The contract’s static reads. CAPABILITIES names "tempo_sync" if and only if the class reads self._transport().

MACRO_LABELS = ('Time', 'Feedback', 'Glide', 'Drive', 'Loop Top', 'Loop Bass', 'Top Cut', 'Low Cut', 'Decay', 'Damping', 'Mix')

Knob names, one per macro index; one mode per label; the private engineering spans behind them; and the patches on the 0-127 grid.

PATCHES = {0: ('Chamber Send', (68, 56, 76, 25, 82, 52, 74, 64, 65, 76, 22)), 1: ('Short Send', (43, 45, 76, 25, 82, 52, 82, 64, 50, 76, 22)), 2: ('Long Send', (107, 90, 76, 25, 82, 52, 74, 64, 77, 76, 22)), 3: ('Dark Loop', (68, 127, 76, 25, 32, 75, 74, 64, 65, 76, 22)), 4: ('Bright Chamber', (68, 56, 76, 25, 82, 52, 114, 43, 72, 110, 22)), 5: ('Driven Tape', (68, 75, 76, 102, 64, 52, 74, 64, 65, 76, 22)), 6: ('Plain Chamber', (0, 0, 76, 25, 82, 52, 74, 64, 65, 76, 22))}

_component.macro_of of the card’s settings; patch 0 is the constructor’s defaults on the grid (DEFAULT_STEPS).

macro(index)[source]

What macro index currently stands for, in its own units: what plays, the default itself on the default’s step.

set_macro(index, value, channel=0, note_id=-1, sample_position=0)[source]

Set macro index from the 0-127 scale. A whole step is a step and plays the default on the default’s step; a fraction is a value of its own. A call that leaves the position where it is (a host echoing get_macro) changes neither.

reset()[source]

Empty the tape and the chamber, restore patch 0, and play from there as a new instance does: the graph is wired again the way the constructor wires it, so its levels land at once rather than walking from the ones the knobs had left.

property output

The node the host plays. Asking for it is the host’s first step to pulling, so from here Mix’s first move off 0 walks in.

property tail_samples

Frames until the output is exactly zero once the input stops, as an upper bound: the tape’s repeats, the Top Cut’s ring and the chamber’s own tail, end to end. The repeats are counted twice over (LOOP_BASS_LAP_FACTOR), a margin measured rather than derived.