little devil (beta)

Live looper with loop length equal to length of a pressed rec key

(8 Votes)
2.0 (Updated 12 years ago)
18.0kB
December 28, 2006
Reaktor 5 or lower
Instrument Sampler

DESCRIPTION



After you open instrument you need to load 2 independent samples of the same length to left and right table (i recon 60 second will do ).

How:

-- keys --

c#5 - hit this key if you want only source input ,now you can swich again to loop playback by pressing F#5, G5, A5(with efx - D#5) or start record new loop with (F5+E5)already recorded material or without (F5)

D#5 - apply efx (work only for recorded loop )

------------------------------------------------------------------------------------
INSIDE INSTRUMENT THERE IS A MACRO CALLED FX INSERT , YOU CAN PLACE YOUR EFX THERE
------------------------------------------------------------------------------------

E5 - mix recorded loop with audio IN (if you press record button now , you will record source input and already recorded stuff , because delay and efx are after table and before feedback , if you turn them on you will record processed loop + unprocessed source input

F5- start recording(loop length is equal to length of pressed rec key , not in declick mode prior to zero point detection but in very many cases no noticeable )

F#5 - stutter playback of source input(pulse osc is used , waveform values : 1 - record 0 - playback) , use G#5 to sync osc to audio source input and then press and hold F#5 use with xtap knob and SQW .

G5 - loop restart(working for rec and play)

G#5 - tap it few times in tempo you want (hold time is discarded , only transition from 0 to 1 is detected)

A5 - playback direction



--- knobs , display ---

delay- add max of 40ms of delay to recorded loop

loop div - divide recorded loop , leaving first segment

loop len - display loop length in seconds

xtap - multiply tap tempo

sq w - that is width of pulse osc used for stutter playback

tapBPM - well.......(before xtap in structure)

stretch - stretch loop (don't affect record clock), minimum = no strech which is default

stretch Spd - speed of stretch osc

play spd - speed of playback , if you change speed to lets say 25% of original speed (max) and record , you will end with loop recorded at lower sample rate which can be desired in some scenarios( if it is not a bug then it is a feature :) )

sample rate - shows current speed recording sample rate

input vol - default 1 X incoming signal

de-click - on /off

A - offset for zero value sample at the begining of loop, values of 50-250 should do
B - offset for last recorded sample which is the same as first one plus offset , values of 50-250 should do

with lower values of A and B sometimes you might have a problem with proper key press sync , with bigger values you get more noticeable clicks obviously,

DE-CLICKER -- after you press rec button (and hold it) it will not record anything until it detects a signal with value of 0 + offset (knob A) on both channels , when you relase rec button it will record until signal with the value of first loop sample plus additional offset (knob B) occurs on both channels ,



about de-click alghoritms:

i have try envelope like click prevention with bad results (will try it again in different ways),

i've also build alghoritm that store first recorded value(at key press) then store last recorded value (after key relase) then it count difference(DI) between those two , divide result by let's say 24 and record additional 24 samples with values going from last value(LV) to first value(FV) of a loop using 1/DI value as a single step : LV + 1/DI = LV1 , LV1 + 1/DI = LV2 and so on until you record all 24 values which mean that you reach first sample value(FV), then playback of a loop begins , so you end with line going from last sample to first sample (24 samples long - 0,5 milisecond) , click is noticeable but is more filtered,

the current alghoritm is good ,but it have one weak POINT - works bad with very low tones


P.S: there are minor bugs , no gui and everything was made in AUDIO domain (yes i know....) , no optimized yet (it is not hungry at all), internal mess......apart from this it's working really well ;)

Bye

COMMENTS  (1)

James Black
12 years ago
keep it comin'. Really looking forward to what you turn this into. No unwanted clicks:)
now