Concept: The delay time is set by an envelope follower that measures the time between attacks and sets the delay to that time or a multiplier of it.

Use a signal with clear attacks for the trigger. Drums and clean guitar work well. If you want a fuzzy guitar, run the clean signal to the trigger and the processed signal to the audio ins. The trigger does not need to be the same instrument as the audio, so for example, drums can control the delay times on a pad (aka sidechaining).

This .bgrp, operating details and revision history are on GitHub in repo rcrath/rhythmecho

revision history

0.8.3 minor tweaks, streamlining, a few more presets

0.8.2 begun 7/21/22 to do:

• rewrite presets
• experiment with restoring positive attack only as an option. up delay buffer to 1,000,000 samples

done

• remove attack clicks when inertia is off using small delay and an negative envelope.
• replace "attack" and "release" with "swell"
• Add internal filters

0.8.1 begun 7/19/22

skipped for bookkeeping reasons :)

0.8.0 begun 6/30/22 2DO:

• Added abs and rel inertia types
• added various inertia algorithms. These two things make inertia a lot more interesting.
• removed 1.5 second limit on delay
• changed delay time variable to int variable
• made envelope sensor into a small group and linked the left params to the right params for advanced sensor adjustments.
• remove exponential env, replace with neg. attack-only envelope. For click removal. Low settings give a funky distortion.
• changed panning panning per delay tap with default at full L, full R. These could be fun to automate! change wet/dry vol to a single wet/dry mix.
• Changed Num/Den to drop down menus

0.7.4 and 0.7.5 added presets

0.7.3

restored presets, added new denominators.

0.7.2

• reconfigure delay calculator to provide more stable delay lengths
• insert envelope after delay line to remove clicks
• inertia is set to 0 when auto inertia turned off delay feedback set to exponential curve ( x^.5) 0.7.1 2/22/10
• put num/den contols into main menu
• added new numerator presets 15 and 17 0.7
• put right and left delay controls into main menu 0.6
• groupified the guts
• made it stereo with cross feedback
• added control to bypass inertia...helps with clicks
• UI reorg
• general dusting and cleaning
• Bypass smoother (inertia) if set to 1 ms in order to reduce clicking sounds

0.5

• Put sanitizers in signal path to prevent denormal problems
• put (RhythmEcho) after cc to param to avoid confusion in midi learn

0.4

• separate trigger signal from sample input
• set output of logic gates to PrevX(1) Fx and fed back into gates as false value prevent output of zeros to keep delay times in bounds.
• added autosmooth function so that smoothing (inertia) time matches (or if over 1 sec, is half of) delay time, making smoothing relate to the timing of the music too.
• added monitoring of delay time in ms to gui
• adjusted min and max of gui volumes so that max=1
• changed '<' back to '>' on envelope matching
• added midi cc to params and midi input so that midi mappings stay put for m-Audio Radium61 keyboard.
• took out spectral processing 0.3
• improved wet/dry mix
• moved multiplier from accum input to after minimum time but before maximum time. This allows for very short delays without unwanted combing and keeps preventing delays longer than the delay can handle.
• changed denominator of multiplier from 32 to 64 to allow shorter delays.

0.2

• changed direction of first binary op from "<" to ">".
• added min (4410 samples) and max (65535 samples) delay times to get rid of combing effects.
• changed accum factor from single number to two integer sliders to give fractions from 12/1 to 1/32
• changed mixing from wet/dry signal to single in single out with balance.
• added spectral freq and magnitude gates before envelope to increase sensitivity of beat detection.
• re-mapped midi parameters

0.1

• The beat sensing is not so great right now. It needs to 1)reject times under 50 milliseconds and 2) pick up the transients better.
• added feedback loop and limiter to circuit
• added smoother, which slows down the change in the delay times, creating a swooping effect when set above 1.
• Changed first greater than back to less than: if delayed (i.e., older) time times slope ratio is less than undelayed (present) time there has been a jump in volume.