FREQTIKIMPULSE ANVIL DOCUMENTATION
ProductDownload
FreQtik / Impulse Anvil Docs / Found Sounds
Impulse Anvil 1.0.122Windows VST3

Found Sounds as IR Material

A recording does not need to be a measured room response to become useful convolution material.

The easiest way to understand this is simply to try it.

Good first recordings

Record something with a clear character:

  • metal, glass or ceramic hits;
  • cloth, paper or leaves;
  • doors, tools and mechanical objects;
  • water, gravel or footsteps;
  • a clap or snap in an unusual room;
  • a breath or short vocal sound;
  • a short noise or texture.

Build it

  1. Load the recording into A.
  2. Click A so you are hearing it.
  3. Use Start/End to isolate one interesting region.
  4. Link Start and End if you want to scan a fixed-size window through the recording.
  5. Fade the edges if the cut feels too abrupt.
  6. Try Color or Texture.
  7. Use EQ only if the material needs cleanup.
  8. Bake several promising versions.

Listen for the small moments

A long recording may sound obviously like "a recording."

A tiny part of the same file can behave like something completely different:

  • a resonant filter;
  • a strange cabinet;
  • a sharp transient space;
  • a metallic tonal fingerprint;
  • a tiny impossible room.

That is why EDIT is so useful with found sounds.

Scan instead of guessing

Create a short Start/End window and link it.

Then move the window through the recording while audio plays.

  • Shift + drag = fine
  • Ctrl + drag = very fine

When a position suddenly does something interesting to the track, stop and Bake it.

Push it further

Take one good fragment and:

  • phase-color it with Color;
  • Morph it with a normal room IR;
  • Bake it and use it again;
  • create several different fragments from the same recording and Morph those together.

Capturing a resonant object?

For guitars, violins, cabinets and other resonant bodies, continue with Acoustic Bodies & Comparison. It explains what a body IR can transfer, how to compare two matched captures and why one IR should not be treated as a complete physical clone of the instrument.