I created this tool because I couldn't find a simple way to create
Acoustic guitar IR profiles. If you find this useful, please consider
supporting its ongoing development:
Buy me a coffee. Github source:
here.
Source Audio
Drop Pickup WAV here or click to browse
Drop Mic WAV here or click to browse
Configuration
Initializing...
Generation Complete
Preview Output
Source Audio Guide
The algorithm compares the Pickup signal to this Mic signal and
generates an Impulse Response that makes the Pickup sound like the
Mic. Record both signals at the same time into the DAW of your choice
(Garageband, Audacity, etc.) or a multi-track recorder like a Zoom H5.
Ensure both files are max volume without peaking, time-aligned, and
from the same performance for best results. Output to mono wav files.
45-60 seconds is plenty of audio for this process.
Pickup WAV
This is the sterile, direct audio signal recorded straight from your
acoustic guitar's built-in (usually piezo) pickup. This is the
baseline sound that the algorithm will learn to "convert".
Mic WAV
This is the target "studio" sound recorded with a real microphone
placed in front of your acoustic guitar. A nice dynamic mic like a
Shure SM57 is recommended for best results. In my experience,
condensers can be too boomy in the mid range and prone to feedback.
Configuration Guide
Algorithm
Standard (Original): The original algorithm that
generates a single, polished IR. It uses a chunking method to reject
outliers and applies a final Graphic EQ matching process.
Cuki's Algorithm: An alternative algorithm based on
Cuki's open-source project. It returns a zip file containing multiple IR variations (Standard,
M-file, Blended) so you can choose the one that works best for your
instrument. The "M" is Cuki's modified approach to the original
algorithm and is often preferred for acoustic guitar IRs.
IR Length (Samples)
Determines the length of the generated impulse response.
2048 represents roughly 46 milliseconds of audio and
is the sweet spot for avoiding unnatural low-end ringing caused by
mathematically forcing the IR to match the microphone's Graphic EQ.
Increasing length (e.g. 4096) allows for finer detail but can
introduce muddy bass artifacts.
Smoothing (Fractional Octave)
Smoothes the microscopic mathematical peaks and valleys in the
frequency response before generating the final audio. A setting of
1/3 mathematically averages adjacent frequencies,
preventing a spiky "comb filter" effect in the high end, resulting in
a more natural, wooden acoustic tone. Lowering this value will leave
more of those sharp, resonant microscopic peaks intact (potentially
introducing harsh highs), while raising this value will more
aggressively average the frequencies, potentially dulling the unique
character and definition of the acoustic tone.
