Skip to content

Automatic Dropout Restoration

HENDRIX-ZT2 edited this page Jul 8, 2019 · 3 revisions

This tools tries to improve tape sources that have many and strong dropouts. It is not useful for digital (full) dropouts.

GUI preview

dropouts before - after

There are two modes / algorithms available, each with different benefits and limitations.

  1. Heuristic Mode

This mode tries to detect (ie. guess) dropouts from variation in the audio's volume curve. Choosing suitable parameters is crucial. It works in multiple frequency bands to correct dropouts of varying intensity and frequency skew. Only the highest frequency band operates freely, the ones below are tied to it to some extent. This is because dropouts are often most prominent in high frequency bands and you want to avoid fixing pseudo-dropouts in the lower bands. Max Width is the area over which detected dropouts are smoothed and does not affect the detection. Max Slope discards dropout candidates situated around transients by limiting the allowed dB increase across the dropout.

  1. Max Mono

This mode removes dropouts from mono sources transferred as stereo files. It takes the maximum for each spectral bin between left and right channel. This is very effective for dropouts or tape damage that are not the same on both channels. Matching azimuth / phase alignment between channels before running this is essential for a good result; matching volume and EQ is recommended but not as important. Remaining dropouts can be treated with the Heuristic mode.

Clone this wiki locally