theaugy/cdj_mode

bash script for preparing tracks for CDJs, Serato, and any other performance system with continuous playback

Introduction

cdj_mode is a set of bash functions that prepare music for DJ'ing. A input file (anything that can be read by ffmpeg) is converted to a wav file. That wav file is split into pieces. The bpm of the wav file is established (tap tempo or specified). The pieces are copied to any number of target directories.

Intended Use Case

Over many years of DJ'ing, a hard learned lesson is that there is very little you can rely on with respect to any club DJ setup. You cannot expect CDJs that support rekordbox. You cannot expect CDJs with hot cue buttons or particular looping features. You cannot expect that the link jacks will work. Hell, sometimes they aren't even Pioneers. Or it's only Serato.

Having a system for organizing, prepping, and performing that is completely agnostic of any performance technology, while maintaining the most basic features a DJ needs like BPM information, cue points, and library organization, is the problem that cdj_mode was intended to solve.

cdj_mode will work with any performance setup that supports:

• USB media (FAT file system)
• Gapless playback

This is known to include CDJs (turn off autocue), Serato, "other" brand CDJs, Traktor and Mixxx and probably any other "in the box" DJ performance software.

Prerequisites

• sox
• ffmpeg
• tapbpm

Configuration

cdj_mode will copy the prepared tracks to multiple target directories. I will plug in both of my performance USB sticks, mount them at /mnt/rb3 and /mnt/rb4, and ensure that cdj_target_dir is /mnt/rb3,/mnt/rb4. To change these, find the line near the top of the cdj_mode file and update the cdj_target_dir= line to point at the directory or directories you want. Comma-separate multiple directories.

Workflow

This is intended to be run from a bash shell. For macOS, Linux and similar operating systems, this is easy. Windows users should consider Windows Subsystem for Linux.

First, put your shell into music mode:

$ source path/to/cdj_mode/cdj_mode

Load a track (called some_song.mp3 in this case)

$ cdj_track music/some_song.mp3

Set the bpm (120 in this case):

$ cdj_bpm 120

Find your cue points by scrubbing to a point in the song (28 seconds in this example):

$ cdj_scrub 28

This will play 1 second of the song starting at the offset. If you're off by a little bit, you can nudge it forward or backward by changing the offset. You can use decimal places to get it exactly right (you rarely need more than the first decimal place):

$ cdj_scrub 28.14

If you need to hear more than 1 second of context, pass the number of seconds to play as the second argument (we'll listen to 5 seconds to be sure we got it right):

$ cdj_scrub 28.14 5

When it sounds right, add a split:

$ cdj_split

You can scrub anywhere else and add another split. You can add any number of splits. Don't worry about the order; when you finalize the track, your splits will always be created in chronological order.

To hear all the splits you've created so far:

$ cdj_split_preview

When you're done with the track:

$ cdj_done

cdj_done will break the track into pieces and copy it to the target directories. So if your target directory is /mnt/cdj_usb_key, and you created splits at 28 and 64, and the bpm is 120, then cdj_done will generate the following files:

/mnt/cdj_usb_key/120.00_some_song [000].wav
/mnt/cdj_usb_key/120.00_some_song [028].wav
/mnt/cdj_usb_key/120.00_some_song [120].wav

Each piece starts at the exact offset where you put its split.