Bounce and export

To bounce is to turn your song into a sound file you can play anywhere, send or upload.

Bounce the song

Set the cycle around the part you want, or leave it off to bounce the whole song.
Choose File ▸ Bounce… (⌘B).
Choose the format and where to save, then press Bounce.

Order says what is bounced: the Song as it lies, or an arranger chain (see Arrangement ▸ Arranger chains) — the whole chain, in its order, exactly as it plays, without flattening it first. While a chain plays, the dialog starts on that chain, because a bounce is what you hear. A chain is bounced whole, whatever the cycle says.

In the usual Offline mode Nebula renders the song faster than real time, with the very same sound you hear while playing (see Realtime bounce for external gear). When it is done it tells you how loud the file is (LUFS) and its highest peak.

Realtime bounce

Mode in the Bounce dialog says how the song is rendered:

Offline: as fast as the computer allows. This is the default.
Realtime: Nebula plays the song through the audio interface and records it as it plays, so it takes as long as the song.
Automatic: Offline, unless the song plays through something only the audio interface can hear.

A song needs Realtime when it uses external gear: an external insert (an outboard compressor or reverb patched through the interface) or a hardware instrument played over MIDI. An offline render cannot hear them. The line under Mode says what will happen and why, for example that Automatic chose Realtime because of the external insert on the Bass. If you choose Offline for such a song, it tells you to choose Realtime or Automatic.

Formats

WAV and AIFF: full quality, for mastering or another studio. 24-bit is the usual choice.
FLAC and Apple Lossless: full quality in smaller files.
AAC: small files for listening and sharing, from 128 to 320 kbit/s.

You can also choose the sample rate (44.1 kHz for most music) and dither. Dither adds a tiny noise that makes quiet endings smoother when you go to 16-bit. Use it for 16-bit files; leave it off for 24-bit.

Surround bounces

A surround song bounces every channel into one file, in the format's order: L R C LFE, the surrounds, then the heights. Its stems are surround too, so they add back up to the mix.

Choose WAV or AIFF for any format, or Apple Lossless up to 7.1. An AIFF names every layout inside the file; a WAV names it for LCR, 5.0 and 5.1, and keeps the same order for the others.
A surround bounce renders offline, at the song's own speed, and is levelled to a peak rather than to a loudness.
To bounce a stereo file of a surround song, tick Stereo downmix in the Bounce dialog (or ask Claude for "the stereo downmix"). The centre and the surrounds are folded in at −3 dB, and the LFE is left out.

Loudness

Level can make the file a chosen loudness, for example −14 LUFS for streaming, without going over a peak ceiling. Nebula never squashes the sound to get there: if the mix has very sharp peaks, the file comes out quieter, and the report says so.

Stems

Stems are one file for each track or bus, so someone else can mix the song.

Choose which tracks, buses and stacks to bounce, or only the chosen ones.
Choose whether the Stereo Out's plug-ins are on the stems (off by default) and whether the effect returns are included.
Name the files with {song}, {track} and {index}.

Every stem lines up with every other to the sample.

Export Ranges

Export Ranges bounces several parts of the song in one go, each to its own file: the songs of an EP laid end to end, every section of a cue, a radio edit and its intro.

Choose File ▸ Export Ranges….
Under Ranges from, choose Markers (each marker to the next, the last to the song's end), Sections (each section of the Arrangement track), Cycle, or List to type ranges by bar with Add Range.
Tick the ranges to export. Beside each, type its ISRC if it has one. The ISRCs are kept with the song.
Under Formats, tick one or more: WAV, AIFF, FLAC, Apple Lossless, AAC. Every range is written in every format ticked.
Under Each range, choose Mix, or Mix and stems for each range's stems beside its mix.
Under File name, set the pattern. $project is the song's name, $section the marker's or section's name, $index the range's number (01, 02…). The default is "$project - $section - $index", and an example name is shown under it.
Click Export… and choose the folder.

Bit depth, sample rate, level and tail are as set in the Bounce dialog, and each file is a whole bounce with its own tail. A file whose name is already in the folder is numbered rather than replaced, unless you tick Replace files of the same names. The count beside the buttons says how many files will be written.

While it works, the banner shows which file it is on. When it is done, a report lists every file and how it went: for example "5 of 6 files written · 1 failed". Each file is tagged as its own track: the marker's or section's name is its title, its place in the batch is its track number, and it carries its own ISRC.

Claude can do this for you: ask it to bounce every section as WAV and AAC, or each marker to its own file.

Song Info

Song Info is what every bounce of the song says about it. Choose File ▸ Song Info… and fill in Title (the song's name unless you type another), Artist, Album, Track number of how many, Year (four digits), Genre, ISRC (two letters, three letters or digits, then seven digits) and Comment (the tempo and key unless you type another). Click Save; it is saved with the song as one undo step.

Every format carries it in the way its players look for it: WAV as BWF and RIFF tags, AIFF as ID3, FLAC as Vorbis comments, AAC and Apple Lossless as iTunes tags. Claude can fill it in for you: ask it to set the artist, the album or the ISRC.

Album and DDP

The album page puts the songs of a CD in order and makes the DDP image a CD plant presses from: the disc's audio, its PQ codes, its CD-Text and an MD5 of every file. The album is kept with the song you make it in.

Choose File ▸ Album….
Add the tracks. Add Song adds this song from its start to its end. Add Part adds the part a marker or section starts, as Export Ranges sees it, so an EP made in one session can be one album. Add Other Songs… adds other Nebula songs by their files (see below). Add Bounces… adds bounced files of other songs, referenced where they lie and never copied into this one; any sample rate is fine.
Drag a track in the list to change the order, and select one to edit it underneath. Title, Performer and Songwriter are its CD-Text; empty takes the song's title, the part's name or the file's name, and the album's performer. ISRC is the recording's code: for this song and its parts it is the one on Song Info, so every bounce says the same.
Under Before it, set the pause (seconds of silence before the track, 2 by default) or a crossfade (seconds the track overlaps the end of the one before, equal power so the join keeps its loudness; it replaces the pause). The first track always starts after the disc's two-second pregap. Set Gain, Fades in and out, and Trim to cut a bounce's count-in or the silence after it.
At the top, fill in the album's Title, Performer, Songwriter, Composer and UPC/EAN (twelve or thirteen digits; the check digit is checked). CD-Text is Latin-1 on one line: a player cannot show other characters, and the page says which field has them.
Click Measure to measure every track's loudness as it plays. Click Match Loudness to set every track's gain to one loudness, as one undo step: the Loudness target, or the quietest track's when it is empty, so nothing is turned up past its peaks. A track whose true peak would pass the ceiling (−1 dBTP) stops short, and the page says which.
Click Make DDP… and name the DDP's folder.

Nebula renders every track (this song's parts through the one bounce, files converted to 44.1 kHz), lays out the disc, applies the gains, fades and crossfades, and dithers the disc to 16 bits once. Pauses and the pregap are digital silence. It writes DDPID, DDPMS, PQDESCR, CDTEXT.BIN and IMAGE.DAT as DDP 2.00, with CHECKSUM.MD5, then reads the whole fileset back, its MD5s and every PQ entry, before it says it is done. A PQ Sheet.txt beside it lists every track's start, length, pause and ISRC for the plant.

Other songs on the album

An album is mastered while its songs are still changing, so a track can be another Nebula song rather than a bounce of it. Nebula renders it from its file as it was last saved, offline, without opening it here: the song you are in keeps playing as it was. The render is kept on this Mac, and the list says where each song stands: Up to date, Changed since its render when the song has been saved since, Not rendered yet, or Song missing when its file has gone.

Click Update Songs to render every changed song again from its file. Measure, Match Loudness and Make DDP… do it too, so a disc is never pressed from an old mix.
A song is rendered from its start to its end, as Bounce would with your tail setting. A song that needs a realtime bounce (an external insert, a hardware synth) cannot be rendered offline: open it, bounce it, and add the bounce instead.
Its title is the song's file name until you give it one, and its ISRC is typed on the album.
The song you are in is added with Add Song, not as another song.

Each track starts on the sector its first sample is in, so a gapless join is never broken. Where the album cannot be a disc, the page says why before anything is written: more than 99 tracks, a track shorter than 4 seconds, more than 80 minutes, a crossfade longer than half a track, an ISRC used twice, or a part whose marker is gone.

Claude can do this for you: ask it to put the album in order with two-second pauses, add your other songs, bring their renders up to date, match its loudness and make the DDP.

Broadcast WAV and time stamps

A WAV file from Nebula is a Broadcast WAV: it carries a bext chunk and a small iXML chunk. Besides the song's and the part's names and when the file was made, the bext holds a time stamp: where the file starts in the song, counted in samples from the song's start. Drop the file into another DAW or a video editor that reads it, and it can place the file at the bar it played at in Nebula. A part from Export Ranges starts at its marker or section, so it lands back in place.

Bounce in Place

Bounce in Place turns regions into audio where they are: a MIDI part printed to audio, or effects committed.

Right-click a region, or several on one track, and choose Bounce in Place… — or drag a range with the Marquee tool and choose Edit ▸ Bounce Range in Place….
Choose whether the track's insert effects and its volume and pan are in the audio, how long a tail to keep for what rings on, and whether the originals are muted or keep playing.
Press Bounce.

The audio goes on a new track under each track, exactly where the material was, and plays to the same output and sends. Volume and pan you leave out of the audio stay on the new track's fader; effects you leave out are not on the new track, so the audio is dry and ready for new ones. Muted originals stay in the song; a track with nothing left playing is hidden. Undo takes the whole bounce back in one step.

Process with Plug-in

Process with Plug-in renders audio regions through one effect and puts the result in their place — a reverb swell on one hit, a filter on one phrase — without an insert on the whole track.

Choose audio regions on one track, then Edit ▸ Region ▸ Process with Plug-in…, or right-click one and choose it there.
Choose the Plug-in. Its window opens on the regions, which loop while you set it; Audition plays them through it, Stop ends that, Open Editor brings its window back.
Tick Keep what rings on to let a tail lengthen the regions, and set how long a Tail may be.
Press Apply. The regions are rendered exactly as Bounce in Place would with that insert, as one undo step: Undo Process with Plug-in.

Process with Plug-in works on arrangement regions. A launcher clip is not processed in its cell, and Nebula says so: place it in the arrangement, process it there, and put it back in the launcher.

Region ▸ Revert Processing brings back the audio a region had before its last pass. Claude can process a region for you too (process_region_with_plugin and revert_region_processing).

To repair a recording — hum, clicks, steady noise — process it with Nebula Restore: learn the noise in its window (Learn Noise while a stretch of noise alone loops, or From Selection), then Apply. To give a region another's tone, process it with Nebula Match EQ after learning both spectra. Both are described in The Tuner, the Analyzer and Nebula's effects. Asked to, Claude learns Restore's noise from the region's quietest second, and Match EQ's reference from the region you name.

File ▸ Share hands a bounce to the Mac's own share sheet — AirDrop, Mail, Messages, Notes and whatever else the Mac offers. Nothing leaves the Mac until you choose a service there.

Last Bounce… shares the file of your last bounce.
Bounce and Share… bounces the cycle, or the song with the cycle off, with the Bounce window's last choices, then shares it.
Add Last Bounce to Music puts the last bounce in the Music app's library. Music plays WAV, AIFF, Apple Lossless and AAC, not FLAC.

Song and track notes

Song Notes and Track Notes are words you keep with the song: lyrics, a to-do list, who played what, a reference to match. They are not MIDI notes.

Choose View ▸ Song and Track Notes, or the Inspector's Notes tab. The song's notes are at the top, the selected track's below.
Click Edit, or double-click the notes, to type; Done shows them formatted again. What you type is one undo step when you finish: Undo Edit Song Notes.
Two stars round words make them bold and one makes them italic; "- " at the start of a line makes a list, and addresses become links.

Claude reads and writes them too (get_notes and set_notes): ask it what is left to do, or to add a line.

MIDI files

File ▸ Export ▸ MIDI File… (⇧⌘E) saves the song's notes as a MIDI file for another program. File ▸ Import MIDI… brings one in, with its tempo and its track names.

Tracks from another song

File ▸ Import Tracks… brings tracks from another Nebula song into this one. Choose the song, tick its tracks, buses and folders, and say for each what comes along: its regions, its plug-ins and their settings, its sends, its automation, and its take folders and alternatives. A send goes to this song's bus of the same name, or to a new copy of the other song's bus — your choice. The audio is copied into this song; the other song is not changed. Everything keeps its place in beats, so bar 5 there is bar 5 here even when the two songs' tempos differ. Undo takes the whole import back in one step.

In the Library, Songs lists your recent songs: open one to see its tracks, and drag a track onto the arrangement to bring it in the same way.