Why Does My Exported Track Sound Different From the DAW?

Why does my export sound different from the DAW?

A bounce that sounds different from the session has one of two causes: the file really is different, or you are playing it back differently. The first comes from master plugins and export settings. The second comes from encoding, streaming normalization and the device you listen on.

Work out which one you have before changing the mix. The table covers the common symptoms.

Are master bus or monitoring plugins printed into the export?

Everything on the master channel is printed, including the fader. In Ableton Live, the main export option renders the post-fader signal at the Main track's output.^[1] That is fine for your mix processing, and a problem for plugins that were only meant for your ears.

Room and headphone correction is the classic example. Sonarworks says its SoundID Reference plugin must be bypassed every time you export, otherwise the correction is imprinted on the mix and adds your system's coloration to the file, and that turning off the plugin's Calibration button is not enough: it must be switched off in the DAW's insert.^[2] The same applies to crossfeed, mono-check and "car speaker" simulators.

• Use a monitoring bus Sonarworks recommends a dedicated monitoring bus outside the master chain, or a DAW monitoring section that is bypassed during rendering.^[2]
• Stems missing effects When exporting individual tracks in Live, the Include Return and Main Effects option adds the return tracks those tracks use plus effects on the Main track.^[1] Without it, stems can sound drier than the mix.
• Hardware in the chain Ableton notes that rendering is usually offline, but runs in real time when the Set routes through an External Audio Effect or External Instrument.^[1] Outboard gear set differently from your last playback changes the result.

Which export settings change the sound?

Several export options change the audio on purpose, and some are switched on in templates without you noticing. Check this list against your render dialog before your next bounce.

Playback shortcuts that switch off during render

Some CPU-saving features behave differently in the render. Image-Line notes that FL Studio's Smart disable is active only during live playback and is temporarily disabled when rendering.^[4] The same page states that underruns do not happen in rendered audio, so glitches heard only in the export point to a plugin behaving badly; the DAW crackling guide covers how to isolate it.

Should you dither when exporting?

Dither only when you reduce bit depth for a final file, and only once. Ableton's manual says dithering should only be applied once to any audio file, recommends rendering at 32-bit if further processing is planned, and warns never to use its POW-r options on files going to a further mastering stage.^[1] Image-Line gives the same rule for FL Studio: dither applies to 16-bit WAV and MP3 files and should be applied once, during the final render.^[3]

Dither adds very low-level noise, so it is rarely the reason an export sounds obviously different. If a mix sent to a mastering engineer was dithered to 16-bit with a limiter on the master, the limiter is the bigger issue. The full workflow is in dithering for 16-bit delivery.

Why does my song sound quieter on Spotify?

Streaming services turn tracks up or down to a common loudness, so a heavily limited master gains nothing and may sound flatter than quieter, more dynamic songs. Spotify normalizes to -14 LUFS by default, measured with ITU 1770, and Premium listeners can choose Loud at -11 LUFS or Quiet at -19 LUFS.^[5]

Louder masters get negative gain; quieter ones get positive gain while Spotify leaves 1 dB of headroom for lossy encoding.^[5] If you compare your upload with the session, compare at matched loudness, not by ear at different volumes. The background is in what is LUFS and how streaming normalizes.

MP3 encoding, mono playback and the listening device

Lossy encoding changes the sound most on loud, bright masters. Spotify recommends keeping true peak below -1 dBTP, and below -2 dBTP for masters louder than -14 LUFS, because louder tracks are more susceptible to extra distortion when encoded for streaming.^[5] Ableton's MP3 export option renders a CBR 320 kbps file.^[1] Deliver WAV to distributors, and use an MP3 bounce only to preview what encoding will do.

The playback device matters just as much. LANDR points out that a phone held upright stacks its two speakers, producing a much more mono sound.^[6] Wide stereo tricks can cancel in mono and take a lead or pad with them; the mono compatibility check walks through the test. On Windows, the player itself can alter playback: Microsoft notes that audio enhancements can cause audio issues and can be switched off in Sound settings.^[7]

How to compare an export with your session

A null test tells you in one minute whether the file differs from the session. If the two cancel to silence, the file is identical and the difference is in how you are listening.

1. Bounce a clean WAV
   Same sample rate as the project, 24-bit, Normalize off, no dither, monitoring plugins bypassed.
2. Import it into the session
   Place it on a new track at bar 1, aligned sample-for-sample with the mix, routed straight to the master outputs.
3. Flip its polarity
   Use a utility plugin's phase invert on the imported track and play both together.
4. Read the result
   Silence or near-silence means the file matches. Audible sound means something in the render differs: check master plugins, export options and randomized plugin settings.
5. Compare playback fairly
   Play the WAV in the same player and at the same level as your references, with enhancements and normalization settings known.
6. Check the encoded version
   Encode a test MP3 and listen for crunchy highs or smeared cymbals; lower the limiter ceiling if they appear.
7. Check in mono and on a phone
   Sum to mono and play the file on a phone before you call it finished.