You searched for track sounds worse after export, so you know the moment: the session sounded huge, you bounced it, and the WAV that came out feels small, dull, and quiet. Here is the reassuring part. The export is almost never broken. Rendering is deterministic math: the same plugins processing the same audio produce the same file, apart from a few real exceptions named further down. What actually changed is the listening situation, the loudness, or one real technical setting. This guide runs through the causes in order of likelihood, with the fix for each, and ends with a two-minute test that proves whether your export changed at all.
Louder reads as better, and your export is losing that contest
The most common version of the problem: you drop the bounce into a playlist next to commercial releases and it sounds weak. Those commercial masters sit at minus 7 to minus 8 LUFS. Your bounce, if you mixed with sensible headroom and light master processing, sits closer to minus 14 LUFS. That is a 6 to 7 dB loudness gap, and the ear reads even a 1 dB difference as "better" before it reads it as "louder." Nothing about your mix got worse. It is simply being judged against tracks playing twice as loud. The same trap exists inside the DAW: if your master chain had a limiter pushing level, or your monitoring volume was higher during the session, the export plays back quieter than your memory of the track.
The fix is to compare at matched loudness, and always by turning the louder file down, never the quieter one up. Pull the commercial reference down 6 to 7 dB until both tracks feel equally loud, then judge balance, width, and punch. Most exported tracks survive that comparison far better than expected. How loud the final master should actually be is a separate decision, covered in the loudness target guide, and the discipline of level-matched comparison is the core of working with reference tracks.
Your bounce plays through a different chain than your session did
In the studio, the session ran through your audio interface into your monitors at a volume you chose. The bounce gets played in a media player, in a browser, on your phone, or through a streaming app, and every one of those changes the sound. Windows media enhancements and "loudness equalization" reshape the dynamics. Phone speakers cut everything below roughly 300 Hz, which removes your entire low end and most of the punch. Streaming normalization is the big one: Spotify turns masters down to roughly minus 14 LUFS and Apple Music to around minus 16 LUFS, so a master you pushed to minus 8 LUFS gets pulled down 6 dB or more, while a quiet bounce gets turned up only cautiously, and on some platforms not at all.
The fix is to remove every variable except the file itself. Drag the bounce back into your DAW on a fresh track, or play it through the same interface and monitors at the same volume as the session. If the mix sounds different in your media player but identical through the session chain, the player is the problem, not the export. Disable enhancements, disable any player EQ, and only then trust what you hear.
Fresh ears versus a fatigued session memory
After a 3-hour session your ears have adapted to the mix. Harsh highs stop registering as harsh, a 4 dB low-mid buildup becomes the new normal, and a buried lead sounds intentional because you have heard it five hundred times. Then you bounce, sleep, and play the file the next morning. The flaws jump out immediately. It feels like the export ruined the track, but the timeline is the giveaway: the file did not change overnight, your ears reset. The bounce is honest. The session memory is not.
Treat next-morning disappointment as free information instead of a rendering bug. Note the two or three things that bother you, fix exactly those, and bounce again. If you want a read on the file that does not depend on what your ears did yesterday, upload the bounce to TrackSensei and look at the measured loudness and frequency balance. Numbers do not adapt after three hours of looping the drop, so they settle the argument between the session memory and the morning playback.
The export settings that genuinely change the sound
Real technical causes exist. They are just last on the list, and each one leaves evidence. A limiter or maximizer set differently for export than for playback changes the level and the transients. Dither applied twice, or applied on a 24-bit file that never needed it, adds a noise floor. Sample-rate conversion from a 48 kHz session to a 44.1 kHz file through a poor converter subtly softens the top end. An export range that starts after bar one misses automation ramps, so a filter opens too early or a send never returns to zero. Plugins with free-running LFOs, random modulation, or unsynced delays render a different take every bounce by design. And in rare cases an overloaded CPU during offline render bakes clicks and dropouts permanently into the file.
Check the export dialog against a short list. Export at the same sample rate as the session, so no conversion happens at all. Use 24-bit for any master in progress. Apply dither once, and only when you are finally going down to 16-bit. Select the full project range including the reverb tail. If a modulation-heavy plugin sounds different on every render, freeze or print that track first, or render in real time. These same settings decide the quality of the final file, which is why the home mastering guide treats the export stage as part of the master, not an afterthought.
A bounce is a commitment, and commitment always sounds worse
The last cause has nothing to do with audio. Inside the session, the track is still potential: every flaw has an imaginary fix attached, every section could still become the better version in your head. The bounce removes that comfort. It is one fixed file, and what you hear is what the track actually is. That psychological drop reliably reads as "the export sounds worse," even when a null test proves the audio is identical.
The cure is exposure. Bounce early and often, not just at the end, so a rendered file stops feeling like a verdict. Listen to bounces outside the studio: in the car, on earbuds during a walk, on the kitchen speaker. The producers who finish tracks are the ones who got comfortable hearing their work as a file instead of a session.
The null test: prove your export is identical in two minutes
When you want certainty instead of reassurance, run the null test. Import the bounce back into the session on a new track. Line it up sample-accurately with the start of the arrangement. Flip the polarity on the imported file, route everything to the same output, and press play. If the result is total silence, the two signals cancelled perfectly, which means the export is bit-for-bit what the session produced. The file is fine, and the difference you heard came from loudness, playback chain, or your own ears. If something survives the null, listen to what remains: it points straight at the plugin or setting that rendered differently, usually one of the technical causes above.
That is the whole diagnosis. The export is honest; the question is only which part of the listening situation changed around it. If you want that verdict in numbers, upload your bounce to TrackSensei and read the loudness and frequency-balance sections. It measures the file against released tracks in your genre, so instead of wondering whether the export got quieter or duller, you see in LUFS and dB exactly where it stands.