Exporting to MP3 and WAV dramatically increases volume

I completed effects (normalization/Compression etc.) on my final product and the audio is basically what I want. HOWEVER, when I export to both MP3 and WAV, the volume is increase to the point of distortion.

Thanks in advance for how I may correct this situation

Hey @Dawacs ! We’re taking a look into this issue now and attempting to reproduce it within the team. Would you be able to share any specific reproduction steps or perhaps the file you were using where you observed the issue to help with this process?

That REALLY shouldn’t happen if you normalize. Audacity itself uses floating-point internally so it can go over 0dB without clipping. But you’ll still clip your DAC and hear distortion if you play it at full volume.

Regular (integer) WAV files cannot go over 0dB so you can get clipping in your exported file if you don’t normalize or otherwise lower the peaks below 0dB before exporting.

Here’s an experiment for you -

  • Make sure Show Clipping is enabled.

  • Normalize a “good sounding” file to -1dB. (You should no-longer see red.)

  • Export As WAV.

  • Open the new WAV file. (You still should not see red and it should still sound good, assuming “CD quality” or better).

  • Run the Amplify effect and note the default. It should default to +1dB, indicating that your current peaks are -1dB (you should have -dB of headroom). (You can cancel the Amplify effect if you just want to check the peaks without changing anything.)

A few more comments…

Show Clipping isn’t looking at the wave shape and checking for distortion. It just looks at the level and indicating potential clipping. You can have a clipped-distorted wave and if it doesn’t hit (or go over 0dB) Audacity won’t show red. Lowering the volume on a clipped file can “hide” the clipping but it doesn’t remove the distortion.

Or Audacity or certain file formats can go over 0dB without clipping and Audacity will still show red for potential clipping.

MP3 can go over 0dB without clipping. (It has a limit but I don’t know what it is.) And, since MP3 is lossy and it changes the wave shape, the MP3 often has higher peaks than the original uncompressed file. A LOT of the MP3s I ripped from CDs show red when opened in Audacity. And although the MP3 itself isn’t clipped, you can clip your DAC if it’s played at “full digital volume”. As far as I know, this slight clipping isn’t audible but some people normalize to around -1 or -2dB so the resulting MP3 doesn’t go over 0dB.

Compression (including limiting) can sound like distortion if “overdone” or with the “wrong settings”.

Maybe you accidentally nudged the track’s Gain slider away from its centre (zero) position.

Gain

Did “Normalize” Checked Remove DC offset and Normalize peak @ -1.0

Compressor at -18/-40/2.5/.8 /11.1 and checked Makeup gain

Filter curve EQ did treble boost and bass boost

Like I said, in Audacity, the sound was GREAT. It’s just when I used export for BOTH WAV and MP3 that the loud/distortion occured.

THANK YOU for your help

If I export with NO enhancments (Normalize/Compress) the sound is fine.

How many tracks do you have?

  1. Four-part harmony with three tracks per part

Looked, but it was all good.

Ah, I think I see what’s happening here: Track mixing is additive, so by adding 12 tracks normalized to -1dB (=0.89 amplitude), you’re ending up at 12*0.89=10.68 (or about +20dB). Audacity with its 32 bit float audio can handle this, but most audio formats and players cannot.

The solution here is called “gain staging” - basically, as soon as you’re working with more than 1 track, you don’t really want to put any one of your tracks to -1dB but instead tune each track such that the resulting mix ends up at -1dB.

If you have all 12 tracks on timelines, wouldn’t the bouncing sound meter show you the overloaded mix?

Koz

1 Like

I am running into similar issues. I have recorded several episodes of my podcast with no issues. Latest one seems to have hit a snag. I can play it fine in Audacity, but then when I do an export (whether it be MP3/WAV, any quality settings), I run into distortion on my voice.

I have tried the normalize and either I am not doing something right, or it doesn’t seem to do anything. I even tried to do the normalize loudness, which seemed to make a difference when playing in Audacity, but then when I export, it seems like none of those settings were saved on the export.

I did do an update (not sure what version from) prior to doing this recording. I tried installing the 3.3.0 version instead of the current 3.3.1 and it made no difference (at least with that recording).

Hoping someone can help me out. Would super appreciate it!

Thanks!

Missed a step there. How are you playing the finished work? What happens if you open the saved work in a fresh Audacity?

Koz