I’m starting a podcast. The episode is two hours long and my initial file size was 110 MB. I know some other podcasts that record episodes of about that length at approximately 50 MB. When I went to export my audio, I went to “Options” and changed the quality setting from 128 kbps to 64 kbps. That gave me the file size that I want. The audio quality on the speaking parts is fine, but the music sounds really awful now.
For other settings. Bit Rate Mode is set to “constant”
If I go to Edit - Preferences - Quality I see the following values:
Default Sample Rate: 44100 Hz
Default Sample Format: 32-bit float
Real Time Conversion
Sample Rate Converter: Medium Quality
Dither: None
High Quality Conversion
Sample Rate Converter: Best Quality (Slowest)
Dither: Shaped
Is there a better way to reduce my file size without producing such a noticeable effect on the music quality?
It is highly recommended to Export your podcast Edit Master as WAV (Microsoft), 16-bit. From that edit master you can make the MP3 compressed versions of the show. Never open an MP3 for editing. Always go back to the WAV Edit Master. Multiple MP3 passes degenerate the sound quality and it can go quickly.
As a fuzzy rule, 64 is the limit for stereo and 32 is the limit for mono. So one simple way to slice the size of the podcast with very little quality change is post it mono.
This multiple pass MP3 thing can kill you in unexpected ways. Your voice segments are live, right? The music segments are all download MP3? So your 64 quality music is folded into your show and then exported at 64 which gives you a clear 64 quality voice, but 32 quality music.
iTunes has started to offer better quality music at higher cost. That’s their $1.29 music cuts up from the older 99c cuts. Also, the higher quality cuts don’t have FairPlay copy protection.
What effect do those other settings have on this? Would changing the project rate or any of those other settings help reduce the file size but maintain sound quality?
44100 is the sample rate of a Music CD. You can reduce that, but then your show will start sounding like a business voicemail system. “If you know your party’s extension…” If you go down far enough, you get AM Radio.
Audacity always runs internally at 32-bit to make the effects and filters work, so you can’t change that.
Nope. You need to use higher quality music, or more accurately, fewer compression steps. If you’re using MP3 for everything—a very new user mistake—just bumping your production quality to WAV may do it. If all you have now is the MP3 masters, it’s already too late.
Perfect quality WAV isn’t free. It can take between 5 and 10 times the drive space as MP3.
There is another variation. You can Save Audacity Projects. Nobody else but Audacity can open and play them, but they have perfect quality and they have the ability to preserve all your tracks, cuts, and production tricks, But not UNDO. Projects do not save UNDO.
Exporting a WAV is recommended anyway. Projects can be a little temperamental and brittle.
If you’re using MP3, selecting Variable bit rate, VBR, rather than Constant bit rate, CBR, can reduce the file-size whilst maintaining quality, but not by much, ~5% , depending on the ratio of speech to music in the show.
Best to use WAV for your podcast rather than using MP3, otherwise you risk spoiling the ship for a ha’penny-worth of tar. The podcast hosts usually transcode the uploaded files to their preferred specifications, adding (further) compression-damage.