We Are Constantly Looking For Writers And Contributors To Help Us Create Great Content For Our Blog Visitors.
Eleven seconds into Auphonic’s demo, the speaker moves away from the phone. The next sentence is still there, just much quieter. Anyone who has turned up a podcast to hear the guest, only to scramble for the volume when the host comes back, knows the problem.
Auphonic’s Adaptive Leveler adjusts volume within a recording. We wanted to know how much difference it made when the same person spoke from two distances. So we measured the original and processed versions of Auphonic’s public phone recording, using the same time windows in each.
We used FFmpeg’s EBU R128 meter to measure the average perceived loudness of each passage. The unit is LUFS. A more negative number means a quieter passage.
The original measured −24.4 LUFS. The leveled version measured −18.4 LUFS.
The original dropped to −35.2 LUFS. The leveled version held at −18.9 LUFS.
That took the gap from 10.8 loudness units to 0.5. A single volume adjustment to the whole file would have kept the original gap. Here the distant passage gained 16.3 units while the close one gained 6.0. You can play the first two versions side by side and listen for the move at 11 seconds.
We measured a sample selected by the vendor. The figures tell you how evenly the two passages play. They tell you little about how a voice sounds when a microphone also picks up traffic, room echo or another person talking over it.
Imagine a guest leaning back to think, then leaning in to whisper the punchline. Both moments register as quiet speech. Only one is a mistake. Auphonic’s editor lets you change the gain at a particular moment, so you can bring back the whisper and keep the correction on the quieter answer. That is the control I would look for before building a weekly show around automatic leveling.
Noise reduction is a separate choice. If a café or a music bed is part of the story, try the leveler alone first. Auphonic offers different modes for steady noise, changing noise and speech isolation. Its speech-isolation mode can remove music too. Listen through the transition into the music before approving the export.
Both passages start quiet. Slide the manual editor override on the Whisper Punchline to bring back its intimacy while keeping the boost on the quiet answer:
Click a mode to inspect what survives in a story recorded with café ambience and a music bed:
On November 17, Auphonic released a plug-in for DaVinci Resolve Studio. According to the release notes, you can select a clip or track, apply a preset and bring the processed audio back onto a new track in sync with the picture. It uses your Auphonic credits and requires the Studio edition. Our measurements above cover the public audio demo only.
Four 25-minute episodes add up to 100 minutes, inside the monthly two-hour free allowance. There are two catches for that same show. Free exports carry an Auphonic jingle, and a free multitrack production must run under 20 minutes. A 25-minute episode mixed to one file fits the allowance. A 25-minute session with separate host and guest tracks exceeds the multitrack limit.
For a podcast, training video or interview series where people drift off-mic, Auphonic is worth a trial on an episode you know well. The demo shows it can bring a quiet answer close to its neighbor in volume. For a scripted story built around whispers, music and room tone, keep time in the edit for a human listen.
Start with one difficult minute from your own work. Pick a quiet answer, a laugh and a deliberate pause. Compare both versions at the same playback volume. If the answer is easier to follow and the laugh still lands, you have a reason to use it on the next episode.
Discover AI.
Learn Faster.
Build the Future.
Download Now!