Adobe has recently released a new tool called Adobe Speech Enhance, a part of their Adobe Podcast suite of AI tools. This product enhances voice recordings, especially those captured under suboptimal conditions, by recreating the speaker’s voice. I assume.
Performance
It works.
Here is a recording I took from my iPhone:
And here is the output from Adobe:
How does that compare to full studio recordings?
This is a direct recording using a Heil PR40 in a treated room.
Ease of Use
To use the web app, simply drag in your audio file and wait for it to render. As of this writing, expect it to process about 10 seconds of audio in a minute.
Could not be easier. This is next-level user interface.
A Bad Recording Day
Use Cases
Rescuing Backup Recordings. Sometimes your primary recording device fails. When it does, you can use whatever backup audio from whatever mics happen to be around.
Cleaning Up Walk-and-Talks. If you record in a noisy environment with traffic, or other noise.
Lecture Rescue. Lectures are often known for having bad audio. This just fixes that.
Travel Recordings. Pro travel setups can be really annoying and heavy. New plan: Leave the setup at home and record remotely with your phone.
Remote Interviews. Conducting interviews over video conferencing software can result in varied audio quality due to different mics and environments. Adobe Speech Enhance can unify the sound quality for a more professional output. You don’t need to care if your guest has a great mic setup!
Public Speaking Events. This tool can clean up live event recordings for later broadcasts or archives.
Historical Archiving. Old audio recordings that suffer from degradation or noise could potentially be cleaned up for historical or educational purposes.
Original recording from the Hindenburg Disaster:
Processed recording from the Hindenburg Disaster:
Original recoding from FDR’s speech:
Processed recording from FDR’s speech: