Beta Caption Export Significantly Reduced Music Volume in Final Video

Hello Adobe Podcast Team,

I am writing to report an issue I experienced while testing the beta captioning feature in Adobe Podcast.

First, I want to say that I am extremely impressed with Adobe Podcast overall. I use it for voice enhancement and noise reduction, and the results have been outstanding! The captioning feature also looked excellent visually. However, I encountered a significant audio problem after exporting the captioned version of my video.

Here is what happened (with explanation courtesy of ChatGPT Plus):

I uploaded a completed video containing a carefully balanced mix of voiceover and background music. Before adding captions, the voiceover was strong and clear, while the music remained fully audible and provided an important emotional and dramatic component to the production.

After adding captions in Adobe Podcast and exporting the captioned version:

• The voiceover remained strong and clear.
• The background music was dramatically reduced.
• On playback, particularly after emailing the exported file to myself and listening on another device, the music sounded as though it had been turned down almost to zero.

I compared the original final video with the Adobe captioned export. In spot checks of several sections, the underlying audio/music level appeared to drop by approximately 23–24 dB in the captioned version. Interestingly, the overall integrated loudness of the two files remained relatively close because the voiceover itself stayed strong. This means a basic overall loudness reading may not reveal the problem, even though the balance between voice and music has changed dramatically.

The approximate overall loudness readings were:

• Original final video: –22.8 LUFS
• Adobe captioned export: –22.2 LUFS

The issue therefore appears to be related to the mix balance during the caption export process rather than a simple reduction in overall output volume.

For context, the original video had already been fully mixed before captions were added. The voiceover and music levels were intentionally adjusted and balanced separately, and the final version sounded exactly as intended before being processed through the beta captioning feature.

I hope this feedback is helpful because the captions themselves looked very good. I would be happy to use the feature once the export process preserves the original audio mix without suppressing the music bed.

Attached is a comparison of the finished mixed video WITHOUT CAPTIONS through Adobe vs. the same version with captions added in Adobe.

Thank you for continuing to improve Adobe Podcast! It has already become an important part of my voiceover and future podcast production workflow.

I

Best regards,

Linda Kirk