Once an audio track is watermarked, can I process it further?
It is recommended that audio watermarking be the last step in your audio process. Post-processing of a watermarked track can disrupt the watermark.
What environments are best for detection of watermarked audio?
There are two basic rules about detecting audio watermarks:
- It works best he closer you are to the sound source.
- It works best without a disruptive amount of background noise.
Rezolve watermarked audio has been successfully read from broadcast audio in living rooms and cars, as well as audio played in retail settings.
My prepared audio will be broadcast as part of a video stream. Will the broadcast compression hurt the watermark?
Rezolve has successfully tested compressed audio (for example, audio broadcast over cable boxes) with great success. In general, starting off with a less-compressed file in such situations is best practice. Use .wav when possible, or high-bitrate .mp3 (192kbps or better).
How does crowd noise affect audio scanning?
We have done some testing to see how volume levels of crowd/ambient noise affects the ability to scan an audio watermark. The results we have obtained show us that at the location of the consumer phone doing the detecting:
- Crowd noise can be up to +8dB louder than the watermarked audio and the Rezolve-enabled app will be still be able to detect the watermark. At this level, some words in the audio start becoming hard to understand.
- However, when crowd noise grows above +12dB louder than the audio (and the whole audio file becomes difficult to hear) the Rezolve-enabled app will start experiencing difficulties detecting the watermark (e.g. detection will take a long time).
This all assumes that the speakers are of good quality, and that a suitable audio track has been chosen. See What Sort of Audio Files Can I Watermark? to learn best practices.
So, for successful scanning, speaker volumes should be adjusted to be no more than -8dB quieter than the crowd/ambient noise (i.e. you can still hear most of the words in the recording).