The Audio Effect SDK seems to be a good option when it comes to filter non-stationary noises, which would benefit the far-end user a lot during video conferencing application.
However, is there any chance where the technology can be used so that it benefits the near end user at the same time? One Possible occasion I can think of is the terrible audio/Speech sent from the far end. If the audio algorithm is powerful enough so that the unwanted noises from the far-end can be filtered, therefore maintaining a much more pleasant experience for the near-end user regardless of the equipments used by the far-end user.
Is this already available in the Maxine Audio effect?
This is possible by filtering the incoming audio through the SDK. This was done in the NVIDIA Broadcast app by creating virtual speakers that denoise the incoming audio prior to playing it to the user. The same can apply for the room echo cancellation filter. This can help a lot with poor quality remote audio as well.
Great to know that this is already available.
Does it significantly deteriorate the voice clarity when the algorithms are applied?
Thanks in advance!
It doesn’t typically degrade the voice clarity, but this depends a bit on the nature of the sound and the settings you use. If you put the effect at 100%, it will filter out very extreme background sound, but often has some impact on the voice as well. We typically recommend around 75% or so as a starting point.