
Noise in multi-channel contact centers
- On August 9, 2020
- contact center, noise cancelling software
Today, some of the legacy call centers are being converted to multi-channel contact centers. In addition to taking phone calls, the agents also provide service via other non-voice channels like: live-chat, social media and emails. These multi-channel contact centers are noisy like the legacy call centers but is there a difference in the approach for noise removal?

Is the noise different?
The noise that is generated in voice-only call center contains mainly the voices of other agents talking over the phone in the background but there are also some ambient non-human sounds like phone ring. In case of multi-channel contact centers the noise contains human voice of other agents talking over the phone and agents chatting among themselves but there is also a fair amount of non-human sounds like: agents doing paperwork and other general office sounds.
As we can see in both cases the noise contain both ambient human voices and non-human sounds and both types of noise should be removed. How these two type of noise could be removed?
Removing non-human sounds
The algorithm that removes non-human sounds uses a single microphone. It listens to the audio signal of the call and extracts only the human voice while attenuating the non-human sounds. From this respect the behavior is similar in both call center and multi-channel contact centers.
Removing ambient human voice
The algorithm that removes ambient human voice needs to identify the physical location from which each human voice is generated in order to decide if it should be attenuated or not. In order to do this operation, the algorithm listens to multiple audio streams that are coming from different locations in the contact center and then by correlating the audio from the different locations the noise-map can be generated. In case of a call center each agent that is talking over the phone provides and audio stream for the correlation. In case of a multi-channel contact center the number of audio stream for correlation depends on the number of agents that are on the phone. In rare scenarios that this number is extremely low – let’s say only 1% of the agents are making phone calls then we can use few “dummy phones” to make sure that there are enough audio streams to provide a high quality removal of ambient human voices
Noise removal for at-home agents
In case of at-home agents, the logic is the same. The noise may contain non-human sounds like a dog barking and also ambient human voice like other people talking in the background. The same technology that works in the office can be easily deployed for at-home agents.
Conclusion
If you are looking for a professional solution to handle all type of noise and work both from the office and at-home, you should work with the experts. Contact us and we will be happy to show you how our unique Noise Firewall can help you.