Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Phone Noise Cancellation Missing?
- Thread starter ghsDUDE
- Start date
- Sort by reaction score
Well noise cancellation on a smartphone is 'better than nothing' at best (it sucks)..... Over the last decade you guys haven't had a phone calls where you couldn't hear the person due to background noise?
FaceTime is likely going to be the testing grounds for it and updated for phone calls eventually. This is just a wild ass theory but maybe some sort of legality hoops to jump through with the cell providers for directly processing voices? I don't know seems like a stretch but you never know.
The actually problem ---- Background noise interfering with phone calls.
ANC uses hardware as its solution. It listens for lower background sounds with a mic then using the DSP attempts to cancel them out. The problem is the hardware isn't that good at it. A tiny surface mount mic (MEMS?) and the low powered DSP in a phone can't hear and/or eliminate the spectrum of sound found in a noise background. If the background is louder than you, all bets are off. Even the energetic high frequency sounds of bird chirping nearby can cut right through ANC on a smartphones phone call.
Voice isolation (not a new tech) is a software solution and it approaches the problem more directly. While its "software" it requires the dedicated hardware for the software (Apples A12 or newer). Using AI it knows what its listening for and the rest can be ignored/cut. Another benefit to this is since its using AI it not limited to Voice Isolation it can also use a wide spectrum to somewhat "equalize" all the sounds the mic pics up. Loud sounds can be reduced soft sounds can be increased all at the same time.
Standard active noise cancellation using the microphone on a smartphone can not due this, not even close.....
The problem with Voice Isolation and Wide Spectrum modes when used on a phone is its very complex but it has to be done completely in real time and it has to be done very accurately. There is software for Mac/PC to isolate specific sounds in movies and music (so you can remove vocals from a song for example) and while I haven't used newer software it was also a processing load that wasn't in real time and always required a human ear to perfect.
FaceTime is likely going to be the testing grounds for it and updated for phone calls eventually. This is just a wild ass theory but maybe some sort of legality hoops to jump through with the cell providers for directly processing voices? I don't know seems like a stretch but you never know.
The actually problem ---- Background noise interfering with phone calls.
ANC uses hardware as its solution. It listens for lower background sounds with a mic then using the DSP attempts to cancel them out. The problem is the hardware isn't that good at it. A tiny surface mount mic (MEMS?) and the low powered DSP in a phone can't hear and/or eliminate the spectrum of sound found in a noise background. If the background is louder than you, all bets are off. Even the energetic high frequency sounds of bird chirping nearby can cut right through ANC on a smartphones phone call.
Voice isolation (not a new tech) is a software solution and it approaches the problem more directly. While its "software" it requires the dedicated hardware for the software (Apples A12 or newer). Using AI it knows what its listening for and the rest can be ignored/cut. Another benefit to this is since its using AI it not limited to Voice Isolation it can also use a wide spectrum to somewhat "equalize" all the sounds the mic pics up. Loud sounds can be reduced soft sounds can be increased all at the same time.
Standard active noise cancellation using the microphone on a smartphone can not due this, not even close.....
The problem with Voice Isolation and Wide Spectrum modes when used on a phone is its very complex but it has to be done completely in real time and it has to be done very accurately. There is software for Mac/PC to isolate specific sounds in movies and music (so you can remove vocals from a song for example) and while I haven't used newer software it was also a processing load that wasn't in real time and always required a human ear to perfect.