Since the subject of noise cancelling headsets has come up here a number of times recently I thought a few folks might be interested.
Grocery Outlet, a chain store spread across the U.S. that specializes in closeouts and returns, got a shipment of Phillips HN050 noise cancelling headsets in. $50 suggested retail, $10 at Grocery Outlet while they last. Some local stores sold out in less than a day, but others still had a few on the shelf days later.
Specs on the back say Active Noise Attenuation 50-1500hz >10dB at 300hz
The ones I got were the "wrap around the back of the head with little outside the ear pads." There was one other model that was an earplug version. I didn't get one of those, maybe I should have, didn't check the specifications or model number, definitely should have. It was also $50/$10.
So I bought two and tried one on my usual test case, street noise and the roar inside city busses. In my opinion there is a definite but barely audible reduction in the low frequency noise with these. If I'd paid $50 a copy for these I'd be pretty upset.
Comparing these to failing memory of my really old ANR, if I remember correctly, noise cancelling headset that I paid about $100 for maybe ten years ago, and which went intermittent and I haven't gotten around to fixing, made me think that the Phillips were substantially less effective. I seem to remember the ANR almost completely cutting the roar inside the busses.
Unfortunately, when it cut all the roar I then found I had to listen to fifteen ongoing conversations that had been previously masked.
So, finally, a question I've never been able to get what I thought was a really dependable answer to. If processing power went up say twenty fold, sample rate went up twenty fold, atod had four more bits of accuracy, is there some fundamental reason we couldn't cancel the speech up to say 4500hz and increase the dB reduction? Or is there some wall that the current specs are up against and it would require vast increases to get even modestly better results?