I think they all privately found they cause hearing damage. Since NC is sending out a cancelling wave to your ears doesn't mean they are actually reducing the total air pressure force inside your ears itself. To avoid future liability they pushed silent firmware updates that decrease the total sound pressure in the ear generated by the NC function while not telling anyone to avoid automatic future guilty verdicts.

Do you have a reference for this? I can't find anything that supports it (only the opposite!), and, intuitively, it doesn't make much sense.

My ear drum, and the hair cells behind it, do exactly one thing: respond to changes in air pressure. If the ANC isn't reducing air pressure forces, then why do I hear so little? The only way to hear less is to have less changing forces to disturbed them, is it not?

I assume the good ones designed with lowpass and aliasing filters to block out anything high frequency, but inaudible, that might damage hearing. These are absolutely not air tight, so there's nothing static, and the lack of bone-conduction sounds from walking makes me think it's keeping lower frequency pressure very well under control.

ANC is not actually cancelling out the second sound wave, it's exploiting an attribute of human hearing that two simultaneous out of phase sound waves are perceived as silence AFAIK. It doesn't mean that sound pressure goes away, and hearing damage can come from sound pressure that is too high for the hairs in your ear and damages those hairs as a result (ex: a lot of loud sound) is my hypothesis.

When I asked an audiologist if there has been research about ANC causing hearing damage, he said that it would be hard to get a grant structure that would work with it, since nobody is making money from it, and there hasn't been.

In tinitus, the frequency that people hear is around 4000hz, because that is where the first major bend in the cochlea is. It means that the loud sound that you hear doesn't necessarily have to be a specific frequency to damage hearing in that frequency either.

Two simultaneous out of phase sound waves do cancel each other out, there's not some 'silent sound pressure' left.

I think his idea is that some non planar wave within the ear would mean that the wavefronts would not be exactly matched for all reflection paths to and within the cochlea. This would mean the phase would not remain as they travel through the cochlea and could construct at some point around the hair cells.

But, considering the wavelength of 20kHz is 1cm and the ear canal is about 0.7, I can't imagine there's much room for anything but a nice planar wave heading down the canal, let alone within the cochlea.

But the QC25 remains as powerful as the QC35 ever was, so it wouldn't be that simple.

If a thing changes, people will complain. Thanks to the internet, we can find them. (This is true even if literally nothing changes except the version number.)

My Bowers and Wilkins PX has much worse battery life after updating its firmware. I've had them for two years and neglected to update the firmware till a week ago. Very disappointing when companies do this. We take the risk of bricking our devices all for a successful upgrade to brick it for us.