Probably not. I have not directly measured the APM before and after all firmware updates, but this is six traces of the APM taken over time (from November 2021 to November 2022) that include :
- four different samples
- several different firmwares
- pads in a different state (the most different trace was measured with brand new pads, the others are all broken in pads, for a few traces the same ones) - so far that's the most significant factor I've measured.
And, mind you, averages to averages differences even when measuring the exact same sample.
View attachment 246980
Other than the feedback mechanism operating in ANC off mode and requiring a broad signal to update its filters (which can occasionally lead to inaccuracies when measuring them with sweeps in that mode with the usual multiple seatings, without priming them properly), I have not detected so far any DSP trick. The more recent H2 APP2 are a bit more complex but I don't think that it applies to the APM.
For me as well, for a fairly simple reason : most of the measurements I've seen so far are not representative of how they actually behave on my head.
This is the sample I have at home currently vs two different ways to guesstimate what the Harman target is meant to sound like, one an average of several large open headphones EQed to the target according to Oratory's presets, the other one an average of several high-scoring passive headphones from Harman, and my own preference :
View attachment 246981
If I compare the error curve vs these estimations and my own target, and the error curve vs Harman from measurements performed on 711 systems, including ASR (in red), I get this :
View attachment 246982
In other words as far as I'm concerned (and it could be different for someone else... or not, we'd need to actually measure the APM's on head behaviour on a cohort of real humans) the situation, while still not ideal, is a lot less problematic than what the ear simulator measurements would suggest, and it isn't a question of sample variation.
A recent presentation from Sean Olive on the question of coupling issues :
https://danishsoundcluster.dk/wp-content/uploads/2022/11/Olive_DSD_2022.pdf