Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

petterihiisila

macrumors 6502
Original poster
Nov 7, 2010
404
304
Finland
Edit: Doesn't look like it does. But the ANC performs well nevertheless.

Original post:

***

Here's an observation. I got my pair yesterday. Out of the box the ANC was pretty much on par with my Bose NC700, more or less. There are certain rattling fans and whirring freezers in this house that are very hard for ANC to block completely: a combination of white noise and treble-rich peaky buzzing that a bad bearing might make. It's repetitive but high-pitched enough to make any ANC work hard.

So, I've got this one fan in the house that I use as my highly non-professional benchmark of ANC performance. I could swear I heard it yesterday when I made my initial judgment: "yep, about the same as Bose ... it's almost gone but it's there."

But today I didn't hear it at all, had to check that it's on. I brought my head right next to it. It's gone. Yet I see it spinning and can hear it like always in Transparency Mode. The same familiar rattle.

Is there any anecdotal or factual evidence that the Max is actually learning over time? I don't mean the first few seconds it takes to adjust, but calibrating itself over hours/days: "Hey, I've heard this pattern before, here's how to to erase it."

... or am I just a victim of good old-fashioned placebo, confirmation bias or a broken memory? That damn fan is now gone, that's for sure, and it is not with the XM4, NC700 and APP, I've had them all.

Technically it's plausible that the Max learns over time, though the same could be said about the APPs. The H1 is a powerful computer on its own and it has the benefit of the zillion operations per second that the iPhone's ML cores could provide for pattern matching on the phone, if these are programmed to do such offloading. Plausible isn't the same as true though.

The press release talks about ANC being adaptive, but only mentions "fit and movement in real time" as input parameters. That's not the same than an overnight pattern matching exercise.
 
Last edited:

Lartymarf

macrumors regular
Aug 11, 2007
132
6
If you wanted to repeat your test, you could do a Factory Reset of your APM and set it up and try again. If you hear the fan after reset, and then a day or 2 later you don't hear it anymore, then maybe it does learn.
 
  • Like
Reactions: petterihiisila

petterihiisila

macrumors 6502
Original poster
Nov 7, 2010
404
304
Finland
If you wanted to repeat your test, you could do a Factory Reset of your APM and set it up and try again. If you hear the fan after reset, and then a day or 2 later you don't hear it anymore, then maybe it does learn.
I thought of that, but it wouldn't still be 100% conclusive. If there really is some kind of long term learning going on, as there plausibly (but I'm willing to admit not very likely) could, those learned patterns could plausibly be restored upon pairing. Just like an Apple Watch restores its identity from a backup...

Anyway, a factory reset was just completed, and before I had even set them up, the fan is already gone. I guess that falsifies the original hypothesis almost conclusively then, which is a bit sad, because it would have been awesome if true. ?

A more likely explanation for the phenomenon — assuming that it's not a false memory to begin with — is that the Max ANC just handles this kind of noise profile really well, and the fan had a particularly bad day yesterday, making it about as audible as it is with the other ANC headphones. Air humidity varies a lot at this time of the year and in this house especially, because it is old and there's no filtering or conditioning for the incoming air. Occam's Razor etc.

So, "nothing to see here" I guess. Unless some patent troll wants to run with the above idea before it's prior art!
 

Lartymarf

macrumors regular
Aug 11, 2007
132
6
Worth a try! I was secretly pulling for you to be right about the learning. Although if it did include some type of learning for ANC, I would imagine Apple would have included that as a selling point in the marketing materials.

I was also going to suggest after you do the Factory Reset to pair it with another iDevice that hasn't been paired to the Airpods Max before so that the Airpods Max wouldn't have the data from the learning before the reset. But, you said you reset it and right away the fan noise was gone, so has nothing to do with any information on previously paired devices.

Thanks for sharing your story.
 
  • Like
Reactions: petterihiisila
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.