RE: https://mastodon.macstories.net/@macstories/116732061203928721
Hell of a photo.
(You’ll know the one I mean when you see it.)
| Blog | https://marco.org/ |
| Podcast | @atpfm |
| App | @overcastfm |
RE: https://mastodon.macstories.net/@macstories/116732061203928721
Hell of a photo.
(You’ll know the one I mean when you see it.)

It's my last day at Apple Park for my seventh in-person WWDC, and as I'm waiting for my final briefing just outside the Steve Jobs Theater – ever so magnificent in its polish, and yet always so strangely calm a place – I keep returning to a thought that's been circling my head, begging for
Because, of course, I would LOVE to build a feature into Overcast that detects "Siri" or "Alexa" during podcast playback over speakers or CarPlay and applies the same filter!
But I can't test the effectiveness of such a filter if I can't reproduce the effect even with other tools.
I've been able to nearly-perfectly reproduce this frequency reduction with an FFT filter in Audition, but I cannot reproduce the alleged effect.
When I play a recording of myself saying "Hey Siri, what time is it?” from speakers near my iPhone, it activates the iPhone every time, whether this filter is applied or not. (Same deal if I just start at "Siri" instead of "Hey Siri”.)
A 2019 Alexa Super Bowl commercial had a similar trick, and I couldn't reproduce it either: https://www.amazon.science/blog/why-alexa-wont-wake-up-when-she-hears-her-name-in-amazons-super-bowl-ad
hot take
two-space indentation is the code-formatting version of Alan Dye's zero-contrast distinctions between states