wearables

Brief Sennheiser Momentum Sport Data Analysis

Zë Au Sol

15 Apr 2024 — 4 min read

Sennheiser released their Momentum Sport headphones in 2024-04, an interesting partnership between Polar (they who invented the chest-worn ECG) and an audio company.

Historically, the ear has been an excellent point to measure both temperature and heart rate from, so logically these headphones should have been able to readily replace other devices (eg a Polar H10 and CORE body temperature sensor).

I took them out for a couple of runs and a ride. To generate the data below, I ran for an hour, recording data from the headphones via the Polar app on my phone, and from the H10 with my Garmin. Below are two graphs, one with, and one without cadence.

Heart Rate

They're not the cleanest, but even without formatting one can see how erratic the Sennheiser headphones are. Quantified, there's a correlation of 0.82 between the two series.

I included a graph with cadence, as sometimes optical sensors mistake running cadence for HR, and I wanted to show this wasn't the case for me: it's just a bad HR sensor, at least for running.

Temperature

Unfortunately, I haven't found out how to export the temperature from the Polar app, and as I don't have a Polar watch (does anyone, aside from sports tech reviewers?), this is the only way I have of recording the temperature data, so screenshots will have to suffice.

This is from the same run above. Below is a screenshot of data captured by the CORE sensor I was wearing at the same time:

The CORE captures three data: skin, core, and HR (from the H10). Without looking at the numbers, the graph shapes between the CORE and the headphones are different. Looking at the numbers, my core temperature increased from 37.06ºC to 38.43ºC, and my skin temperature fluctuates. The headphones show my temperature as static for the first 15m (bad data, should be rejected), and that my core temperature never goes above 37ºC. I reject the headphones ability to measure temperature with their current firmware based on the following arguments:

the numbers are different
the graph lines are different, both core and skin (ie Sennheiser/Polar can't argue they're tracking skin instead of core)
bad data at the start (the static line) isn't rejected

Sadly at the time of writing - press time - there isn't a way to capture the temperature data from the headphones without either the Polar app or a Polar watch, even though it looks to be exported via a Bluetooth standard (ie it could be, if the receiving device supported the protocol).

Audio

Ignoring the above, the audio on these is superb. They are the first headphones I've been able to hear background noise (cars, birdsong) but without wind noise. This alone is almost enough to justify their exorbitant price to me, as I prefer audio books while I exercise.

Can the above problems be fixed in firmware, or is this a bad marketing exercise, and a confusing one given Polar's reputation for HR data quality?

DCRainmaker Anecdote

This should not have been a difficult run – heck, as I often joke, even the Whoop band got it right this time (mostly). Granted, removed from this graph is the horrific accuracy of the Sennheiser Momentum Sport optical HR headphones (because it was so messy you couldn’t otherwise make sense of the other watches on this graph).