Suunto 7 Accuracy – Production Release
Return to the main Suunto 7 Review
I’ve finished my first tranche of sensor testing with the Suunto 7. These results are from a production watch, not a beta watch.
At first, I was initially very impressed with the oHR but I’ve had unexpected dropouts when running outside in certain, repeated circumstances. Some I can’t explain but some are linked to the watch moving closer to my wrist bones. Either way, it’s a new oHR chip being used by Suunto, so things should improve in the coming months with tweaks to the algorithms. (I’ve been told by a 3rd party it is by PixArt or LifeQ). Considering this is a new chip for Suunto I would say the results for ME are much better than Suunto’s first attempt with the Valencell oHR chip a few years back.
GNSS/GPS is also using a new chip, this time it’s the Qualcomm Snapdragon 3100. The GPS is actually pretty good much of the time and sometimes produces an EXCELLENT track for extended periods. However, I have two general criticisms; firstly the track can sometimes be a little bit jaggedy and secondly, the Suunto 7 seems to have more issues than it should in situations where there is some form of slight difficulty in the GPS reception.
Although some of these technical results DO SHOW a need for improvement, I must say that I’m impressed with the wider abilities on the Suunto 7 watch (review to follow). I’m sure Suunto devotees WON’T be impressed with a smart sports watch like this but the Suunto 7 could be a commercial success for Suunto which might then enable Suunto to invest more in their traditional customer base…so, Suunto lovers, don’t knock it too much! It might be the financial route to give you want you want elsewhere on the Suunto platform.
Suunto 7 – Optical HR Accuracy
I am normally a BAD candidate for optical HR. Regardless of the vendor it rarely seems to work well for me.
For me, it’s been a mixed bag with the Suunto 7’s optical HR and it produces similar results to what you can expect to find across the same sports with different vendors’ watches. That means that your indoor workouts will likely show good results unless there are WEIGHTS involved and your outdoor’s workouts will be great until you either fiddle with the sensor or bang it in some way (bumpy road).
TO BE CLEAR: oHR varies from person-to-person, from sport-to-sport and varies from one environment to another. You could get better or worse results than me. Some forum chat and feedback I’ve had say the the Suunto 7 is the best-ever from Suunto and, literally, others say it’s the worst-ever from Suunto. #Sigh
Meh, not so good.
Sweet as cherry pie.
This is a similar run and here I was fiddling with the Suunto 7 at the start as it was too close to my wrist bones and I quickly noticed that the readings were too high. I moved it 1cm away from my wrist bones and Voila! All was good. Take note of positioning folks. During exercise, you might want to wear the S7 slightly further from your wrist bones than normal.
A nice long run next. Lovely jubbly.
And another good one
Despite the body being more static than when running you will probably grip hard at times and also move your wrist around on the handlebars and it’s those movements just cause some slight losses in the signal that you might otherwise have expected to see earlier with running on a treadmill. This is passable though.
These runs demonstrate that more tinkering with the algorithms is needed, although it can be OK in some circumstances.
This effort was totally acceptable. the sections where there are peaks and troughs are when I was running slowly on a muddy riverbank jumping around very large puddles.
Sustained and good in places and, err, less so at other times
finally this one
This long, and sometimes bumpy, bike ride was EXCELLENT with the Suunto 7. It only had one moment of indoor madness (red, circled) when the expresso arrived! Even the OH1+ had a few dropouts (blue, circled) which is very unusual (after a puncture).
This was also pretty good apart from one notable dropout.
This one is probably slightly on the lower side of acceptable
Weights – Upper Body
This is not so good, 40 minutes solely on upper body stuff. Stretching period for last 20 mins or so. The S7 was a couple of CMs from my wrist bone and the gym was hot. As you can see these are generally low levels of heart rate, so I would imagine the errors are solely caused by wrist movements. I get this type of result with any wrist-based oHR device doing weights.
Suunto 7 – GNSS Accuracy – GPS Accuracy
Suunto state that the Suunto 7 uses GPS, Glonass, Galileo via the Qualcomm Snapdragon Wear 3100 Platform. This is new hardware for Suunto to integrate & support. So expect some teething issues. Having said that, things don’t look too bad at all at launch (Feb 2020).
It’s likely that the accuracy of the Suunto 7’s GPS will change so I’ve not done all the tests I would do on a product whose development is relatively stable. Having said that I’ve done a few 😉 Here we go…
1. Cycling, Multiple repeats of a circle/roundabout in open sky conditions.
Key: Suunto 7 (Red), Stages Dash L50 (Green), Wahoo Bolt (Yellow, back pocket), Garmin 945 (Pink)
I thought this was interesting in the sense that it showed the repeated accuracy/inaccuracy of each device. They were all fairly consistent with the previous attempt(s) and against the others. In this instance, the Wahoo Bolt probably best reflects reality but I use that device A LOT and it is not especially accurate, per se.
Here are the same devices on the same ride. This time at a different circle/roundabout. I’ve highlighted in red where I actually was (UK cyclists use the other side of the road).
That is relatively normal. Maybe the trees have slightly thrown the devices off course with a reflected or obstructed signal? Sometimes you will see an inherent North-South or West-East skew in your track, I’m not sure of the exact reasons for that and it could be linked to slight ephemeris errors.
Fancy words aside, the track is good enough for me on a bike.
2. Suburban Running, round the block
Key: Suunto 7 (Blue), Garmin 945 (Red, Galileo), Coros Vertix (Green)
So this looks fine, right? At a high level, this is the sort of accuracy of track that every GPS watch vendor should be able to deliver today. Some can’t.
If we drill down into a bit more detail we can see some deviations from the pavements I actually ran on. Like this
If you saw that second image you’d probably be less happy. What’s happening here, compared to the bike tracks from earlier, is that there is probably an effect from GPS signals bouncing off the nearby buildings, that bounced GPS signal travels just that little bit further and hence your watch interprets that as you being ‘somewhere else‘. Sometimes that can produce awful, squiggly tracks which don’t look good but, more importantly, it means that the short term distances that are being estimated are wrong and hence your current running pace/speed is wrong. That effect is worsened at running speeds rather than cycling speeds.
Anyway, the Suunto’s track is probably the most accurate one there but they all produce an equally-pretty overall track AND the instant pace from each should be similar and, in this case, everyone’s instant pace was sufficiently accurate.
However, once you get much larger buildings involved in an urban environment and, in this case, a tunnel as well, then the errors become potentially catastrophic and far from acceptable. The Suunto 7 (Red) here is actually a relatively good track for this particular stretch as I run it VERY many times. But imagine running in downtown Manhattan where you will get this kind of error every single block..block after block after block.
There’s only so much technology can do and, no, your car’s Satnav won’t be much better in downtown Manhattan and nor will your iPhone – although at least being in the road rather than on the pavement can help a little. But you have to ask how the Suunto 7 can produce a better track than one from the year-old Garmin 945 ie from sports tech’s leading supplier and Garmin started out with SatNavs…
3. Running Under Trees
You can never truly tell if the GPS readings have been smoothed but here the 945+Galileo (green) tracks where I ran but the Suunto wasn’t too far off. This was underneath fairly dense, low and small trees.
But earlier on the same run, both devices struggled a little under slightly bigger and more sparse tree cover. I’ve drawn a black line to indicate the actual route. The round structure was one of King Henry VIII’s ice houses, they didn’t have fridges in those days and the ice stayed frozen for longer in these ‘houses’. Now you know. #LocalHistory
4. Cycling In Surrey
Here both the Garmin Forerunner 945 and Suunto 7 outperformed the Dash and Bolt at, err, cycling stuff. I did a couple of laps of Box Hill/Juniper and it’s good to see the nice consistency of both the S7 and 945. All are fine for my cycling needs though.
5. Formal GPS Test & Techy Bits
I also perform a 10-mile run on a pre-determined circuit so that I can compare any new device to previous efforts. It started out as a means to see how different devices handled different GPS conditions but then I eventually realised that my standard run formed the basis of a good test to produce a ranking as the route just so happened to have quite a variety of GPS ‘hazards’
I explain more here: GPS Test Methodology & Results where you can download any of the data files to look at yourself. Yep, even the Suunto 7 file is there. FWIW I tested the Suunto 7 alongside the greatest Forerunner 945 with GPS and Galileo enabled and the well-established and GPS-fine-tuned Garmin was only marginally better overall. #Sigh.
The Suunto 7 scored 69%. That sounds high but it is nothing special. It’s pretty much the same standard as the best Garmins but a fair way behind some of the great results from Suunto and Polar back in the good old days when, yes, even GPS test results had a rose-tinted hue of wonderfulness.
I take this test result as just one of the inputs into my overall assessment of GPS accuracy.
6. Instant Pacing Accuracy
Whilst the overall distance recorded is generally ‘about right’ this does NOT translate to a correct instant pace on the Suunto 7. In my case the Suunto 7 typically has me running at least 5 seconds per km faster than reality, sometimes 20-30 seconds faster. That’s NOT GREAT if you want to rely on pace for your training but, unfortunately, it’s the same on most other devices.
If you run in straight lines in open spaces then the Suunto 7 and most other devices will show reasonable instant pace. The reality is that few of us do that and the only solution for accurate instant pace is a footpod. My regular reader knows that it’s at this point I mention STRYD as the best solution 🙂
Note: Suunto’s pace is based on their branded FusedSpeed, which augments GPS speed/pace with wrist/arm movements. I’d thus hoped that this algorithm would improve over time as it got to ‘understand’ me better. But it didn’t.
I don’t plan to do any more elevation tests. This one was from 3 hours of reasonably hilly cycling and the Suunto 7 is good, note there is a DEM/SRTM-corrected track which the S7 is very close to, perhaps the closest of all?
On a shorter ride, you can see here that all the devices follow the correct elevation reasonably well, so your ascent/descent figures should be about right on other rides but clearly the absolute levels are somewhat ‘off’. I didn’t calibrate the elevation but the Garmin normally auto-calibrates, so I’m not entirely sure if the differences are a weather (air pressure) effect.
For those of you who want super-accurate elevation then Garmin has pretty much nailed DEM-based elevations for most topographies where an elevation can be assigned to a GPS position (DEM elevation NOT GPS elevation, they’re different). But these elevations based on a barometric altimeter look decent.
I might cover My Radar in the main review, that’s a cool and more serious weather app.
Other Bloggers Results
As you might guess most of the bloggers writing about endurance tech have back channels and sometimes I make that public. Here are some very interesting results from Eduardo at Correr una Maratón, his oHR results show better accuracy than mine. A word of warning though, you may well soon see some less good oHR results from other writers! As I’ve said above…results vary.
Eduardo’s stats also show numerous running cadence dropouts. I have NEVER got that.
As you can see he is using DCR’s analyzer tool, which is a good tool to compare multiple recordings of the same workout. I’ve started to use it as well for power comparisons and might use it longterm, depending on what happens with SportTracks next month.
Suunto 7 Sensor Accuracy – Summary
Suunto is ‘about there’ but need to do a little more work with on both the oHR and GNSS/GPS
GPS accuracy looks solid much of the time. With a bit of tinkering around the edges hopefully, further improvements are also possible.
oHR has some predictable dropouts for me in specific circumstances that occur too frequent for my liking. Despite that, Suunto is in a better position with oHR than with the Spartan’s oHR at launch. Others have found oHR to be VERY GOOD.
Return to the main Suunto 7 Review