23 thoughts on “Garmin 955 – Most Accurate Ever ? Accuracy Report”

1. Not surprised about the instant pace – having more accurate GNSS data aren’t going to make it update any faster – it’s just going to make (potentially) what it does report more accurate.

   1. I am surprised about the instant pace. It doesn’t have something to do with faster gps updates.

      GPS can be inaccurate up to a couple of meters. I’m running at about 3 or 4 meters per second. So every second I move 3 meters forward, but gps inaccuracy could measure me even 5 meters back. To work around this garmin/polar/suunto/coros don’t calculate instant pace over a 1 second period, but over a few seconds. This makes instant pace slightly more accurate but you notice speed ups with a delay.

      If GPS would be more accurate the instant pace can be calculated over a shorter period because strange spikes due to inaccurate occur less (if at all)

      1. Instant pace also incorporates accelerometer inputs. In Suunto’s terms, Garmin has Fused Pace. Sensor ‘fusion’ is applied to many of the metrics Garmin produces. eg altitude comes from a barometer, DEM map and/or 3-D GPS/GNSS.

         As an example of pace, yesterday I ran through a tunnel that was not far off 100m long and the pace was accurate to +/-10 secs/km, yet no correct GPS signal could possibly have been received. The reading came from the accelerometer (stryd was installed and disabled)

         Instant pace is also not ‘instant’ I can’t exactly remember but it will be something like a 3- or 5-second average.

2. I am not really blown anyway by the GPS improvements but then I never had a problem with the Fenix 6 either. With tree cover the new chipset seems to hold the track better, but with solid buildings close by the new chipset is doing worse than the Fenix 6 as it seems to throw it off by 2 to 3 meters vs. the actual track run. I don’t think the cost of battery life is worse the marginal improvements, and in sone cases it seems worse, not better.

   1. +/-2-3m needs to be seen as very good.
      yes i don’t think the extra accuracy is worth the extra cost or the lower battery life, at least not to me. the good thing about it is that people now have the option

      1. ‘+/-2-3m needs to be seen as very good’ – I half agree and half don’t! – an example if I am running down a typical UK street (llke your ‘shop test’), down the pavement with peoples front doors exiting onto the pavement, with the new chipset I am seeing my tracks showing me IN people’s houses rather on the pavement I’m running on surprisingly often. This surprised me, because the F6 tracks have me on the pavements (typically quite narrow) whereas the new chipset has me IN them. I kept double taking the tracks in DCR’s analyser thinking I was mixing up the watches…it’s weird.

3. Lot of ground covered, well done !
   oHR hiccups are a bit of a concern given the warm temperatures. Haven’t seen any of these with the Venu 2. The wrist-hugging velcro probably helps. What strap did you use ?
   Needs some testing on the track to get a sense of “real” accuracy (yest I know about the track mode, doesn’t work on 250, 300 or 350 meter tracks tough). I’ll try to compare it against my Venu 2 that’s been the most accurate GPS device I’ve seen on a track so far.

   1. standard garmin strap.
      i also prefer the velcro ones but the standard one fits snugly.
      point taken on the track GPS testing but i wont be doing that. (I do have a track nearby, St Mary’s where my 10-mile test starts from)

4. HR – better than Polar Pacer ohr?

   1. it will be results specifc to you probably.
      as you can see these hr results (above) were good and better than pacer pro but i think it’s just the warm weather this time round.

5. How accurate is the fr955 for respiration rate? I find my fr255 very inaccurate with OHR. Every morning I perform the health snapshot and at the beginning it shows me respiration rate 15 (while it should be <10). If I intentionally slow my respiration rate down to 4-5 per minute, then it start showing me 17. If HRV, stress score, … are of the same quality, then why bothering with using such metrics for any conclusions?

   1. using ohr at rest: it is consistent for me at 13 breaths per minute overnight which is probably right. for an athlete, i cant see the point in a health snapshot other than to take an hrv measurement at the same time each day
      i think what is important is the long term changes to respiration rates over the days/weeks when comparing the nightly averages.
      respiration is one of the pointers to changes to your readiness/recovery/stress state…body temp, hrv and HRrest are others.
      there is science that backs up the validity of looking at each measure in this kind of way.
      however all the wearables companies seem intent on creating a single ‘readiness’ type metric that combines all these individual measurements in some mysterious, unscientific way.

      1. Lets distiguish consistency from accuracy. For my older Forerunner 645 in cold temperatures I would would consistently see HR 180 for the initial 5 minutes of a running activity. Consistency =100%, accuracy=0%. Your respiration rate of 13 overnight looks plausible and you mentioned also consistent. Is it accurate though? As this is accuracy report, can you check yoga or breathing activity? Start with normal breathing and check if the results are accurate. Then start breathing very fast & at some point of time very slow to see if the results make sense. For me the results are consistently inaccurate (yoga showing 2-3x too low, while morning report doesn’t react to my breathing pattern at all). I know that respiration rate may not be important to you, but quite some people may find it usefull. Cheers!

         1. Garmin reports slightly lower than Oura Ring for average overnight respiration (14.7), although in recent weeks my oura readings have fallen perhaps due to easier training.

            my resonant breathing level is higher than what garmin state. i thus suspect garmin is slightly underestimating

            all other tools I’ve used in the are slightly higher than what garmin state

6. So do you still recommend running with a Stryd set as your Speed and Distance?

   I got one this week and did an indoor track run with it. Pretty happy with the results. But not sure it is all the useful outdoors in the open with my 955.

   1. i will definitely still be using stryd

7. Always love your reviews. Any chance you’ll consider a quick & dirty comparison of oHRM & GPS accuracy, maybe also battery, vs the old 945 (not LTE) for those like me evaluating an upgrade?

   1. sure 😉
      ohrm – probably better
      gnss – definitely better…it’s actually now officially GOOD ! before? less so
      battery – better
      so long as the 945 battery is good i would buy a stryd and decent hrm rather than upgrade.
      if you want a new toy…you know what to do (you deserve it and your partner wouldn’t spot the difference 😉 )
      personally I would and did upgrade but that choice is also a factor of this blog

      1. If it had had real native power, I wouldn’t have thought about it one second.. but as it is, I’m thinking 😛

         Still, will probably end up buying it. Thanks a lot for your reply!

8. How are you finding going back to this display after having used the EPIX for the past few months? In addition to better clarity, the EPIX displays more information.

   I’m still debating whether the lighter weight and better comfort of the 955 is worth the sacrifice of not having the EPIX’s excellent AMOLED screen…especially if it’s your day-to-day watch (to accumulate the data needed for the new training/recovery metrics).

   1. it’s just not as good on the screen front.

9. Your urban center test was fluff.
   How about a test in a real city, marathon quality city, like New York or Chicago?
   In those conditions you’ll probably see your track going wonkers, blocks off actuality.

   As for elevation, I was in Death Valley with my Fenix and used the opportunity to calibrate the watch with the posted elevations there. STILL not right. What’s really weird is if my course is circular, and I start and end at the same point, elevation will be different. ???!!

   1. I suspect my budget won’t stretch to flying to NYC from the UK and London (marathon-quality city) is just mostly a larger version of where I tested in Kingston
      I specifically say that I haven’t done cities like NYC, I can’t do everything. Few people are (surprisingly) interested in accuracy (I am)
      I don’t doubt that somewhere like NYC is bad for gnss reception.
      elevation – yeah that is weird. did you calibrate at the start

Comments are closed.