Running pace vs 965

MKPotts

I used an Ambit 3 Peak for about 2 years before switching to a S9B. For me, instant pace was just as realistic, stable and responsive on the S9B as the A3P. Location accuracy with the S9B was noticeably poorer in difficult conditions (tree cover etc) when I first got it - which affected pace sometimes - but now it’s more or less as good as the A3P was.
I used a Garmin Enduro for about 2 months, which I believe has the same Sony chip as the S9B. Instant pace was much less reliable with the Garmin - often unrealistic changes or a lack of responsiveness.
I always understood the calculations used to filter and average the data and how data from the other sensors is/isn’t used plays a big part with instant pace. Even if you assume the watch has a positional accuracy of 1m in good conditions, that will still give significant variability for each (1 sec) data point, especially if running more slowly. In my limited experience - comparing Garmin and Suunto watches that used the same Sony chip - Suunto was better at handling the data to give a usable instant pace.
I’m not sure the newer (multi-satellite and/or dual band) chips will make much difference to instant pace, at least under good signal conditions. I’ve not noticed any difference in instant pace with my S9PP compared to my S9B - both generally good and both useless on the London marathon around the 30km point, where the route weaves between high buildings.

zhang965

@MKPotts said in Running pace vs 965:

I used an Ambit 3 Peak for about 2 years before switching to a S9B. For me, instant pace was just as realistic, stable and responsive on the S9B as the A3P. Location accuracy with the S9B was noticeably poorer in difficult conditions (tree cover etc) when I first got it - which affected pace sometimes - but now it’s more or less as good as the A3P was.
I used a Garmin Enduro for about 2 months, which I believe has the same Sony chip as the S9B. Instant pace was much less reliable with the Garmin - often unrealistic changes or a lack of responsiveness.
I always understood the calculations used to filter and average the data and how data from the other sensors is/isn’t used plays a big part with instant pace. Even if you assume the watch has a positional accuracy of 1m in good conditions, that will still give significant variability for each (1 sec) data point, especially if running more slowly. In my limited experience - comparing Garmin and Suunto watches that used the same Sony chip - Suunto was better at handling the data to give a usable instant pace.
I’m not sure the newer (multi-satellite and/or dual band) chips will make much difference to instant pace, at least under good signal conditions. I’ve not noticed any difference in instant pace with my S9PP compared to my S9B - both generally good and both useless on the London marathon around the 30km point, where the route weaves between high buildings.

My experience on pace stability :

(Ambit 3 peak ≈ Spartan race ultra ≈ early firmware s9b )
Better than
(Later firmware S9B ≈ S9P ≈ S9PP)

my experience on gps tracker stability :

Ambit 3 peak ≈ Spartan race ultra ≈ later firmware s9B ≈ S9P)
Better than
(Early firmware S9B ≈ S9PP)

S9PP has a bigger Standard deviation, it’s painful for me.

MKPotts

@zhang965
What do you mean by “gps tracker stability”? How accurately the track reflects where you actually ran?
I’ve not seen any difference in the reliability of instant pace between A3P, S9B and S9PP. But I rarely run in areas where GNSS signal is challenging - not sure if you do?
Because the S9PP uses a new chip, it’s maybe not surprising you’re having problems with pace, given there hasn’t been any firmware update to optimise data processing.

zhang965

@MKPotts said in Running pace vs 965:

@zhang965
What do you mean by “gps tracker stability”? How accurately the track reflects where you actually ran?
I’ve not seen any difference in the reliability of instant pace between A3P, S9B and S9PP. But I rarely run in areas where GNSS signal is challenging - not sure if you do?
Because the S9PP uses a new chip, it’s maybe not surprising you’re having problems with pace, given there hasn’t been any firmware update to optimise data processing.

For example, I run the same route in Paris ( I don’t think it’s more challenging than London)

If I compare the gps trackers, other watch’s Standard deviation is pretty small,

But s9pp has the best gps tracker and the worst, it makes me uncomfortable, but I think it can be fix by future update ( anyway I’m not waiting it anymore)

MKPotts

@zhang965
Understood now and my experience is similar. Most of the time my S9PP track accuracy is very good. I’ve run the London Marathon the last 3 times with S9B, S9P and S9PP and the S9PP gave the most accurate track (better than a friend got with a Fenix 7X). But occasionally the track is quite a bit poorer. Whereas S9B/P are consistently good.
Back to the original question about instant pace, I think this just reinforces that data processing is the key factor when it comes to variability.
Hopefully your Vertical will be better. I think it has the same Sony chip as the S9PP, so guess they will already have knowledge of how to optimise.

isazi

@MKPotts next firmware update for S9PP should include a new Sony firmware.

suzzlo

@isazi any idea if S9P will also be updated?

MiniForklift

@André-Faria said in Running pace vs 965:

…but there is one thing that during the run is not identical, is the instant PACE.
Like sometimes I see 6min/h on the 965 and at the same time 4:50min/km on the Ambit 3, but in the end the average is similar.

Why does this happen? I am curious about the technical side of it.

This is something I also noticed when I was using my Garmin Enduro. Often while running the current pace on the Enduro is incorrect, for example I know I am running a sub 5min/km but the pace reads 6min/km. It’s odd and frustrating but as you mentioned, at the end of the run the average pace seems to be correct and all is good. When I first noticed this I actually thought I had the wrong setting as it was so far off what it should be. I don’t use that watch any more but I would be interested to know why it does that but the average pace works out to be correct in the end

FWIW my S9B has always been great with regards to current & average pace, but as @MKPotts alluded to mine does struggle when running in the forest/under tree cover

@MiniForklift said in Running pace vs 965:

@André-Faria said in Running pace vs 965:

…but there is one thing that during the run is not identical, is the instant PACE.
Like sometimes I see 6min/h on the 965 and at the same time 4:50min/km on the Ambit 3, but in the end the average is similar.

Why does this happen? I am curious about the technical side of it.

This is something I also noticed when I was using my Garmin Enduro. Often while running the current pace on the Enduro is incorrect, for example I know I am running a sub 5min/km but the pace reads 6min/km. It’s odd and frustrating but as you mentioned, at the end of the run the average pace seems to be correct and all is good. When I first noticed this I actually thought I had the wrong setting as it was so far off what it should be. I don’t use that watch any more but I would be interested to know why it does that but the average pace works out to be correct in the end

FWIW my S9B has always been great with regards to current & average pace, but as @MKPotts alluded to mine does struggle when running in the forest/under tree cover

Except that in my case is the garmin that is correct .
My intention with the thread was not discussing if one brand was more accurate than the other , was trying to understand why this happens

isazi

@suzzlo it should, but no new GPS firmware.

MKPotts

@André-Faria
In my first post I mentioned the Enduro because I had experience of using it at the same time as the S9B and they use the same Sony GNSS chip (as far as I know). So the differences I saw in the reliability of instant pace (S9B much better than the Enduro) must have been caused by factors other than the chip and its firmware (written by Sony?).
My assumption is that it is the use of other sensor data (accelerometer?), to compare against the location (GNSS) data, that is important (don’t Suunto call this “FusedPace” or similar?). This I believe helps the watch decide if you’ve changed pace versus noise/error in each individual (1 sec) location point. Over longer time periods, I guess these errors tend to cancel (and/or additional data processing?), hence lap or total run pace looks the same (as it did for my S9B vs Enduro).
If your 965 is more accurate for pace than your Ambit, I guess that means Garmin have managed to very effectively optimise data processing with the new chip (Mediatek?).
Like you, I’m also interested to learn more about how watches calculate pace, so would love to hear from someone with expertise in the area.

Dimitrios Kanellopoulos

You are probably seeing speeds ups and drops due to speeding up from static position etc. And that is fine if that is the case.

The instant pace in Suunto watches (As already discussed here ) is much better than most competitors (and a reason for this discussion from prev members such as @skyrunner etc) .

What you need todo in order to report a bug or put weight in the discussion is to compare FIT files on the speed chart.

Notice that Suunto watches use INSTANT speed / pace not averaged (5-10s) such as for example coros and garmin etc. That is a trade off for the instant speed.

Additionally with older watches than the vertical , running with left / right hand will have different paces on track and field due to the offset if the watch is not having a good signal.

@Dimitrios-Kanellopoulos said in Running pace vs 965:

You are probably seeing speeds ups and drops due to speeding up from static position etc. And that is fine if that is the case.

The instant pace in Suunto watches (As already discussed here ) is much better than most competitors (and a reason for this discussion from prev members such as @skyrunner etc) .

What you need todo in order to report a bug or put weight in the discussion is to compare FIT files on the speed chart.

Notice that Suunto watches use INSTANT speed / pace not averaged (5-10s) such as for example coros and garmin etc. That is a trade off for the instant speed.

Additionally with older watches than the vertical , running with left / right hand will have different paces on track and field due to the offset if the watch is not having a good signal.

I don’t want to point a bug or complain. Your last sentence gave me a clue, I was always with the ambit on right hand.
Thank you for taking the time to answer me and better understanding how things work, which was my aim.