Suunto 9 baro very low elevation gain vs other brands

Brad_Olwin · 31 Dec 2020, 20:53

@stromdiddily said in Suunto 9 baro very low elevation gain vs other brands:

@Brad_Olwin your experience does not match my own. But I will leave this topic here.

May “missing” elevation be the worst of our problems in 2021

My point from your last post was the A3P may not be providing the “correct” altitude gain/loss for comparison. I agree that larger gains/losses typically are more accurate for me and the low altitude runs are substantially more variable. However, when comparing with correction algorithms I find the watches are overestimating more often than underestimating. Altitude is tough to get correct and my benchmarks when running or on skis is to compare the actual summit altitude with the watch, the watches are often quite close. As has been discussed before, clothing, wind, and I suspect arm movements affect the calculations and the sensor. In my experience none of the Suunto watches I have are different from each other. My only other comparison is with an Apple Watch and I found excellent agreement with gain/loss on Apple Watches compared to Suunto. I can post a large number of those comparisons too.

On a final note, many of you know I field test for Suunto. This is part of the field testing and many testers compare more than one brand. Based on what I know and what I have observed from my experience and others I don’t believe there is a consistent problem here with watches that have barometers. The GPS altitude watches are another story and I have a different opinion there.

fazel · 31 Dec 2020, 22:24

@Brad_Olwin how are you running both watches? Pairing and unpairing in Suunto App? I’m still likely to run with both watches to satisfy my own curiosity. Nice “hill” repeats

Brad_Olwin · 31 Dec 2020, 22:57

@fazel said in Suunto 9 baro very low elevation gain vs other brands:

@Brad_Olwin how are you running both watches? Pairing and unpairing in Suunto App? I’m still likely to run with both watches to satisfy my own curiosity. Nice “hill” repeats

Thanks @fazel!. I run often with 2, 3 or 4 watches. Two on one arm, one on one wrist and holding one or one under a hat. I connect/disconnect all to SA on one iPhone and delete all duplicates. I have my exercises set to private so they post only when I have edited those out I am not keeping…daily routine of a tester…

fazel · 31 Dec 2020, 22:58

@Brad_Olwin Thank you!

Brad_Olwin · 31 Dec 2020, 23:04

@fazel said in Suunto 9 baro very low elevation gain vs other brands:

@Brad_Olwin Thank you!

Sure, sometimes not all that much fun when I am hitting the lap button on 2 watches. Today I used S7 and S9B for hill repeats lap button on S7 and testing S9B SuuntoPlus Climb, which lags behind a bit unfortunately. Here is altitude, HR (both on OHR, S7 is better but not perfect) and Stryd watts. S7 OHR lagged on the first and 3rd intervals, S9 OHR too high at the beginning. Watts were good!
Workout after warm up: 1 mi uphill, 0.75 mi downhill, 2 min rest, 0.75 mi uphill, 0.5 mi downhill, 2 min rest, 0.5 mi uphill, 1 mi downhill, 2min rest and cool down.

fazel · 31 Dec 2020, 23:06

@Brad_Olwin holy moly. I checked some runs on TP after I read your first reply of the day and noticed the corrections look staggered some. I’m wondering if you’ve ever seen that? I was thinking it might have something to do with using the Stryd for distance and the watch drawing the GPS track on top of it?

Like this…

Brad_Olwin · 31 Dec 2020, 23:12

@fazel FusedAlti will continuously correct so the staggered start is likely getting the altitude dialed in, I often see the same and you can see that in my post from TP here. I tend not to worry about it too much. My experience with Stryd for distance is not great, way short on trails. All of my posts here are trails, I do all my interval training on trails (sometimes wide and flat or wide and hilly like today) but hardly ever pavement. If your Stryd is accurate that may not be a bad idea. I have my Stryd paired as a powerpod only. If you pair as a footpod, the Stryd distance, pace and cadence will override the GPS data. Make sure you set auto calibration to off in the footpod menu or it will mess up the Stryd metrics for distance and pace.

Edited per your photo…I have never seen this! That does not look good. Do you have pause or auto pause on? I see breaks in the corrected.

fazel · 31 Dec 2020, 23:16

@Brad_Olwin so what is your advice for someone like me who needs it paired as a foot pod for the treadmill and runs gravel roads far more often then trails? The way I see it I could

leave it paired as a foot pod but only use it for the treadmill
leave it paired as foot pod, use it for the treadmill and roads but not trails
pair it as a power pod for outdoor use, then disconnect and re-pair as foot pod for treadmill use

Correct?

Brad_Olwin · 31 Dec 2020, 23:19

@fazel said in Suunto 9 baro very low elevation gain vs other brands:

@Brad_Olwin so what is your advice for someone like me who needs it paired as a foot pod for the treadmill and runs gravel roads far more often then trails? The way I see it I could

leave it paired as a foot pod but only use it for the treadmill

leave it paired as foot pod, use it for the treadmill and roads but not trails

pair it as a power pod for outdoor use, then disconnect and re-pair as foot pod for treadmill use

Correct?

Test your distances with and without the Stryd on pavement and trail. Just run without and with, I would leave connected as footpod. If you find discrepancies then you can decide.

fazel · 1 Feb 2021, 19:12

@Fenr1r no detailed analysis yet, but I can tell you they were within 6%.

9 Baro: 2549’ of ascent
Ambit3 Peak: 2726’ of ascent
Buddy’s Garmin 920XT: 2703’ of ascent
Buddy’s Garmin 235: 2558’ of ascent (no barometer)

What is more interesting are the distances.

9 Baro paired to Stryd: 22.74 miles
Ambit3 Peak via GPS: 21.58 miles
Buddy’s Garmin 920XT: 21.00 miles
Buddy’s Garmin 235: 20.83 miles

Links to Activities
Ambit
9

Wow - the Ambit draws pretty tracks…

Oktan · 3 Jan 2021, 08:22

Is there an email / link to support contact form (I looked and didnt found one), where users can complain about the elevation issue?

Dimitrios Kanellopoulos · 1 Mar 2021, 08:41

@Oktan here is one place we are about 5 fieldtesters here for suunto here , and 3 debating.

Dimitrios Kanellopoulos · 1 Mar 2021, 09:59

@fazel I spent some time again with your files and passed them through various services and still the ascent is reported lower from all dbs except Strava.

Try loading your ambit file to runalyze

So you can see above corrected and uncorrected.

Now I loaded both files to qs.

The altitudes are not so similar. One of the 2 had a fault probably due to some hand movement.

Take a look

You can clearly see that the ascent was missed by the s9 not due to some software but rather to different readings especially at the end that the diff is quite close to the ascent diff.

Again the algo is the same for all watches , what you experience is more of pressure sensor changes.
You ask 2 things:

To lower the threshold. Due to filtering out outliers this doesn’t sound right and many testers and SUUNTO don’t support it. We play on the safe side. Climbing a mountain is ok. Climbing your porch bench 2000 times and hitting everest or running on the beach facing wind and getting 2000+ ascent (I have examples) I don’t endorse personally nor from what I understand suunto.
To change the hw or position of the sensor ? (Not possible)
To change the gps chip or ad an antenna? Not possible atm.

Now, while this conversation might help someone , I will be leaving this thread for my sanity.

It’s pretty much the Nth time we look into this and actually there is no issue. I would like to spend my time with improving qs.
I cannot debate anymore on a saturated issue that has been iterated and tested from people (100+ fieldtesters) to machines that run the watches on rails. While we can argue that you want to register even 1 cm chnages as ascent Suunto and the RD team has chosen diffently and has chosen the filter and algos for each sensor that makes sense for the atheltes. Each watch from a3 to ssu to s9 has a different sensor and different sensor position with different pros and cons. For example a3 is prone to sweat while ssu is prone to wind and jackets. If you are not pleased guys and thus really bothers you to loose those 0.34 meters climbing everest or lifting your hand, I understand that but I personally don’t won’t to be bothered anymore. You can continue the debate here or anywhere else (contact support etc) but this is very very improbable to change any time soon. This might be improved at some point as a generic change for all watches and there are testers and a bigger team to validate that.

Have a happy new year , and my advice is to try to be healthy and not dictated by a tenacity. It is what it is.

Over and out from here.

And hey!!! Important:

You are doing great guys into looking into these stuff. I would do the same. I am not at all trying to stop you! Just want to be clear ️ . There is no problem adding any feedback here @fazel @Oktan etc. Please go on.

Egika · 3 Jan 2021, 10:11

@Dimitrios-Kanellopoulos and that post should be made a sticky one

Oktan · 1 Mar 2021, 10:49

OK. then last sentence from my side:

I’m a programmer too.
If I had in my program a constant (that its value can be argued) which fits 90% of the users, but the other 10% nagging the same issue over and over and consuming 20% of my time for explanation & proof - I would consider to let them change it (just a bit) for their own fine tuning ( and hapiness ).

Dimitrios Kanellopoulos · 3 Jan 2021, 11:07

@Oktan the best would be (although not easy) to allow you to change that. Suunto cannot change it for all because then the other percentage will complain then for example.

I think that is the best solution for all. User power.

Oktan · 3 Jan 2021, 11:28

I meant that 3 will be the default threshhold in the setting so the 90% will obviously wont change it, hence will not complain.
Anyways, thank you for your time, patience and all that you do for us.

Dimitrios Kanellopoulos · 3 Jan 2021, 11:45

@Oktan yeah I got it pal.

fazel · 3 Jan 2021, 15:09

@Dimitrios-Kanellopoulos Sorry - I thought my data was unconcerning and should have been clear. For me - this was within an acceptable range of deviation. I was relieved to be honest. You didn’t have to do all of this. I apologize for frustrating you.

Dimitrios Kanellopoulos · 3 Jan 2021, 15:42

@fazel no no. It’s ok. I just wanted to explain that many times we can miss some fault etc.

To be honest if I could I would make you a tester as you did take the time Todo that great run and compare.

Actually at your data the s9 I think did a boo boo at the end as I said which I reported.