Treadmill distance too high in SSWHR Baro

aldnet

Thanks for the suggestions.

@oeagleo so, which data synchronize Milestone with Suunto, distance and pace? something more?

oeagleo

In my experience, judging from the inside walk I did yesterday, speed, distance, cadence, and pace show in Movescount after a “USB Sync”. The Milestone App also shows Ground contact, stride length, and foot strike data for runs. I don’t do run, as my knees don’t work so well anymore. In addition, the Milestone app shows graphs of Pace vs. Cadence, and Pace vs. Stride Length, and Pace vs. Ground contact time. Those would be nice to be imported into movescount, somewhat like the HRM-Run/Tri does for Garmin devices.

Darren_M

The easy answer to this is to add a setting to the Treadmill sport mode called ‘Stride length’. Essentially this is just a mathematical multiplication factor. One pulse from the accelerometer multiplied by the stride length and you achieve a set distance. Accuracy obviously depends on the accelerometer. However, it’s worth seeing stride length come to the watch settings as it will get you fairly close if the sensor is any good at all. Relying on the watch calibrating after running isn’t really a fix because if you run with someone slower, or vary your pace significantly, or even run up hill during a run, you’ll change your stride length accordingly and that has to result in an incorrect calibration. Also, the watch could produce estimated stride length from runs with GPS i.e. by distance/steps=stride length. I know from my Garmin my exact stride length at different paces, and they repeat to within 0.01 metres. With this stride length info you could alter the number in the stride length setting to suit the type of treadmill training you’re doing that session. The more data collected, the more scope you have for reasonably accurate treadmill training at different pace. Should be easy to implement, surely?

Dimitrios Kanellopoulos

@darren_m said in Treadmill distance too high in SSWHR Baro:

The easy answer to this is to add a setting to the Treadmill sport mode called ‘Stride length’. Essentially this is just a mathematical multiplication factor. One pulse from the accelerometer multiplied by the stride length and you achieve a set distance. Accuracy obviously depends on the accelerometer. However, it’s worth seeing stride length come to the watch settings as it will get you fairly close if the sensor is any good at all. Relying on the watch calibrating after running isn’t really a fix because if you run with someone slower, or vary your pace significantly, or even run up hill during a run, you’ll change your stride length accordingly and that has to result in an incorrect calibration. Also, the watch could produce estimated stride length from runs with GPS, i.e. by (steps/distance=stride length). I know from my Garmin my exact stride length at different paces, and they repeat to within 0.01 metres. With this stride length info you could alter the number in the stride length setting to suit the type of treadmill training you’re doing that session. The more data collected, the more scope you have for reasonably accurate treadmill training at different pace. Should be easy to implement, surely?

nice but why do you exclude cadence?

SlaSh

@dimitrios-kanellopoulos he already mentioned it by saying steps higher spm shorter stride length.

Dimitrios Kanellopoulos

@slashas an how would that be precise? For example, my cadence is 90/180 no matter what speed (up to 7:00m/km). However, my stride varies.
(playing the devils advocate here but I might be missing something)
Doesn’t then this make the stride a variable and not something constant?

If you have a milestone or a Stryd you will see this.

I can provide screenshots if needed

SlaSh

@dimitrios-kanellopoulos you are correct stride is variating depends on pavement and etc, but on treadmill it is almost constant.

Darren_M

@dimitrios-kanellopoulos said in Treadmill distance too high in SSWHR Baro:

@slashas an how would that be precise? For example, my cadence is 90/180 no matter what speed (up to 7:00m/km). However, my stride varies.
(playing the devils advocate here but I might be missing something)
Doesn’t then this make the stride a variable and not something constant?

If you have a milestone or a Stryd you will see this.

I can provide screenshots if needed

Strange because, running speed is the product of stride frequency (SF) and stride length (SL), and both are shown to increase when runners increase their speed. Anyway, putting the way the Garmin works it out to one side because the Garmin may be using a more complicated algorithm to produce stride length, it is average stride length, so maybe it’s sampling and looking for patterns rather than just raw maths. I obviously don’t have that information. I just know using stride length in the settings worked for me when on a treadmill. If you run on a treadmill at a certain speed you should be able to work out the error between watch and treadmill and add a number to a stride length setting to correct that error for next time. It will never be totally accurate, but it must be better than a 45% error.

Dimitrios Kanellopoulos

@darren_m said in Treadmill distance too high in SSWHR Baro:

@dimitrios-kanellopoulos said in Treadmill distance too high in SSWHR Baro:

@slashas an how would that be precise? For example, my cadence is 90/180 no matter what speed (up to 7:00m/km). However, my stride varies.
(playing the devils advocate here but I might be missing something)
Doesn’t then this make the stride a variable and not something constant?

If you have a milestone or a Stryd you will see this.

I can provide screenshots if needed

Strange because, running speed is the product of stride frequency (SF) and stride length (SL), and both are shown to increase when runners increase their speed. Anyway, putting the way the Garmin works it out to one side because the Garmin may be using a more complicated algorithm to produce stride length, it is average stride length, so maybe it’s sampling and looking for patterns rather than just raw maths. I obviously don’t have that information. I just know using stride length in the settings worked for me when on a treadmill. If you run on a treadmill at a certain speed you should be able to work out the error between watch and treadmill and add a number to a stride length setting to correct that error for next time. It will never be totally accurate, but it must be better than a 45% error.

If you take a look how stryd works or milestone you will see that your stryd length varies.
That said I am full on with your proposal and unfortunately suunto does not have that except an automatic calibration.

However back to our nice discussion if you do some research you will see that good running posture should have a stryd variance not cadence (stryd frequency).
For example the faster you go the bigger the stride should be. If you go slower and instead of deacresing the stride you decrease cadence then you are overstriding. And that is not good.

Anyways. I know there are plans in suunto for this but not in the very near future.

Dimitrios Kanellopoulos

@slashas said in Treadmill distance too high in SSWHR Baro:

@dimitrios-kanellopoulos you are correct stride is variating depends on pavement and etc, but on treadmill it is almost constant.

I am not 100% sure about this. Perhaps less variant but still my stats show differently.

SlaSh

@dimitrios-kanellopoulos you can check my treadmill activities most of them are 54spm cause I set constant speed on the tradmill so my spm is always constant in most cases

Dimitrios Kanellopoulos

@slashas said in Treadmill distance too high in SSWHR Baro:

@dimitrios-kanellopoulos you can check my treadmill activities most of them are 54spm cause I set constant speed on the tradmill so my spm is always constant in most cases

true that but how about if that changes? I never use constant on TM.

In a case of constant speed or same style lets set it like so everything is ok. It’s the happy flow in other words.

Darren_M

@dimitrios-kanellopoulos How accurate is the auto calibration method Suunto use? The OP has this massive error of 45% more distance. Is that kind of error common? That would be bad for those who run on them as part of their regular training plan.

Dimitrios Kanellopoulos

@darren_m said in Treadmill distance too high in SSWHR Baro:

@dimitrios-kanellopoulos How accurate is the auto calibration method Suunto use? The OP has this massive error of 45% more distance. Is that kind of error common? That would be bad for those who run on them as part of their regular training plan.

The calibration afaik happens based on GPS. So outside runs calibrate it.

The TM is an indoors device that shows you the pace and distance so I suppose (not really knowing) trying to match that with a watch will always produce an error so Suunto has not put effort into that.
That is because Suunto focuses on outdoors with it’s proclaimed FusedSpeed and not indoors.
In short, if you lose GPS you won’t want to have an “indoors” calibrated value dictating your speed but rather your outdoor calibrating value doing that.

As I ve read around even Stryd has some idoors issues with TM.

Might worth reading the :

Mysteriously Low Treadmill Pace

here https://the5krunner.com/2017/02/10/stryd-announcement-interesting-development-for-footpod-pace-and-treadmill-users/ and https://blog.stryd.com/2017/02/10/mysteriously-low-treadmill-pace-2/

also

https://www.dcrainmaker.com/2017/02/5-random-things-i-did-this-weekend-39.html#2-tunnel-testing

Dimitrios Kanellopoulos

However, just to add, there is a request and work on manual calibration and setting the above factors from what I have seen.

But again perhaps I have to displease you as I am kinda sure it wont be on the next update.

Darren_M

@dimitrios-kanellopoulos Ok thanks. It doesn’t upset me at all. In fact if I’m even looking at a TM I’m either recovering from injury or there’s something pretty running on it, haha! I’m all for the outdoors when running, especially on the trail. I am injured at the moment and about ready to test the damaged muscle, so that’s how I got onto the subject. I’m also an Instrument Technician, so I repair, maintain & calibrate precision instrumentation, therefore I don’t like errors, on any instrument, in any mode.

SlaSh

@darren_m for me on the beginning massive errors were thrown, but seems auto-calibration works it is now within 10-20% and I dont care much as I am focusing on HR and time as doing warm up treadmill incline speed walk.

Dimitrios Kanellopoulos

@darren_m said in Treadmill distance too high in SSWHR Baro:

@dimitrios-kanellopoulos Ok thanks. It doesn’t upset me at all. In fact if I’m even looking at a TM I’m either recovering from injury or there’s something pretty running on it, haha! I’m all for the outdoors when running, especially on the trail. I am injured at the moment and about ready to test the damaged muscle, so that’s how I got onto the subject. I’m also an Instrument Technician, so I repair, maintain & calibrate precision instrumentation, therefore I don’t like errors, on any instrument, in any mode.

Speedy recovery man!

Regarding the upset I am the middle man here and sometimes my position has to bare bad news for example (not sure how bad this is but I have seen people attacking me personally in several forums like I am the only dev/headmaster/etc on Suunto). Perhaps this has become my complex. (Just speaking my mind out loud here). However this place looks more sane

Darren_M

For the benefit of the OP and his question. I tried the watch on the treadmill today, but only did 2 km to check my injured calf muscle. I was impressed with the results. I wound the TM speed up to 12 Km/h (5:00/km) which took 0.11 km to achieve and then start the watch. I stopped the watch at an indicated 2.15 km on the Treadmill. Here’s what the watch produced:-

I know this is only a short distance, and it’s only one test so it may not repeat, but if it did produce this kind of accuracy repetitively, I’d be fairly happy with it.

Dimitrios Kanellopoulos

@darren_m to be honest I am surprised 🤩