Apps WorkOutDoors: New workout features

[cfc]

@cfc Thank you for detailed answer. The most popular services like Strava, Runalyze, intervals.icu don't smooth the track so option to choose between apple solution and Raw would be a nice solution

I expect that most of those services do smooth the track in terms of calculating the distance and therefore pace. They may still display the GPS points that are in the file, but they do not calculate the distance by simply adding together the distance between each point. If they did that then the distance would usually read too long on straight stretches of the route due to the unavoidable zigzag of a raw GPS route.

No GPS signal is perfect so no decent workout service would calculate the distance and pace without some sort of processing/smoothing.

As I say I may offer Apple's smoothing as an option in the future, but there are higher priority features at the moment.

 

[Maryn]

I expect that most of those services do smooth the track in terms of calculating the distance and therefore pace. They may still display the GPS points that are in the file, but they do not calculate the distance by simply adding together the distance between each point. If they did that then the distance would usually read too long on straight stretches of the route due to the unavoidable zigzag of a raw GPS route.

No GPS signal is perfect so no decent workout service would calculate the distance and pace without some sort of processing/smoothing.

As I say I may offer Apple's smoothing as an option in the future, but there are higher priority features at the moment.

I understand. I think they take the distance and pace from file because distance is 1:1

 

[NME42]

I understand. I think they take the distance and pace from file because distance is 1:1

In case you are interested: you can recalculate distance with mygpsfiles.

 

[noel18]

Sure. Screenshots from runalyze in highest zoom level.

WOD previous version:

View attachment 2011559

WOD latest version:

View attachment 2011560

WOD previous version:

View attachment 2011561

WOD latest version:

View attachment 2011562


I am running with an AW7 cellular, but almost always have cellular disabled to not be disturbed by messages or calls.

I'm currently injured so I haven't run on the new version yet so I can't confirm or deny that my tracks look different but I saw this in the change log:
"Added ability to export FIT files for workouts (now the default for Strava uploads)"

Is it possible that you had it set to TCX for Strava export and now it switched to FIT export and maybe Strava has different smoothing logic for FIT uploads? Maybe switching it back to TCX would give the same results you got on previous versions.

 

[NME42]

Is it possible that you had it set to TCX for Strava export and now it switched to FIT export and maybe Strava has different smoothing logic for FIT uploads? Maybe switching it back to TCX would give the same results you got on previous versions.

Yes, as I am using a Stryd, I also export FIT files now. Not to Strava though. The Screenshots are from runalyze, as I wrote.

And sure, @cfc said, the new version has more data points. If they are also in the TCX files, I don’t know and care.
But don’t understand your point, sorry. Why should I switch back to a slightly worse result?

 

[Maryn]

It doesn't matter whether Strava or Runalyze. In both cases the routes look bad

 

[cfc]

It doesn't matter whether Strava or Runalyze. In both cases the routes look bad

The displayed routes are the raw data. As I say, smoothed data looks nicer but won't affect the distance and pace because the receiving system is probably smoothing the data to calculate them, even if it is not smoothing it to look nicer on the map. This is what Strava does for sure.

Most devices (i.e. smart watches) that send workouts to Strava etc have a tiny fraction of the processing power and battery size of an iPhone, so they probably send raw GPS data, just like WOD does.

I may offer some smoothing capability in the future, if only for cosmetic reasons, but it is currently not high priority.

 

[jeon]

Now I'm using the updated WOD.
There is no change in the distance recorded in the WOD and strava since the update, but the exercise record displayed in the fitness app on the iPhone is marked fixed distance and is displayed shorter than the distance measured in the WOD.
Which distance is the right one?

 

[cfc]

View attachment 2011997
Now I'm using the updated WOD.
There is no change in the distance recorded in the WOD and strava since the update, but the exercise record displayed in the fitness app on the iPhone is marked fixed distance and is displayed shorter than the distance measured in the WOD.
Which distance is the right one?

Is WOD configured to show GPS distances or Apple's calibrated distances?

 

[noel18]

Yes, as I am using a Stryd, I also export FIT files now. Not to Strava though. The Screenshots are from runalyze, as I wrote.

And sure, @cfc said, the new version has more data points. If they are also in the TCX files, I don’t know and care.
But don’t understand your point, sorry. Why should I switch back to a slightly worse result?

I guess I skimmed the posts too quick and got confused. I thought people were complaining that the new version was producing bad results. @cfc confirmed that nothing changed. I was just pointing out that the default export format changed which may cause services to apply different smoothing algorithms (which is obviously out of @cfc's control).

 

[Duffman19]

A couple questions:

1. Does WOD contribute to Apple's calibration process? Or do only activities completed using the Workout app contribute?

2. a) Does using a screen with a map consume more battery than a screen without?

b) Does the map continue to update/redraw even when it is not the current screen?

c) Would removing all map screens from an activity conserve energy usage?


To clarify why I'm asking #1, I've had some issues with Apple's calibration, mainly that it often shorts distance by a decent amount. I've finally got it to a point where it is somewhat accurate and would like to keep it there. For #2, I often don't need the map feature and am looking to save some battery life as I'm pushing it close as is. I noticed it drained a bit quicker than the native Workout app, but that could have been other factors.

(Sorry if these have been asked elsewhere. I didn't see them in the last few pages, though. Thanks!)

 

[cfc]

A couple questions:

1. Does WOD contribute to Apple's calibration process? Or do only activities completed using the Workout app contribute?

2. a) Does using a screen with a map consume more battery than a screen without?

b) Does the map continue to update/redraw even when it is not the current screen?

c) Would removing all map screens from an activity conserve energy usage?


To clarify why I'm asking #1, I've had some issues with Apple's calibration, mainly that it often shorts distance by a decent amount. I've finally got it to a point where it is somewhat accurate and would like to keep it there. For #2, I often don't need the map feature and am looking to save some battery life as I'm pushing it close as is. I noticed it drained a bit quicker than the native Workout app, but that could have been other factors.

(Sorry if these have been asked elsewhere. I didn't see them in the last few pages, though. Thanks!)

I don't know if third party apps contribute to the calibration so my assumption is that they do not. Apple only mention using their app.

The map hardly affects battery usage directly. However people tend to look at the screen more often when it has a map (especially when first using the app), in which case the watch will use more power. But if you don't look any more frequently then it won't use noticeably more power.

Note that I say "hardly affects" and "noticeably more". This is because there will be a very slight extra power usage for the CPU and GPU to load and render the maps, but it is negligible compared to the hardware power draws from the sensors (HR, GPS), communications and the screen.

If the screen does not have a map then the app will not draw it. It only renders it "on demand" when you first display a screen with a map. So if you never display a map then it will never use any power to display it.

 

[jeon]

Is WOD configured to show GPS distances or Apple's calibrated distances?

The exercise recorded as fixed distance proceeded distance & average face source with apple. I changed it to gps now. I don't remember what settings I used before the update.

 

[jeon]

A couple questions:

1. Does WOD contribute to Apple's calibration process? Or do only activities completed using the Workout app contribute?

2. a) Does using a screen with a map consume more battery than a screen without?

b) Does the map continue to update/redraw even when it is not the current screen?

c) Would removing all map screens from an activity conserve energy usage?


To clarify why I'm asking #1, I've had some issues with Apple's calibration, mainly that it often shorts distance by a decent amount. I've finally got it to a point where it is somewhat accurate and would like to keep it there. For #2, I often don't need the map feature and am looking to save some battery life as I'm pushing it close as is. I noticed it drained a bit quicker than the native Workout app, but that could have been other factors.

(Sorry if these have been asked elsewhere. I didn't see them in the last few pages, though. Thanks!)

I feel the same symptoms. For example, if you run 10 kilometers, you will be recorded in the iPhone fitness app for 9.8 kilometers or shorter. Which is the more accurate distance?

 

[cfc]

Apple's fitness app always uses Apple's calibrated distances, even when WOD is using GPS, so WOD will show that you have done 10km according to the GPS (and will send this to Strava etc), but Apple will still show their calibrated distance.

Which is more accurate depends on whether the watch was well calibrated and how strong the GPS signal was.

 

[jeon]

...

Thank you for the explanation. I hope the distance measured by the WOD is more accurate. I trust my AW7 because I worked out with Garmin user and recorded the same distance as Garmin, but I am very concerned about the distance correction of FITNESS app.

 

[Cowboy8467]

Hey there @cfc , I just purchased your app and I'm going through the setting and everything right now. Looks very thorough and I'm looking forward to using it. One question I have and I apologize if this was answered earlier in the thread. I currently use a Polar H10 strap and I can't figure out if I'll gain anything by pairing it with the WOD app or if I can continue to pair it directly to my AW like I used to. This is the only external bluetooth sensor that I use so there is no issue with needing to have multiple sensors connected.

 

[Duffman19]

Thank you for the explanation. I hope the distance measured by the WOD is more accurate. I trust my AW7 because I worked out with Garmin user and recorded the same distance as Garmin, but I am very concerned about the distance correction of FITNESS app.

I will sometimes run with multiple devices for comparison's sake (fun!). Recently I've been using a Garmin 945 (my "base" for comparison as I've had it longest), a Coros Pace 2, and an AW7. My takeaway after a couple dozen or so runs is that Garmin and Apple apply quite a bit of algorithmic calculation to the recorded GNSS distance to come up with the recorded activity distance. At first, the AW was always shortest, followed by Garmin, and then Coros the longest. More recently, the Garmin has been short, followed by AW, and then Coros.

Interestingly, the Coros does NOT appear to apple any calibration to its recorded GNSS distance. That is, the GNSS distance is the activity distance (I use https://quantified-self.io for comparisons). Further, if I go back and use a tool like Komoot or https://onthegomap.com/ to calculate an activity's distance, the Coros is usually closest with the Garmin and AW coming up short.

Which is really correct? Who knows. Does it really matter? No. But I do find it interesting that the supposedly intelligent calibration process might actually be doing more harm than good. Knowing this, it may be best to use WOD's GPS distance option for most of your activities. At least then you'll have that to compare to Apple's calculated distance in the Fitness app.

 

[cfc]

Hey there @cfc , I just purchased your app and I'm going through the setting and everything right now. Looks very thorough and I'm looking forward to using it. One question I have and I apologize if this was answered earlier in the thread. I currently use a Polar H10 strap and I can't figure out if I'll gain anything by pairing it with the WOD app or if I can continue to pair it directly to my AW like I used to. This is the only external bluetooth sensor that I use so there is no issue with needing to have multiple sensors connected.

The only advantage of connecting via WOD is that you will get heart rate variability (HRV) data, which is not available if you connect via watchOS.

 

[Cowboy8467]

The only advantage of connecting via WOD is that you will get heart rate variability (HRV) data, which is not available if you connect via watchOS.

That makes sense, although I thought that HRV was only measured when at rest. Is that not the case?

 

[cfc]

That makes sense, although I thought that HRV was only measured when at rest. Is that not the case?

HRV is calculated from the time intervals between each beat, so can be calculated at any time. These "RR" intervals are provided by some HR sensors (including the H10) and the app calculates the HRV from them. For those who are interested the app determines the HRV by calculating the RMSSD over a 5 minute window.

 

[Cowboy8467]

Gotcha, thanks for the quick response. Was able to test it out on a bike ride this morning and the wealth of information that I can get from this is so much more than the default AW workout app. Great job!

 

[Duffman19]

One more Q for you, @cfc:

I've noticed that the elevation numbers in WOD seem a bit low. How are these figures calculated?

For some specifics, I've run the same trail 4 times over the last 2 days, each time wearing 3 devices, so 12 sets of data ('cause I had some time, I guess). The average elevation gained across all devices was 343 ft. This seams believable based on the terrain. However, using WOD for 3 loops gave an average of 257 ft.

I also ran the trail once using the stock Workout app and got 346 ft, right on average with the other devices.

Interestingly, if I export the WOD GPX files (as I was doing to compare tracks), they have numbers much closer to the average. Over the same 3 loops, the GPX average was 357 ft.

So is WOD applying some sort of "smoothing" when calculating elevation? Would you think this number would be more accurate than those from other devices or apps?

(Again, sorry for the barrage of info. I've probably been diving a bit too deep into this lately.)

 

[cfc]

One more Q for you, @cfc:

I've noticed that the elevation numbers in WOD seem a bit low. How are these figures calculated?

For some specifics, I've run the same trail 4 times over the last 2 days, each time wearing 3 devices, so 12 sets of data ('cause I had some time, I guess). The average elevation gained across all devices was 343 ft. This seams believable based on the terrain. However, using WOD for 3 loops gave an average of 257 ft.

I also ran the trail once using the stock Workout app and got 346 ft, right on average with the other devices.

Interestingly, if I export the WOD GPX files (as I was doing to compare tracks), they have numbers much closer to the average. Over the same 3 loops, the GPX average was 357 ft.

So is WOD applying some sort of "smoothing" when calculating elevation? Would you think this number would be more accurate than those from other devices or apps?

(Again, sorry for the barrage of info. I've probably been diving a bit too deep into this lately.)

Yes, all apps do some degree of smoothing to calculate the ascent and descent. Otherwise sensor variances would mean much higher values than expected. However the algorithm vary. If you load exactly the same GPX file into Garmin, Strava etc they can produce very different results due to the smoothing. This is particularly true on relatively flat routes where sensor variance makes more of a difference. With WOD I tried to produce something between Strava and Garmin, but it really depends on the route.

 

[jeon]

I tried to pair my friend's garmin running dynamic pod with WOD. However, I failed to pair both iPhone and AW7, and only another friend's garmin 245 watch itself was paired. How do I pair to RDP? I want to know the ground contact time and vertical amplitude. Can I see this information in the exercise record when I pair RDP or Stryd with WOD?I am trying to purchase one of the two, but I can't find a review of pairing with an updated WOD on the Internet.