Transmit power data from Skierg PM5 to Garmin Epix2

[jorgensor]

Have connected my Skierg PM5 to my Garmin Epix watch.
I get all data except the most important- the power/speed.
As I understood it is som flag in the protocol that indicates that the data is from a virtual skier - and that the Garmin watch cannot interpret that (probably an easy fix for Garmin)
Anybody knows any way to fix this yourself w/o using 3rd party apps?
Can I for example change the flag to ”rower” (which probably will work)?
I don’t want to sync C2 loggbook to Garmin since you don’t get the complete picture of of that.

[Sakly]

Recognized the same, no solution.
It should be really easy for Garmin to fix that, I have no idea why they don't have an indoor ski activity to connect to a trainer.

[gvcormac]

I use ErgData to upload data to log.concept2.com, which I have configured to forward to connect.garmin.com I don't use the watch at all (for ergs).

That satisfies my need for a detailed log. Not sure if it would satisfy yours.

[JaapvanE]

As I understood it is som flag in the protocol that indicates that the data is from a virtual skier - and that the Garmin watch cannot interpret that (probably an easy fix for Garmin)
Anybody knows any way to fix this yourself w/o using 3rd party apps?
Can I for example change the flag to ”rower” (which probably will work)?

It isn't some flag. In ANT+ (see thisisant.com) there are official specifications each device must adhere to. Each sport has its own pages and definitions, and both Garmin and Concept2 must adhere to them without exception for things to work. As do other manufacturers (I am the lead developer of OpenRowingMonitor, and we implemented this ANT+ spec as well). Companies do not get their ANT+ certification if they only loosely follow the specs. Adding power or pace to a skierg will require a formal change of these specifications and both parties implementing these pages on their devices.

I don't know if a SkiErg PM5 is capable of behaving like a rower. As some of the physics might be different (for asymmetric pushes), it probably wouldn't work.

[jorgensor]

Thanks for great replies! Since both Garmin and Concept2 got this working for both Rowing and Cycling I could not imagine it would be very hard to get this going if they really want. Understand that they need to adhere to the ANT+ protocol specs ...

[HornetMaX]

And in case garmin and concept2 finally decides to fix that, please also have a look at the 'handling of rest intervals" which is currently very broken, even on a RowErg (IIRC, that may be more a concept2 issue than a garmin one, still annoying).

[JaapvanE]

And in case garmin and concept2 finally decides to fix that, please also have a look at the 'handling of rest intervals" which is currently very broken, even on a RowErg (IIRC, that may be more a concept2 issue than a garmin one, still annoying).

Development on ANT+ has ceased, and Garmin will not add new sports.

Handling rest intervals is quite a nasty beast. I see the same within fit-files (which essentially are recorded ANT+ messages). The key issue is that you can flag an interval as 'rest', but then you explicitly are not allowed to put any progress reports in the rest interval. But to 'zero' pace, power, etc. you really want to. In essence, you want the last position report in the active interval to have these fields filled, and the first in the rest interval to be zeroed.

Easiest approach for Garmin/Strava/etc. is to respect that a rest interval indeed is a rest interval, and zero all active parameters on their side.

[HornetMaX]

And in case garmin and concept2 finally decides to fix that, please also have a look at the 'handling of rest intervals" which is currently very broken, even on a RowErg (IIRC, that may be more a concept2 issue than a garmin one, still annoying).

Development on ANT+ has ceased, and Garmin will not add new sports.

Why not ? They already have some crazy "sports" in their list (uh, gaming ?!) ...

Handling rest intervals is quite a nasty beast. I see the same within fit-files (which essentially are recorded ANT+ messages). The key issue is that you can flag an interval as 'rest', but then you explicitly are not allowed to put any progress reports in the rest interval. But to 'zero' pace, power, etc. you really want to. In essence, you want the last position report in the active interval to have these fields filled, and the first in the rest interval to be zeroed.

Easiest approach for Garmin/Strava/etc. is to respect that a rest interval indeed is a rest interval, and zero all active parameters on their side.

Makes sense to me. I just don't understand why nobody seems to care: seems like a big fat bug (given it also apply to indoor rowing).

[Sakly]

Handling rest intervals is quite a nasty beast. I see the same within fit-files (which essentially are recorded ANT+ messages). The key issue is that you can flag an interval as 'rest', but then you explicitly are not allowed to put any progress reports in the rest interval. But to 'zero' pace, power, etc. you really want to. In essence, you want the last position report in the active interval to have these fields filled, and the first in the rest interval to be zeroed.

Easiest approach for Garmin/Strava/etc. is to respect that a rest interval indeed is a rest interval, and zero all active parameters on their side.

That doesn't make sense to me. A rest interval is meant to have no serious output in terms of the workout goal, but the user can decide to stop completely or do some light paddling (for the example of rowing). The most sense would make to update the data during rest periods based on the user action. If some strokes are detected, send data accordingly. If no stroke is detected anymore, send zeros. Update the data as it was a normal work interval, so all metrics would appear with valid data.
Of course, averages for power and pace would include the rest intervals (as they do already), but the calculations would match the reality.

[JaapvanE]

Development on ANT+ has ceased, and Garmin will not add new sports.

Why not ? They already have some crazy "sports" in their list (uh, gaming ?!) ...

These do not use ANT+ to get the data for these activities. ANT+ is abandoned as the EU requires data about health to be encrypted upon user request, and ANT+ as a protocol can't fulfill that requirement. But most gym and home exercise equipment is ANT+ based, so it is going to be a challenging change.

They key problem is that Garmin wants/must move to BLE variants, and a PM5 currently only supports 1 BLE connection. And some people want to connect to their logbook or a game like EXR AND record on a watch. The protocol and most hardware can support much more connections, but it requires work as only one should be allowed to control training parameters (like target distance, ect.).

Easiest approach for Garmin/Strava/etc. is to respect that a rest interval indeed is a rest interval, and zero all active parameters on their side.

Makes sense to me. I just don't understand why nobody seems to care: seems like a big fat bug (given it also apply to indoor rowing).

It is, but it essentially is cosmetic as in a FIT-file you must rely on the reported totals that are part of the FIT-file, not on what you assume happened. Everybody, aside Strava, respects that (that is why max speeds in rowing are sometimes way off in Strava). So the gradually decreasing pace during a pause might look funny, but the totals are not affected in decent tools.

That doesn't make sense to me. A rest interval is meant to have no serious output in terms of the workout goal, but the user can decide to stop completely or do some light paddling (for the example of rowing). The most sense would make to update the data during rest periods based on the user action. If some strokes are detected, send data accordingly. If no stroke is detected anymore, send zeros. Update the data as it was a normal work interval, so all metrics would appear with valid data.

That is not how Garmin specifies it. And they determine how the spec works.

[Sakly]

That doesn't make sense to me. A rest interval is meant to have no serious output in terms of the workout goal, but the user can decide to stop completely or do some light paddling (for the example of rowing). The most sense would make to update the data during rest periods based on the user action. If some strokes are detected, send data accordingly. If no stroke is detected anymore, send zeros. Update the data as it was a normal work interval, so all metrics would appear with valid data.

That is not how Garmin specifies it. And they determine how the spec works.

Sure, but that does not make it any better. A specification could be changed, if it does not fit the needs of a useful metric.
Anyway, the outcome of the workout is not changed by that, luckily

[HornetMaX]

Development on ANT+ has ceased, and Garmin will not add new sports.

Why not ? They already have some crazy "sports" in their list (uh, gaming ?!) ...

These do not use ANT+ to get the data for these activities. ANT+ is abandoned as the EU requires data about health to be encrypted upon user request, and ANT+ as a protocol can't fulfill that requirement. But most gym and home exercise equipment is ANT+ based, so it is going to be a challenging change.

They key problem is that Garmin wants/must move to BLE variants, and a PM5 currently only supports 1 BLE connection. And some people want to connect to their logbook or a game like EXR AND record on a watch. The protocol and most hardware can support much more connections, but it requires work as only one should be allowed to control training parameters (like target distance, ect.).

Hmmm, but then why Garmin still has ANT+ (so unencrypted) on it's newest models (just released stuff, like the Instinct 3) ?
They have it for the RowErg but their justification for not having it for the SkiErg is "Sorry, can't because of EU" ?!

It is, but it essentially is cosmetic as in a FIT-file you must rely on the reported totals that are part of the FIT-file, not on what you assume happened. Everybody, aside Strava, respects that (that is why max speeds in rowing are sometimes way off in Strava). So the gradually decreasing pace during a pause might look funny, but the totals are not affected in decent tools.

I can't remember exactly as I stopped logging on the watch at all (extra hassle for little gain, IMO) but I think the fact the PM5 was not zeroing the power made Garmin consider that during your rest you have output something like the average of your last stroke before the rest and the 1st stroke after that. It wasn't just a graphical weirdness.

Also, I thought FIT files are only involved when you don't log on the watch ](i.e. when c2 logbook forwards to Garmin Connect), no ?

[JaapvanE]

Hmmm, but then why Garmin still has ANT+ (so unencrypted) on it's newest models (just released stuff, like the Instinct 3) ?
They have it for the RowErg but their justification for not having it for the SkiErg is "Sorry, can't because of EU" ?!

The are phased out gradually, but you also can't ignore the huge installed base of RowErg's, Treadmills, Bikes, etc.. But Garmin prefers encrypted BLE now, especially for HRM data.

I can't remember exactly as I stopped logging on the watch at all (extra hassle for little gain, IMO) but I think the fact the PM5 was not zeroing the power made Garmin consider that during your rest you have output something like the average of your last stroke before the rest and the 1st stroke after that. It wasn't just a graphical weirdness.

Yeah, I use OpenRowingMonitor to connect to my Garmin, and then it isn't an issue (we always zero data after a stop). I like seeing the HRM and breathing data, as well as the distance per stroke, as it clearly shows where I started to struggle and left Zone 2.

Also, I thought FIT files are only involved when you don't log on the watch ](i.e. when c2 logbook forwards to Garmin Connect), no ?

The whole Garmin ecosystem is built upon fit-files. Internally, Garmin watches also log in fit-files. Those files are essentially chronological recordings of the messages the watch received. Workouts are also described using FIT-files. The FIT-SDK is actively being developed, and is much data richer that its predecessor (the tcx file) and much more compact (roughly 20 times).

[HornetMaX]

Hmmm, but then why Garmin still has ANT+ (so unencrypted) on it's newest models (just released stuff, like the Instinct 3) ?
They have it for the RowErg but their justification for not having it for the SkiErg is "Sorry, can't because of EU" ?!

The are phased out gradually, but you also can't ignore the huge installed base of RowErg's, Treadmills, Bikes, etc.. But Garmin prefers encrypted BLE now, especially for HRM data.

Well, they seem to happily ignore the installed base of SkiErg

And yes, the Garmin could get HR data via BLE (EU compliant) and rest from the PM5 via ANT.
The PM5 could (at the same time) get the HR data from the HRM via ANT. This is already possible.

Also, I thought FIT files are only involved when you don't log on the watch ](i.e. when c2 logbook forwards to Garmin Connect), no ?

The whole Garmin ecosystem is built upon fit-files. Internally, Garmin watches also log in fit-files. Those files are essentially chronological recordings of the messages the watch received. Workouts are also described using FIT-files. The FIT-SDK is actively being developed, and is much data richer that its predecessor (the tcx file) and much more compact (roughly 20 times).

Sure, but that's not what goes on between the PM5 and the Garmin watch no ?
And my understanding is that at least part of the issue is there (the way the PM5 sends out data for rest intervals is not what Garmin seems to expect).

[JaapvanE]

And yes, the Garmin could get HR data via BLE (EU compliant) and rest from the PM5 via ANT.
The PM5 could (at the same time) get the HR data from the HRM via ANT. This is already possible.

The ANT+ protocol contains a HRM field, as Garmin strived for a single connection. In developing OpenRowingMonitor we decided against including that field as HRM data directly to the watch probably was more accurate. But strictly speaking, nothing is stopping us from flipping the capabilities field and injecting it in the stream. Injecting HRM would actually clean up the resulting FIT-file and remove a lot of overhead as it becomes an attribute of a report, instead of an independent report (as ORM also creates its own FIT-files, we are quite interested in the differences).

The key issue is that the EU requires that for new devices the user can select to encrypt any possibly sensitive data (see https://www.dcrainmaker.com/2025/01/the ... eless.html for a more detailed explanation). Garmin has to comply by summer, and simply decided that ANT+ can't comply without breaking backward compatibility, so they kill the entire ANT+ protocol (slowly).

Sure, but that's not what goes on between the PM5 and the Garmin watch no ?
And my understanding is that at least part of the issue is there (the way the PM5 sends out data for rest intervals is not what Garmin seems to expect).

I honestly haven't looked at what a PM5 broadcasts during a rest interval. I know what they should send, and what we do send, but a Garmin watch simply ignores key flags. For example, in ORM we explicitly raise a "Start" and a "Finished" flag, and the ANT+-protocol describes those flags should automatically trigger a start and stop of the activity on your watch. Completely ignored, never implemented for over a decade. But from a C2 perspective, that is what you should do and the only thing you can do. There are actually quite valid reasons to keep reporting the last reported power and pace, as that is the thing you report when crossing the finishline and due to lags in reporting and processing, you want to capture that last report for sure (ANT+ is far from reliably when human bodies are interfering). With ORM we used a more practical approach and balance on a tight rope where we looked at how a lot of watches react to zero-ing that data in a follow-up report, but as soon as some watchmaker has a completely different approach, we are in trouble.

But Garmin's software development and QA department is a complete mess for years, so fixing these issues in a protocol that is going to be abandoned, isn't going to happen.