--
You received this message because you are subscribed to a topic in the Google Groups "Subsurface Divelog" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/subsurface-divelog/5640mbRaxQE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to subsurface-dive...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/951fd3df-2581-46fc-806f-5fa36f4ca8c1n%40googlegroups.com.
Hi Mikel and Gareth.
Yes, that also worked for me. Thanks!!!!
El viernes, 16 de agosto de 2024 a las 17:06:06 UTC+2, Gareth Morris escribió:
I noticed what I think may be the same thing — when first importing a dive, the dive profile is correct, but the dive summary time is wrong; in my case around 10 mins for a 50 minute dive.
After closing / re-opening the app (I believe just closing / re-opening the logbook is sufficient), the dive summary time refreshes and is now showing correctly for me, perhaps that will work for you too.
On Fri, Aug 16, 2024 at 7:58 AM Mikel Larrañaga <mikel...@gmail.com> wrote:
HI,
I've managed to import FIT files exported in the suunto app (recorded with suunto Ocean) but realized that the dive duration is wrong, Tried with several dives (40-60 minutes). All the dive profile is correctly imported but the imported dive duration is 8-12 minutes, which affects to the statistics (air consume, average depth, etc.)
I have got a sneaky suspicion for what might be happening there - it is likely that Garmin's recent changes to not require entries in a FIT file to be chronologically ordered (discussed here: https://the5krunner.com/2024/02/12/garmin-fit-files-welcomed-improvements-for-2023-4-is-there-a-downside/) are tripping us up, as Subsurface still expects all entries to be chronologically ordered.
I had done a 'kneejerk fix' for new format files produced by
Garmin devices after recent firmware updates
(https://github.com/subsurface/libdc/pull/58/commits/bb502c6d8b53153b6f43469f55f290e470efc105),
but what this does is just completely ignore message types we do
not use in the import, which happens to make it work for FIT files
produced by Garmin devices.
But it looks like the Suunto FIT files need actual reordering of
(not ignored) messages before the import.
Is anyone able to share a FIT file showing this behaviour with me so that I can investigate a bit more?
Ngā mihi
Michael Keller
Hi all.
On Sat, 10 Aug 2024 at 08:52, Gareth Morris <gareth...@gmail.com> wrote:One thing I noticed is the dive is showing a duration of 10 minutes, both in the dive list and the Summary tab -- whereas the dive was actually 48mins long.Ok, both you and Mikel noticed this, and today I did my first dives with this thing, and yes, the duration is off by a factor of 5 on initial import.
You beat me to it, Linus. ;-)
I started looking into the FIT file last night, and saw the samples every second that only contain temperature information. But since I was looking for 'out of order' samples I did not make much of it.
The fix was trivial, and I've done a pull request for it.
Looks good to me.
I'll look at also making the FIT file importer not do pointless empty samples, but the whole "no depth" is actually real, in that it could be a sample with no depth but with tank pressure information. So the FIT file import wasn't *wrong*, it was just a bit silly.
I think what the FIT file importer does makes sense actually - create a sample, use -1 for 'no depth', and then expect Subsurface to throw away the sample but retain an event that contains the additional information in the sample.
If you want to look into the FIT importer, one more thing that is bugging me is that with Garmin's FIT 2.1 spec, they explicitly state that log entries no longer have to be chronologically ordered (see https://the5krunner.com/2024/02/12/garmin-fit-files-welcomed-improvements-for-2023-4-is-there-a-downside/).
Garmin started to make use of this format in a firmware update for their Descent Mk2 / Mk3 devices. Luckily for us, they are only using out-of-order events for data that we are not interested in, so I was able to fix this by dropping these events in the importer in https://github.com/subsurface/libdc/pull/58/commits/bb502c6d8b53153b6f43469f55f290e470efc105. But I suspect that sooner or later they might be sending fields that we care about out of order, at which point the current importer will break. The way around this will be to add chronological sorting of events and samples, either in the FIT importer, or in the generic importer code inside Subsurface proper.
Cheers
Michael Keller