Garmin Swim too sensitive, counting far too many lengths

Hi Gerome,

Have you seen this FAQ? Some of the tips there may help you. In addition, we often find that after 4 or 5 swims with the watch, problems with over-counting lengths disappear.

Ok thanks for your feedback. I was at 4 swims... but I did a reset to see if the situation improved. It happens to be the same (1800 m instead of 1500 m). i'll stick to what you advise in the FAQ and come back in 2 or 3 weeks to update you!

I also have a new Garmin Swim and I am a weak swimmer. On my first swim it over counted 3 lengths, which I am happy to put down to learning curve etc. Having read the FAQ etc I now make a point of avoiding having to slow up for congestion. My second swim it still over counted by one length, no congestion so was a bit mystified. Third swim tonight and again it over counted by one length. Again no slowing up for congestion, but has raised a question mark in my mind that it might have been caused by a swimmer coming in the opposite direction in the same lane. i.e. buffeting in water due to swimmer in the opposite direction. I am now going to watch out for this in future swims to see if this may account for the extra length. Like others have said I like the concept but at present the watch is let down by the software. i.e. Firmware counting extra lengths and inability to combine two erroneous lengths on website. A number of people have asked me about the watch at the pool, but I am advising them not to buy at present and wait six months or so to see if the software improves.

My Garmin Swim sometimes overcounts too. I do freestyle only, and I try to do a controlled, non-splashing stroke. I think the Swim sometimes takes my short half-second pause between each half-stroke to be a turn. I have tried swimming faster, with shorter pauses, and that does make a difference. Then, the Swim is completely accurate.
I read on this forum that the Swim takes some time getting used to the swimmer's technique..so I do hope that in the end it will adapt to my technique.
Alternatively, I would not mind occasionally combining two lengths to one on Garmin Connect, but there's no way to do that..

After 5 sessions, in all of them some overcounts happened, from 2 to 5 per session. Looking at my data, all of them happened during (real) backstrokes lengths. Not only backstroke detection got problematic after v.3.2, but looks like it is more prone to overcount errors.

Hi,

I would like to encourage you to take a look at the Fit File Repair Tool (http://fitfilerepairtool.info) which is used by more than 100 people. This tool does not only repair files but also enables you to edit swim data files - for instance correct the number of lengths (by merging and splitting lengths), the stroke style etc. Some examples on how to edit swim files can be found here: http://fitfilerepairtool.info/quick-tour/edit-a-swim-file/

Regards
Mathias

Fit File Repair - Looks good but it costs money and is not cheap for something that is really a Garmin problem.
Given the speed ( Or more accurately total lack of speed) of Garmin software development it would be good if they
employed somebody like Mathias. Given Fit File Repair is a Windows application, how transferable are the lessons learnt
to a web solution. Does Mathius have the relevant web programming skills?

To the OP (GBILLOIS)... Where are you reading the number of lengths? From the graphs on Garmin Connect? If so, I just discovered that the "Swim Graph" length count is correct but the 3 graphs below it have the wrong count if length is selected on the horizontal axis. Apparently, the 3 analysis graphs count each rest or pause as a length!
Could this be the problem?

The latest software update makes the watch less sensitive. I would recommend updating your watches here.

The latest software update makes the watch less sensitive. I would recommend updating your watches here.

Good news! I'll try this week-end, the problems still happened after 5 sessions, I hope the new firmware will correct that.

Thanks,