Forerunner 965 - Recent Threads

FR965 28.05 — FIT data corruption: lap distances sum to 6.67 km, summary reports 19.90 km

6593293 — Fri, 24 Jul 2026 00:03:00 GMT

Controlled comparative test

Protocol: A single structured interval running session (14 × 400 m with recoveries) was recorded simultaneously by two devices worn on opposite wrists by the same athlete. The Fenix 5 Plus serves as the control device.

Course: Flat, open terrain. No tunnels, no dense canopy, no urban canyon.

3.1 Session-level comparison

Metric	Forerunner 965	Fenix 5 Plus	Deviation
Total distance (summary)	19.90 km	6.90 km	+188%
Sum of individual lap distances	6.67 km	6.91 km	—
Average pace	1:45 /km (34.4 km/h)	5:20 /km	physically impossible
Total ascent	3,015 m	33 m	+9,036%
Total descent	13,569 m	38 m	+35,608%
Elapsed time	34:45	36:49	consistent
Average heart rate	140 bpm	143 bpm	consistent
Max heart rate	170 bpm	172 bpm	consistent
Average cadence	147 spm	149 spm	consistent
Max cadence	203 spm	201 spm	consistent
Average temperature	24.5 °C	25.2 °C	consistent

Evidence that this is data corruption, not GNSS inaccuracy

Finding 1 — The file contradicts itself

The sum of all individual lap distances recorded by the Forerunner 965 is 6.67 km. The session summary in the same file reports 19.90 km.

The control device shows no such contradiction (lap sum 6.91 km vs. summary 6.90 km).

A GNSS reception problem produces positional jitter that inflates both the lap detail and the summary consistently. It cannot produce an internal contradiction between the detail records and the aggregate of the same file.

Notably, the Forerunner 965's own lap-level distance total (6.67 km) agrees with the control device (6.90 km) to within 3.3%.

Finding 2 — Corrupted timestamp records

The file contains impossible duration values:

Lap 1 elapsed time: 1193:01:49 (1,193 hours ≈ 49.7 days)
Lap 18 elapsed time: 1193:00:08
Cumulative time field escalates to: 2386:40:20
Session summary elapsed time: 34:45

Finding 3 — Barometric altimeter failure correlated with the same laps

Total recorded descent is 13,569 m — approximately 1.5× the elevation of Mount Everest above sea level — accumulated in 34 minutes on flat terrain where the control device recorded 38 m.

Critically, the laps containing corrupted timestamps are the same laps containing both zero distance and massive phantom elevation:

Lap	Recorded duration	Recorded descent
1	1193:01:49	7,906 m
3	0:05.2	2,013 m
18	1193:00:08	615 m

Additionally, total ascent (3,015 m) and total descent (13,569 m) differ by 10,554 m on a session that started and finished at the same location — physically impossible.

Finding 4 — The device contradicts its own inertial sensors

Using the Forerunner 965's own recorded accelerometer-derived values:

147 spm × 34.75 min × 1.39 m stride = 7,100 m ≈ 7.10 km

The device's own inertial data implies approximately 7.1 km, consistent with the control device (6.90 km) to within 3%. The device's GPS-derived summary reports 19.90 km.

Finding 5 — Downstream corruption of running power

Garmin's native running power is derived from pace, grade and athlete mass. With 13,569 m of phantom descent injected as grade data and pace inflated by a factor of three, running power output is necessarily invalid. This is consistent with the originally reported symptom of implausible power values, and supports a single upstream root cause rather than multiple independent sensor failures.

Root cause hypothesis

The evidence points to a sensor hub / barometric altimeter synchronization failure, not a GNSS receiver fault:

Distance dropouts, timestamp corruption and phantom elevation occur in the same laps, indicating a shared upstream failure.
The GNSS-derived per-lap distances are broadly plausible; only the aggregation and the barometric channel are corrupted.
Heart rate and cadence subsystems are unaffected.
The failure persists after a full factory reset, indicating it is not a corrupted user-configuration state.

I can provide .fit files from both activities to make the comparison

28.05 — FIT data corruption: lap distances sum to 6.67 km, activity summary reports 19.90 km

6593293 — Fri, 24 Jul 2026 00:23:30 GMT

Device: Forerunner 965 | Firmware: 28.05 | Age: 3 years, daily use, no prior issues

For the past several days my 965 has been recording grossly inflated pace, distance and elevation on outdoor runs, with a green GNSS fix confirmed before every start. Running power is also invalid.

I ran a controlled test to isolate it, and the data shows this is not a GNSS reception problem. The device is writing internally contradictory records into the FIT file.

The core finding

I ran a 14 × 400 m interval session on flat, open terrain wearing two watches, one on each wrist: the FR965 and a Fenix 5 Plus as control.

The sum of the FR965's own per-lap distances is 6.67 km. The activity summary in the same file reports 19.90 km.

The Fenix shows no such contradiction: lap sum 6.91 km, summary 6.90 km. And the FR965's own lap-level total (6.67 km) agrees with the Fenix to within 3.3%.

Poor GNSS reception inflates lap detail and summary together. It cannot make the detail and the aggregate of the same file disagree by 3x.

Side-by-side

Metric	FR965	Fenix 5 Plus
Total distance (summary)	19.90 km	6.90 km
Sum of lap distances	6.67 km	6.91 km
Average pace	1:45 /km (34.4 km/h)	5:20 /km
Total ascent	3,015 m	33 m
Total descent	13,569 m	38 m
Elapsed time	34:45	36:49
Avg / max HR	140 / 170	143 / 172
Avg / max cadence	147 / 203	149 / 201
Avg temperature	24.5 °C	25.2 °C

HR, cadence and temperature match closely across both devices — same athlete, same session, same effort. The failure is isolated to the distance/altimetry subsystem on the 965.

An average pace of 1:45 /km is 34.4 km/h sustained for 35 minutes. That is roughly 60% faster than the world hour record for running.

Interval accuracy

Every rep was a fixed 400 m.

Fenix 5 Plus: 0.40 km on all 14 reps.
FR965: 0.00, 0.00, 0.26, 0.22, 0.24, 0.33, 0.25, 0.33, 0.00, 0.07, 0.29, 0.30, 0.34, 0.35 km

Three reps recorded zero distance while HR and cadence in the same laps confirm I was running.

Three signatures of corruption, not inaccuracy

1. Impossible timestamps. Lap 1 elapsed time reads 1193:01:49 — 1,193 hours, roughly 50 days. Lap 18 reads 1193:00:08. Cumulative time escalates to 2386:40:20. The summary says 34:45.

2. Altimeter failure correlated with the same laps. 13,569 m of descent in 34 minutes on flat ground, where the Fenix recorded 38 m. Ascent and descent differ by 10,554 m on a session that started and finished at the same point. And the laps with corrupted timestamps are the same laps with zero distance and phantom elevation:

Lap	Duration	Descent
1	1193:01:49	7,906 m
3	0:05.2	2,013 m
18	1193:00:08	615 m

3. The watch contradicts its own accelerometer. Using the 965's own recorded values: 147 spm × 34.75 min × 1.39 m stride = 7.10 km. Its inertial data implies ~7.1 km, matching the Fenix within 3%. Its summary says 19.90 km.

Why running power is also broken

Native running power is derived from pace, grade and mass. With 13,569 m of phantom descent feeding the grade term and pace inflated 3x, power output is necessarily garbage. One upstream fault explains all four symptoms — distance, pace, elevation and power — rather than four independent sensor failures.

Root cause hypothesis

This looks like a sensor hub / barometric altimeter synchronisation failure, not a GNSS receiver fault. Distance dropouts, timestamp corruption and phantom elevation all occur in the same laps. The GNSS per-lap distances are broadly plausible; it is the aggregation and the barometric channel that are corrupted. HR and cadence are unaffected. And it survives a full factory reset, so it is not corrupted user configuration.

Troubleshooting already completed — issue persists through all of it

Confirmed 28.05 is the latest public stable release — nothing newer to install
Removed all obsolete paired sensors; confirmed no external sensor set as pace/distance source
Satellite mode forced to Multi-Band / All Systems, not Auto Select
20-minute GNSS soak, stationary outdoors, activity open and unstarted
Ephemeris/CPE refresh via Garmin Connect sync
Soft reset (Light button, 15 s)
Full factory reset — settings and data erase, full reconfiguration, re-sync
Side-by-side controlled test against a second device

Questions

Is this a known defect on 27.09 / 28.05? There are several open threads on this platform reporting pacing and tracking anomalies since 27.09 rolled out in January.
Can a moderator route the FIT files to engineering? I have the original .FIT for both watches from the same session and can send them on request.
Has anyone else seen a lap-sum vs. summary mismatch specifically? That is the part I have not seen reported elsewhere, and I think it is the most diagnostic symptom. If you have had bad distance data on 28.05, it is worth exporting your CSV and checking whether your lap distances add up to your total.

Happy to run any additional test that would help narrow this down.

Rallier un WP en ligne droite

3874348 — Fri, 17 Jul 2026 18:27:17 GMT

Bonjour à tous,

Je cherche désespérément un moyen pour que ma garmin forerunner 965 puisse me fournir une direction (directe= "à vol d'oiseau") en m'indiquant la distance. C'est une fonction que j'utilisais avec ma fénix et que j'adorais. Si quelqu'un a une solution, je serai le plus heureux.

Pierre

RE: Rallier un WP en ligne droite

3874348 — Mon, 20 Jul 2026 14:38:13 GMT

Thank you. I hope Garmin will see this.

RE: Rallier un WP en ligne droite

Bit — Mon, 20 Jul 2026 13:06:06 GMT

Regrettably, the "Direct Route" option isn't available on the Forerunner 955 I have, probably not on the 965 either.
I tested as follows: Select Walk activity, Navigation, Saved Locations, <Location>, Go To.
Not much, but there is still the 'Heading Bug' though.

RE: VO2 max mismatch on watch UI

Jacquers — Sat, 18 Jul 2026 14:14:42 GMT

Could one be for running and the other for cycling?

VO2 max mismatch on watch UI

Prateek — Fri, 17 Jul 2026 17:45:40 GMT

The VO2 max dashboard shows 43 on watch as well as garmin connect app (for cycling), but when I pan the chart, at tip it is showing 50 (lowest point is 47).
Why is there a discrepancy and which is the correct/more reliable number?

Issue with elevation during workout

2104482 — Thu, 09 Jul 2026 07:04:06 GMT

I have had my Forerunner for a few months now and had no issues. But for about the last 3-4 weeks whenever I record an activity - hiking, walking, running or trail running it isn't measuring my elevation.

If there is no workout being recorded it is still measuring it correctly.

I have tried all the normal things - reset, calibration, submerged in water to check the sensor isnt blocked etc.

Anyone got any ideas please?

RE: Issue with elevation during workout

Jacquers — Sat, 18 Jul 2026 14:10:41 GMT

Does elevation work during general use, when looking at barometer glance / widget?

RE: Issue with elevation during workout

5023312 — Sat, 18 Jul 2026 10:41:13 GMT

Same problem for me and my Foreunner 255.

RE: Bosch Live Data Interface

RichardB — Fri, 17 Jul 2026 09:43:38 GMT

www.bike-magazin.de/.../

Bosch Live Data Interface

rene84828 — Wed, 15 Jul 2026 20:46:12 GMT

Hello!

Will the Forerunner also be able to use the data from Bosch Live Data Interface?

problemas de cronometro/horas tras actualizar v 28.05

2270138 — Thu, 16 Jul 2026 22:08:52 GMT

tengo problemas con mi forerunner 965, despues de la actualizacion 28.05, sincronizo la hora ya sea manual/ automaticamente y al pasar los minutos se va perdiendo la hora, es decir la hora real en mi telefono marcha bien pero la del reloj se atrasa, esto sucede tambien con el cronometro y en las actividades el cronometro tambien pierde los segundos/minutos y al cabo de unas horas existe una diferencia significativa de minutos, incluso horas

RE: Bosch Live Data Interface

Bitti — Thu, 16 Jul 2026 18:19:37 GMT

If you are talking about Forerunner 965, it's very unlikely that new features would be added to it any more.

Request: Set training every other day for 14 days

Enduring2111 — Tue, 14 Jul 2026 19:52:37 GMT

I can now only set a training day Monday to Sunday, but I want to train every other day. With a 14-day choice, I can set: Monday, Wednesday, Friday, Sunday, Tuesday, Thursday and Saturday. Then the cycle starts again on Monday. I also want to set the days on which I want to do a long run. Can this 14-day option be added?

Unified Training Status

pnmckenna — Tue, 14 Jul 2026 05:01:27 GMT

is anyone else having issues with Unified Training Status?

since getting an Edge 850 i've had all kinds of weirdness going on with my training status metrics. My best guess is that Garmin has changed something with unified training status which effectively no longer supports the FR965 (probably triggered by the E850).

everything used to work nicely with my E530 but now its like the different devices are talking different languages to each other and the result is a mess.

i've had a support ticket open for 8 months now and the best they can say is "updates to improve unified training are being further reviewed on a long term basis with our engineers" so i'm trying to understand what is happening and how i might be able to get past it.

Need help fixing pool swim file

mweber01 — Fri, 10 Jul 2026 14:45:13 GMT

I did a pool swim today and my fit file is all messed up somehow. It's showing duplicates for my laps when I analyze the file on my computer. I've done the basic corrupt file fixing and nothing is fixing it. Any smart people out there that can help fix my file?

RE: Need help fixing pool swim file

mweber01 — Fri, 10 Jul 2026 19:06:55 GMT

It works! That's so weird. I wonder why it didn't prompt me with that? I can't thank you enough! Now I'm wondering what exactly that FIT File Viewer fixed... I'll drop both files in something like Beyond Compare and see what the differences are. Thanks again!

RE: Need help fixing pool swim file

seilogramp — Fri, 10 Jul 2026 19:04:03 GMT

forums.garmin.com/.../2026_2D00_07_2D00_10_2D00_06_2D00_51_2D00_54_5F00_fixed.fit

RE: Need help fixing pool swim file

mweber01 — Fri, 10 Jul 2026 19:03:13 GMT

That's so strange. When I did it, it just loaded the file up and didn't give me the notification that it was fixed.... Can you send me that fixed file?

RE: Need help fixing pool swim file

seilogramp — Fri, 10 Jul 2026 19:00:04 GMT

Open fit file viewer >> click on "Open FIT file" in the top/right corner of the page >> browse for and upload your fit file >> click on "Download fixed FIT file."

RE: Need help fixing pool swim file

mweber01 — Fri, 10 Jul 2026 17:33:23 GMT

How do you repair a file using FIT File Viewer? I only see how you can upload a file, then download individual CSV files for each individual record?

RE: Need help fixing pool swim file

seilogramp — Fri, 10 Jul 2026 16:37:53 GMT

I used FIT File Viewer to repair the file and successfully tested it by uploading it to my Garmin account. I can’t say for sure if the laps need adjusting, but from this non-swimmer’s perspective, they look good.

RE: Need help fixing pool swim file

mweber01 — Fri, 10 Jul 2026 16:16:27 GMT

Thanks! So, Connect doesn't recognize it. And the issue isn't just as simple as duplicate laps, I think there's more to it but I'm not sure... On my watch it shows the proper distance, but in SwimData.org it shows it's duplicated. It's strange.

Here's the Dropbox link for the file: https://www.dropbox.com/scl/fi/76vdhbvzvmfln7dtg3c2r/2026-07-10-06-51-54.fit?rlkey=ohr56wh2zglthxd99modmlv4w&dl=0

Thanks for the help!

RE: Need help fixing pool swim file

seilogramp — Fri, 10 Jul 2026 15:48:30 GMT

Technically, if Garmin Connect accepts it, then it’s not corrupt, so most online tools won’t have anything to fix. However, some tools, such as FFRT, do offer the option to edit laps.

Zip the .fit file, then in a reply post Insert >> Image/video/file >> click Upload >> browse for file >> OK.
Or upload the fit file to a file sharing service, such as OneDrive, Google Drive, Dropbox, etc., make the file public, then paste a link to the file in your post.