1 sec vs. Smart recording - is there a point?

Garmin Connect certainly does "smart" rendering of your track.

However, on Strava, I can move through my run a second at a time and see the position changing ever second.

On Strava, another "good" reason to use every second is for more accurate segment times although it is fairly well known that you can actually "gain" a few seconds with a smart recorded track.

In terms of the accuracy of the data the watch gives you there is no difference between the two modes. This is a common misconception. GPS and HR etc is still "sampled" every second and the watch processes it. It just not recorded either second with smart recording.

1-sec recording logs more datapoints, which should produce more accurate GPS maps.
Smart recording logs one one datapoint per 5-6secs, possibly by taking some kind of average. This reduces the logsize, and should also smooth out any erratic behaviour.
(Paraphrased from DCRainmaker.com's excellent comprehensive review of this Forerunner 235 model).
You raise a good point that perhaps its the software (GarminConnect, Strava, etc) that may also introduce their own algorithms to draw a smoother GSP route....

Garmin Connect certainly does "smart" rendering of your track.

However, on Strava, I can move through my run a second at a time and see the position changing ever second.

On Strava, another "good" reason to use every second is for more accurate segment times although it is fairly well known that you can actually "gain" a few seconds with a smart recorded track.

In terms of the accuracy of the data the watch gives you there is no difference between the two modes. This is a common misconception. GPS and HR etc is still "sampled" every second and the watch processes it. It just not recorded either second with smart recording.

This is exactly why I'm raising the question.

I manage a local parkrun and have created a few segments on there for people to get competitive over, however the segments have lots of straight lines that cut right over corners. I was hoping to recreate those segments once I'd had a chance to rerun the route using 1 sec recording, however on some recent races using 1 sec the maps kicked out of Strava don't look any different with huge straight lines still there. I have double checked I am on 1 sec and not smart so had wondered how Strava was rendering the map.

Appreciate the actual accuracy in terms of time and distance is not changed with 1 sec recording, it is just the smoothing of the map that I want to get a more accurate path for the segment to take.

If Strava uses the data taken from Garmin Connect, and Garmin does some "smart" rendering before publication, is the Strava data definitely not amended before it is made available too?

Strava just gets your activity FIT file and indeed so does GC so whatever Strava does with it is due to their processing algorithms and similar with GC and there's no co-dependency.

There is a fair bit of tolerance with Strava segment matching due to the small variances you will see in all GPS tracks.
Perhaps you can post some examples so we can better understand what is going on. It is a UK parkrun I might be able to get there myself one day :)

Now and again I have "drawn" a route in some mapping tool and tried to create a pseudo activity from it to make "good" Strava segments. One example was at my local running track where wanted to create all sorts of segments of differing numbers of laps. Most of the time however doing it from a run is not only far easier but generally works well enough.

I found the HRM performed poorly on hill sprints with Smart recording (default option) turned on - smoothed the HR massively and frequently peaked out at 145bpm when I know it should have been ~175bpm. Since I've switched to 'record every second' the performance on intervals has been solid.

I found the HRM performed poorly on hill sprints with Smart recording (default option) turned on - smoothed the HR massively and frequently peaked out at 145bpm when I know it should have been ~175bpm. Since I've switched to 'record every second' the performance on intervals has been solid.

Hmm. How do yo know that Smart recording affects HRM at all? I've heard that it affects GPS only.

Smart recording saves a data point when speed/pace, direction, elevation or heart rate changes. Not sure of the threshold that triggers the decision to keep or discard. The device always records at 1-second intervals. Smart recording just decides which points to keep and which to discard.

support.garmin.com/.../case.faces

https://forums.garmin.com/showthread.php?20550-Smart-Recording-vs-1-second-recording

Just to add to this discussion, i've found little difference in the quality of the data collected between 1sec and smart recording. Many times the smart recording has given more accurate tracks / nicer looking HR curves for an identically run course. Most times the differences are within the "noise".

One thing that did strike me as interesting, for EXACTLY the same 1hr 2min run, same course and pace: 1sec recording (yesterday) used 10.9% battery, while smart recording (today) used 7.7% battery (for both GPS, HR, BT and smart track on). Thats a big difference if activities are the dominant drain on your battery (e.g. running >1 hour most days), and could amount to getting an extra few days per charge.

I'm doing more testing, including multiple identical runs to account for the run-to-run variance in the numbers. I'll post these in a week or so.

This is exactly why I'm raising the question.

I manage a local parkrun and have created a few segments on there for people to get competitive over, however the segments have lots of straight lines that cut right over corners. I was hoping to recreate those segments once I'd had a chance to rerun the route using 1 sec recording, however on some recent races using 1 sec the maps kicked out of Strava don't look any different with huge straight lines still there. I have double checked I am on 1 sec and not smart so had wondered how Strava was rendering the map.

Appreciate the actual accuracy in terms of time and distance is not changed with 1 sec recording, it is just the smoothing of the map that I want to get a more accurate path for the segment to take.

If Strava uses the data taken from Garmin Connect, and Garmin does some "smart" rendering before publication, is the Strava data definitely not amended before it is made available too?

Maybe DRAW the proposed segment with MapMyRun or similar, export the file, import to Strava as an activity and create the segment?

I did an online support chat with Garmin today. The guy (Chris) told me that I change my Fenix 5 watch from "Smart" to "Every second" then I will get *more accurate pace readings while running.*
Everything I've read (in this forum and elsewhere) about this setting contradicts what Chris told me.

Can anyone provide any insight on this? I'm getting wacky pace readings while running and I'm looking for a way to improve that. Will changing to "Every second" help me?