Polar S625X vs Garmin Forerunner

tallbird - cheers

It's a while since I've posted on this thread, but it's good that it's become a useful forum for discussing S625X issues. I've now been using mine for 3 weeks and am very pleased with the results. I raced a half marathon today and it measured 13.05 miles, which is accurate to within half of one per cent. Over the same distance my GPS (Garmin F201) unit is usually about 3% out. The S625X always seems to get well within one per cent. I improved my PB by over 2 min 30, at least partly because of the accuracy of the data I was getting from my watch.

For me that's a huge gain and means I can analyse my training with much more confidence than before. The other huge improvement is having a real time pace measurement. This is pretty hopeless on the F201 but - as far as I can see - far more accurate on the S625X. Finally, the desktop software is better with the S625X. I find the workout graphs are really useful.

What do I miss about the GPS system? Well, it's nice being able to work with maps. And the training "buddy" is pretty inspired. There's no reason in principle (apart from the computing capacity and memory of the watch) why a footpod based system shouldn't be able to do those things too.

Ah well, I guess we'll have to wait a little longer for the perfect gadget.

Riverman,
I've had mine for 30 weeks now and I've never lost my interest with it. I do believe all the HRM manufacturers are in their infancy stages and we will see very cool things in the future. The 625x is Polar's first attempt at the pod, so we'll see major improvements as their experience increases. Same thing with Garmin and the heart strap.

I'm hoping each manufacturer is looking at the features of the others. I would sure like to see Polar with an audio announcement feature similar to MP3Run ( I haven't heard it, but I could imagine it to be very useful).

I would also like to see a GPS mapping accessary that could work with Polar. To maintain the accuracy and dependability, the pace and distance would still come from the pod, but the gps mapping coordinates would be received from a GPS unit and stored in the same file which could then be downloaded to PPPSW. I would love to see where I run, but don't want to depend on satelite signals for my training and race data.

OMM

I agree with all of that - and it's good that you're still enthused by the 625X. Is GPS capability strictly necessary for mapping? I'd have thought the data from the foot pod ought to be enough to calculate position as well as distance and speed. (Obviously you'd have to enter some reference co-ordinates - eg for the start of your racce - if you wanted absolute position).

Riverman,
Interesting idea. But in order for you idea to work, you would get an accurate orientation from electronic compass on the watch. Suppose you ran in a straight line for 10 miles and wanted to be within 50 feet of the center of the road. At the start of the run you would have to know your orientation to within .05 degrees which is probably pretty difficult for a compass.

Also the problem would be that the errors in the calculated x,y coordinates are accumulating to an unacceptable large values.

The advantage of GPS is that the errors are not cumulative.

Just another plug for Polar to muddle your decision. I looked at this same argument from every angle I could and finally took the plunge with the S625X. Only a week using it and I can't believe I didn't buy one sooner. Accuracy 98% without calibration and even better with. Should be good enough for most of us (Greyhound said it best). Also not bulky like Garmin so am not embarrassed to wear it wherever I go. I haven't even begun to tap the capabilities of this thing and look forward to many happy runs to come. It has honestly added more excitement to my running. Must be the elated gadget gene.

OMM

Interesting thoughts. Without knowing anything about how the foot pod works, I've been assuming that it measures acceleration as a 3-D vector, from which the watch ought to be able to calculate the direction in which you're running without a separate compass.

But your argument about accumulated error is obviously absolutely right.

Hi folks
been a while since I posted here, but as Riverman says, it has become a useful forum.

I have a question which I have posted on the Yahoo group, but I'm yet to receive any feedback.

I've had my 625x over a month now and it has quickly become invaluable.

I have noticed a discrepancy between the total distance run as recorded on the watch, and the sum of the distances run as given in the View/Laps tables on the PC.

Last Sunday I ran a distance given as 6.91 miles at the last 5sec sample point. When I added up the total of the lap distances, automatically set for 0.5 miles, the total was over 8 miles !

Any ideas gratefully received.

Have just read back through the thread, this from OMM seems to be addressing the issue.

'Lap distances are calculated internally to the watch and uses 2 sec samples(approximately) for computing distances. These computed lapdistances are later sent to the PPPSW as is. However, the sample point distances are calculated by the PPPSW from the velocities at the sample times you have chosen(5,15,60 seconds) so the curve and listing distances will be more coarse. If you rerun your 10K but use 60 second samples I think you'll see a larger discrepancy.'

OMM, if I have understood this correctly, then the total of the laps is correct and the samples are underestimating! Is that th ecase, or am I just hopelessly confused?

Hi XL-man
I'm puzzled by the the large discrepancy you're seeing. I'm seeing very little difference between the two. The following are the summed lapdistances and last sampled distance for my previous runs.
LastSample LapTotal %Diff
(mi) (mi)
5.529 5.532 .05%
4.249 4.263 .33%
14.083 14.064 .14%
6.802 6.818 .23%
These were slow steady runs with autolapping set to 1mi and sample time at 5sec. Your difference was about 13%.

Hi XL-man
I reviewed Willem's document trying to find the answer to your question about lap distance accurancy vs sampled distances. It's worth reading and is much better than getting it second hand from me.
You'll find a link to his document in the link section at Yahoo's forum.

He claims there are conditions where accumulated lapdistances are more accurate and others conditions where final sample times are. Hows that for a straight forward answer.

He states that each lap interval will add an extra 10-40 meters to the total and so lap intervals should be large to minimize this effect. He calculates that a lap interval of 800m would yield an accuracy of 97%.
After rereading it, I should correct a previous post. Willem proves that PPPSW does not use the lapdata from the watch even though it's sent to it. It recalculates the lap distances from the sample distances.

Morning all,

This is my first post on this thread so can I start by saying thanks to you all for the informed debate on the two products. It certainly helped me make an educated decision when it came to purchasing time.

Like BeepBeep, I too plumped for the 625x a week past and so far so good. My logic for choosing the Polar was that though I run outside for my long runs at weekends, I do most of my weekday running at the gym and I want to accumulate data for all my runs rather than just the weekend one (which the Garmin would be limited to - clearly no signal in the gym!).

I do, however, have one observation, having yet to calibrate the gadget, and that is the slower I run the less accurate the pacing readout becomes - at least it doesn't match the output on the running machine. Total Distance covered seems to be accurate enough, but as I increase speed it seems to get better.

As an example I was doing an interval session and during the effort segments I was running at 5.58 min/mile on the running machine and the Polar was at 6.00 which is fine, but during recovery I was up to 9 min/mile and the watch was at 10 min/mile.

Clearly this is a function of stride length but my concern is that if I calibrate at a certain rate it is going to skew in one direction or the other when I change speed.

The consolation at the moment is that at least it errs on the cautious side ie telling me I'm going slower when I'm going quicker, I can can use the HR function to keep me focussed on my pace too.

Any thoughts would be welcomed. (And sorry about the lengthy post).

Hi Jaffs,

And I thought I was the only one that wore the pod on a public treadmill! The treadmill/watch combination is ideal for measuring heart rate trends for identical runs. Have you tried the overlaying plots on PPPSW. I think up to 5 can be done at one time.

Polar claims 97% accuracy uncalibrated and doesn't mention stride contraints. Calibration is for making accuracy improvements on the 97%. Effect on stride changes will impact you when you try for higher accuracies. Most of the complaints I see on other forums occur when they try for around 99% in which case some Polar bugs come into effect.

I would tend to blame the treadmill. I see speed variations when I run the same speed on different treadmills. You'll probably need to run on several of your gym's mills with the two speeds in question and see if you can deduce the culprit. 10% error you're seeing, sames way too high for a Polar issue.

I think XL-man is narrowing down his problem (see above) to a large sample interval (60 secs), so I think we'll hear a report from him shortly.

OMM

Thanks for your thoughts. I'll try different machines, though typically things are on hold today as throat and head playing up. I am using 5 second observations to keep things as accurate as possible. Weekend run was good: 9.1m on polar vs 8.97m on mapminder. Not too bad.

I haven't tried the overlay charts yet but if things go to plan (ie fitness improves!)that should make for an interesting analysis.

And I must admit I haven't seen any other pods at my gym either - perhaps people just need someone to show them the light!

With all the good reports and recommendations here, I'm very tempted to get a polar S625x, but I've got one question or puzzle. I do quite a lot of my running on very unven off-road terrain, where my stride must vary a lot, both in speed and the way my foot moves. Does it cope with this ? and how ?(no don't try and answer this second part. If you tell me its does work, I'll put it down to another of those mysteries of modern technology!))

Hi folks,
just to confirm OMM's suspicions.
The 625 is MUCH better with 15s sample times instead of 1 minute. Also getting much smaller discrepancies with automatic lap times set to 1 mile.

Like you TmR, I'm baffled by the clever technology, but the gadget is wonderful.