|I had the opportunity last Wednesday to try these two gps units side by side at the start of quite a nice session. A mate, Gavan, allowed me to try his new Etrex Legend Color. I only got to wear it for the first 5 runs as he had to leave early but it makes an interesting comparison.|
I wore the Foretrex on my right wrist/forearm in an aquapack and the Legend on my left bicep, also in an aquapack.
The Legend has the newer Sirfstar3 chip which is reported to be more accurate and much more sensitive to the GPS signal (holds satellite signals better). The internal software seems not to suffer from the 2.4m grid effect that the foretrex does so the one second setting produces a very smooth track and speed graph without the degree of up and down zig-zag we see with tracks from the Foretrex. My tracks viewed in GPSAR v3.3 bear this out as you can see from the screenshots of the graph. (see the pictures section of this website as soon as I can get it posted).
I had a couple of ‘loss of signal spikes’ in both tracks where I submerged both units during water starts and they were edited out in the normal way.
The statistics produced from the two units vary slightly as we would expect. For instance, the 100m for the Legend is only over 5 seconds whereas the Foretrex is over 6 seconds and the Legend analysis gives a slightly faster speed as a result (41.023 v 40.778). It is interesting that in contrast, over 250m the Legend gives a slightly slower speed. (40.083 v 40.171). In the 5 x10 second average the results are very close with the Legend at 37.483 and the foretrex at 37.418.
Of further interest is that the Foretrex track gave a higher 2 sec max calc. by quite a large margin at 42.213 knots whereas the Legend gave only 41.204 for the best 2 seconds and gave 1 sec max of 41.237! I believe this has something to do with the zig-zag and grid effect from the Foretrex track and I tend think the Legends 1 second figure is closer to the real peak speed. My Foretrex showed a peak GPS top speed reading of 41.5. I did not get a top speed reading from the Legend as it had not been reset since its last use.
After these 5 runs I put my weight vest on and did some more runs just with the Foretrex that make up my whole posted session for this day.
your findings confirm my comparison between a Geko 201 and an Etrex Legend C (see www.gps-speed.com/test.html). Looking at the raw data the Legend C indeed gives the full numerical accuracy of 4 valid bytes instead of the truncated 3 bytes of the Foretrex and the Geko, thus there is no 2.4m grid effect. However, as far as I know the Legend C does not use the SiRF III chip, but an older chipset. The Edge205 does not have the many fancy features of the Legend C and a grayscale display only, but it uses the SiRF III and the full accuracy, so I prefer this smaller and lighter unit.
With GPSResults you could have compared the 5sec results more accuractely (with linear interpolation). The 2sec outlier of the Foretrex should be due to the 'noise' from the grid effect. Another important point is the WAAS mode: The latest Garmin info about WAAS says that this method improves localization accuracy only in the US. In other regions it is strongly recommended to switch WAAS off (use Normal mode), since it actually gives LESS accuracy in these regions!
original excerpt in german (www.garmin.de/Begriffserklaerungen.php):
ACHTUNG! WAAS funktioniert ausschlieÃŸlich in Nordamerika, weil die entsprechenden Satellitensignale nur fÃ¼r dort gÃ¼ltig sind. Empfangen neuere, WAAS/EGNOS-fÃ¤hige GARMIN SatellitennavigationsgerÃ¤te auÃŸerhalb Nordamerikas. WAAS-Korrektursignale (erkennbar an der Satelliten-Nummer 35 auf der GPS-Seite im GerÃ¤t), so VERSCHLECHTERT sich die Genauigkeit der Positionsbestimmung.
|Thanks Manfred. I have now looked up the Garmin specs as I should have confirmed before and it seems you are quite correct. The Legend color seems to use an older GPS chip but nowhere could I find what it is. If it the same as that used in the foretrex 201 it begs the question of why Garmin cant (or more to the point, why they have not already) revised the FT 201 software so that it can have the precision of the Etrex C!|
When I ran the Foretrex result through GPS results with interpolation on it did indeed give a result closer to the Legend:
Foretrex 201 100m interpolation on = 41.369 knots (4.7 secs).
Foretex 201 100m interpolation off = 40.778 knots (6 secs)
Legend C 100m interpolation off = 41.024 (5 secs)
Legend C 100m interpolation on = 41.034 (4.7 secs)
The results from the Foretrex 201 seem to be suffering from that grid effect here and the results from the Legend C seem more consistent (dare I say more trustworthy) to me for that distance.
Of course, the use of interpolation makes no difference to the 10 second speeds or the 5 x 10 second average where both GPS tracks come out very much closer.
|Actually , now that I look at it again, the result with interpolation on from the Foretex is not closer to the Legend, just faster!|
To compare the 2 different units that have the same chips.....
I suggest yu put the data into an xls. bar graph....the 10 sec run data & include the data for when yu are accelerating up to speed for your 10 sec run & then include the decceleration speed data.
Yu should find the Foretrex's 2 sec data is very erractic, how smooth is the Legends data for accel-deccel 10 sec run?
Using this method i've compared the Foretrex to the Geko.....the Geko's bar graph data is far superior with smoother accel-deccel.
I'm a bit of a dummy with EXEL so I don't know what you mean. But if my guess is correct, would one not see the same think in the speed graph of other programs like GPSAR?
The screenshot I have sent in for the photo page is not up yet but on it you will see the difference between the smoother track of the Legend and the zig-zag graph from the Foretrex. Do you see a similar difference comparing a geko and foretrex? The only comparison I have seen is the on that Manfred Fuchs did comparing a Geko with a Legend C. In this comparison the Geko displays the same sort of 'grid effect' stepping or zig-zaging as we see from the Foretrex.
To see the graph go to:
Then go to the downloads page and click on the 'test' box.
I've seen the comparison of the Grid that Manfred Fuchs posted.
With this type of graph the Edge 205 looks great & the Foretrex a total loser, which i think it is!
The Geko most likely will display the same grid pattern to that of the Foretrex.
But these graphs only tell 1 part of story.
The reason why the xls bar graph paints a clearer picture is the GPS record data is a pulse, (being 1 or 2 sec) ...when looking at much data the line graph seems to graphical average the data flow out...with the peaks & dips hard to see visually. The speed line graph in GPSAR can be used but yu have to make the graph rather large to see your speed run of accle-your 10 sec speed-deccle.
I can email yu the xls bar graph comparsion, Foretrex vs Geko.
The 10sec speeds are only 35's knots...the foretrex 2 sec speed jumps 4 knots while the Geko 2 sec speed does not jump.
|Michael. Thanks, that would be most interesting to see. sailquik at hotmail dot com|
|Andrew you asked.. "If it the same as that used in the foretrex 201 it begs the question of why Garmin cant (or more to the point, why they have not already) revised the FT 201 software so that it can have the precision of the Etrex C!"|
I emailed Garmin tech dept 6months ago explaining our situation and asking for a firmware update to include that extra byte of info we need but I never got a reply, and no update on their site.
|Hi Daniel, I did exactly the same thing and wrote an e-mail on March 29 to Garmin - unfortunately with exactly the same outcome - no reaction at all. They dont take us very seriously and I guess they better want to sell the new, more expensive devices like the Edge 205 instead of updating the firmware of the older units. Unfortunately they are in a strong position since all other GPS-manufacturers obviously are not able to produce better devices (really waterprotected, full (untruncated) accuracy, bluetooth interface, etc.) Perhaps it would help if all 1000 or so users of this site send similar e-mails to Garmin...|