Garmin Legend – Waypoint Averaging?

September 4, 2003

I just picked a place where I want to place a GeoCache. I want to make sure I have the most accurate coordinates possible. I’ve heard about waypoint averaging and I heard the Garmin Legend will do waypoint averaging but I can’t figure it out. Can someone please help me out?

Does the Garmin Legend have waypoint averaging? If so how does it work? And how do you do it?

Thanks,

Craig@thecarsons.org

1.4k · September 4, 2003

Are you refering to WAAS (Wide Area Augmentation System.)?

http://www.garmin.com/aboutGPS/waas.html

Lend me your ear while I call you a fool.

September 4, 2003

I have a Legend and I dont think it has averaging. Wish it did.

-------------------------------------------------------------------------------

I have never been lost. Been awful confused for a few days, but never lost!

N61.12.041 W149.43.734

September 4, 2003

Oh, sure, I average on my Legend... I walk away and come back several times, and then average my numbers !

38.2k · September 4, 2003

The Legend doesn't waypoint average. I used to do it manually with my Legend. I'd take about 20 readings and average them. I didn't find it to be any more accurate than when I took one reading.

Now I just make sure I have a good sat lock, mark the waypoint and I'm done. I haven't had any complaints about my coordinates. In fact I've had people compliment me on their accuracy.

The one cache where I have received some complaints is one where I averaged.

So in short, don't waste your time with averaging. Place the cache, take a reading and post it. If you subsequently geta lot of complaints, go back and do it again, but it's my bet you won't have to.

"You can't make a man by standing a sheep on his hind legs. But by standing a flock of sheep in that position, you can make a crowd of men" - Max Beerbohm

September 4, 2003

I have the Lengend and do basiclly what Brian does. Make sure you have a good satellite lock and a decent accurracy before you take the reading. You may need to wait a few minutes for this to happen. If you can, set the GPSr down for a while. Usually what I do is take the first reading, walk a few yards away from the cache, bring up the coordinates and then check them to see if they work. If I'm not happy with them I'll repeat the process. I try to get the coordinate accuracy closer for micros.

GF

********************************************

Fool me once, shame on you, fool me twice, shame on me.

9.2k · September 4, 2003

Averaging together coordinates won't somehow give you better or more accurate numbers. Averaging was a way to try and lessen the effects of Selective Availability, and it did a pretty poor job of it. SA has been turned off for over 3 years. There's no need to average anything. Just make sure you've got a good number of sat-locks in a good constellation, and go with it. Don't think you have to be directly over the cache when you take a reading. If you can move 8' away and get a much better view of the sky, do it. You (and the cache hunters) will be better off.

"Don't mess with a geocacher. We know all the best places to hide a body."

September 4, 2003

I agree that waypoint averaging is pretty useless. Getting 13 feet accuracy is great in my book... Prime suspect has an excellent point too with the selective ability issue.

-----------------------------------------------------------

Garmin V

Using Opera 7.1

1k · September 4, 2003

I wish you wouldn't tell people not to bother averaging their waypoints. If you have perfect conditions, with a clear view of the sky, and your GPS has been in place long enough to settle down, then yes - you could get away with taking just one reading.

The same logic for averaging during the days of SA still apply today. If you take the average of multiple readings, you reduce the error.

It only takes a few extra minutes and IMHO it's worth the effort.

-Junglehair

I thought I was wrong once, but I was mistaken.

September 4, 2003

I had a Garmin GPS 128 that had averaging. After they turned off selective availability, I could let it average for 15 minutes and basically I would end up with the same coordinates I started with.

Ducks - Flying, great tasting, geocaches of meat

1.1k · September 4, 2003

quote:
Originally posted by junglehair:
I wish you wouldn't tell people not to bother averaging their waypoints .... The same logic for averaging during the days of SA still apply today. If you take the average of multiple readings, you reduce the error.

It only takes a few extra minutes and IMHO it's worth the effort.

Worth the effort, not so, reduce the error? how do you know (without recording data and comparing it to the real position).

I'll defy anybody to machine average coordinates for a few minutes then GUARANTEE that average is better then the position started with.

Averaging in no way today can be compared to averaging when SA was active.

Cheers, Kerry.

I never get lost everybody keeps telling me where to go

9.2k · September 5, 2003

quote:
Originally posted by junglehair:

The same logic for averaging during the days of SA still apply today. If you take the average of multiple readings, you reduce the error.

Sorry, but no. SA was an artifically induced random error. That's not what you're likely to encounter in the field.

Again, averaging together bad data will not somehow make it good data. Averaging together good data will not somehow make it better data.

"Don't mess with a geocacher. We know all the best places to hide a body."

September 5, 2003

What I do to get the best waypoint possible is to take a waypoint.....walk away.....walk back....take another point. Do that a couple of times. Then I approach the cache, select nearest waypoint, walk farther away (especially if I can get to an open field in an attempt to increase my accuracy) and make sure that the waypoint takes me right back to the cache. I'll keep doing this until I am satisfied with the results of one or two particular waypoints.

Next, I go back to the computer and download the points to mapsource. Then I check out the waypoint(s) I thought were best to see if it is near the middle of the cluster of points I took.

But I think the real key when getting the best coordinates is make sure that when you walk up to the cache location you walk in as straight a line as possible. The GPS just won't pick up those little zig zag movements....especially under a canopy of leaves.

Hook'em hard Muskytooth

Hermits have no peer pressure.

38.2k · September 5, 2003

quote:
Don't think you have to be directly over the cache when you take a reading. If you can move 8' away and get a much better view of the sky, do it. You (and the cache hunters) will be better off.

This is very good advice and something I always do. A good reading 15 feet from the cache is better than a bad reading right over it.

quote:
I wish you wouldn't tell people not to bother averaging their waypoints. If you have perfect conditions, with a clear view of the sky, and your GPS has been in place long enough to settle down, then yes - you could get away with taking just one reading.

I stopped averaging a year and a half ago. I've yet to receive a complaint about my coordinates since. I find it to be a complete waste of time. Besides, if you have bad reception, or a bad sat alignment, you're just getting an average of bad coordinates.

"You can't make a man by standing a sheep on his hind legs. But by standing a flock of sheep in that position, you can make a crowd of men" - Max Beerbohm

September 5, 2003

quote:
Originally posted by zoltig:

Lend me your ear while I call you a fool.

Sorry for the OT but... Zoltig, this phrase reminds me something... Am I right in continuing "You were kissed by a witch one night in the woods"?

L.

1.4k · September 8, 2003

quote:
Originally posted by Acaro:

quote:
Originally posted by zoltig:

Lend me your ear while I call you a fool.

Sorry for the OT but... Zoltig, this phrase reminds me something... Am I right in continuing "You were kissed by a witch one night in the woods"?

L.

YES L.

"...and later insisted your feelings were true"

Bravo!

Myself, I am currently sipping on Glenfiddich Solera Reserve (15y)

Also have Balvenie of Banffshire (10y)

plus, not a single,

Johnnie Walker Blue Label (top notch for a blended).

Cheers

K.

Lend me your ear while I call you a fool.

2.3k · September 9, 2003

Considering the topic here I think the line further down, "Keep looking, keep looking for somewhere to be", has something to do with the placements of caches too!

Anders

September 9, 2003

Sorry, as a signal processing engineer I just can't let the misconception that " averaging bad data won't make it good , averaging good data won't make it better" go unchallenged. Depending on the statistics of the error, averaging may indeed improve your readings. If the errors are "unbiased" (that is truly random with zero mean) then the accuracy of your reading will improve proportionally to the square root of the number of readings taken. That is, four readings will double your accurracy, ON AVERAGE, 100 readings will make it 10 times better ON AVERAGE. Indeed their is no guarantee, but the more averaging you do, the more likely you will fit the statistical profile. This technique for pulling good data out of bad signals is used in everything from data transmission to electrocardiology.

If their is a bias in the error (for instance all readings are on average 25 feet too far to the west) then of course averaging will not remove the bias.

[This message was edited by MacFlash on September 09, 2003 at 01:28 PM.]

September 9, 2003

In order for the averaging to be significant it is NOT the number of readings but the length of time over which the readings are taken.

Accuracy dependent on the satellite geometery and to a lesser degree free electrons in the ionosphere.

100 evenly spaced readings between sunset and sunrise would probably be the most accurate you could get from a commercial GPSr.

BUT, the searcher is probably no better off. They are still dependent on the conditions at the specific time they are looking.

So.... when you get close. Put the GPSr away and use the force.

1.1k · September 9, 2003

GPS errors are certainly not "unbiased" and certainly not random in the pure statistic sense.

And it's simply not a misconception that if one is averaging data that is trending away from the truth (which GPS can do for reasonable periods of time) then the average will in affect be tending to make the position worse.

The problem with GPS is errors (for want of a better term) especially in the days of Selective Availability were certainly not random but certainly not predictable. Generally of any 24 hour period 95% of any errors will be trapped, which will certainly give a reasonable average if averaged over that time frame, however that's certainly not practical. A 30 day cycle is required to trap most errors.

These days without SA averaging doesn't contribute any reliable outcome to the solution apart from knowing there's a 50/50 chance of getting a better result but of course there's a 50/50 change of getting a worse result and only hindsight will verify this.

Even though GPS accuracy is based on statistics the real world affects of averaging both pre/post SA are quite different

Practical example of GPS averaging

More practical aspects of GPS averaging

Some aspects of GPS averaging with respect obstructions

Cheers, Kerry.

I never get lost everybody keeps telling me where to go

4k · December 8, 2003

All that being said, I still say I've had nothing but positive success with averaging on my Legend. I've played with it in all types of environments, from light to heavy canopy, clear to cloudy sky, morning, noon, night, and temps from 20 f to 95 f, and still find that when I average a waypoint, it gets a better coord.

And THAT being said, let me say this: I didn't average when I placed my one and only cache so far, and everyone that has looked for it has found it. When it comes down to brass tacks, put away the GPSr when your within 20 ft and use your eyes and brain....or as someone posted earlier in this thread, use the force!

December 10, 2003

What I've done with my Legend is to set the GPSr down for a few minutes and let it get a good lock. Mark your waypoint and then walk away. When you get back to your starting point act just like someone looking for the cache. I tell it to goto that waypoint and see how close it gets me. I've found that this way I usually am within about 15 ft of the cache. I hope this helps.

December 10, 2003

Bravo!  Myself, I am currently sipping on Glenfiddich Solera Reserve (15y) Also have Balvenie of Banffshire (10y) plus, not a single, Johnnie Walker Blue Label (top notch for a blended).

It seems that I'm destined to OT on this topic... This is my new username (In fact I'm "Acaro of Team TAR")... I forgot to look on this thread to see if Zoltig confirmed my "interpretation"... Now I see that also Anders likes good music (good music = the one I like ).

Interesting the connection between JT and single malt... but it works for me too, even if I never noticed!

Personally I'm fond on 18 Y Glenmorangie over anything else, then 15 Y Macallan and Highland park (any aging)... but also the more aged Isley (Ardbeg and Laphroaig, for what I tried) have ALL my approval!

Now back on topic. On this thread I proposed a way to do a "Magellan like" wp averaging on a Legend/Vista using the active tracklogging at the maximum time resolution. We discussed there with Kerry the usefulness of the pratice, and I promised that I would have tried to test what he said (which is pretty much the same he says here). I made a couple of tries and have to say that he seems to be quite right: at least in the open (I haven't tried in the woods, too much witches there kissing people ) averaging does not seem to improve the precision of the wp.

The random component of the position error, as a consequence, seems to be finally smaller to the contribution due to algoritym convergence, geometry, atmosphere model approximation or anything else involved.

I don't have a decent statistical sample yet, only few sparse tries, but this seem to be the trend.

Maybe averaging many positions taken at distant times (one each day, for example) would be somewhat beneficial, due to the loss of correlation in satellite geometry and ionospheric situation between the measures. I don't know. But on short timescales ( 1 hour or less) averaging seems to be useless. :huh:

4k · December 10, 2003

Bravo!  Myself, I am currently sipping on Glenfiddich Solera Reserve (15y) Also have Balvenie of Banffshire (10y) plus, not a single, Johnnie Walker Blue Label (top notch for a blended).

It seems that I'm destined to OT on this topic... This is my new username (In fact I'm "Acaro of Team TAR")... I forgot to look on this thread to see if Zoltig confirmed my "interpretation"... Now I see that also Anders likes good music (good music = the one I like ).

Interesting the connection between JT and single malt... but it works for me too, even if I never noticed!

Personally I'm fond on 18 Y Glenmorangie over anything else, then 15 Y Macallan and Highland park (any aging)... but also the more aged Isley (Ardbeg and Laphroaig, for what I tried) have ALL my approval!

Now back on topic. On this thread I proposed a way to do a "Magellan like" wp averaging on a Legend/Vista using the active tracklogging at the maximum time resolution. We discussed there with Kerry the usefulness of the pratice, and I promised that I would have tried to test what he said (which is pretty much the same he says here). I made a couple of tries and have to say that he seems to be quite right: at least in the open (I haven't tried in the woods, too much witches there kissing people ) averaging does not seem to improve the precision of the wp.

The random component of the position error, as a consequence, seems to be finally smaller to the contribution due to algoritym convergence, geometry, atmosphere model approximation or anything else involved.

I don't have a decent statistical sample yet, only few sparse tries, but this seem to be the trend.

Maybe averaging many positions taken at distant times (one each day, for example) would be somewhat beneficial, due to the loss of correlation in satellite geometry and ionospheric situation between the measures. I don't know. But on short timescales ( 1 hour or less) averaging seems to be useless.

Huh? That's WAY over my head. I'm just going on personal experience with the averaging I've done, and will restate that my own tests proved that averaging increases accuracy. However, I will state again, that if you're in the ballpark with your waypoint when looking for or placing a cache, 20 feet typically isn't going to be an issue. It's at that distance that you need to drop the GPSr and start using your brain and eyes.

Now, what did she say again?

December 10, 2003

If your unit has an auto-averaging feature go ahead and use it. Otherwise don't waste your time. The manual method with a dozen or so positions does not increase positional accuracy to any significant degree. Thirty is the minimum. 60 is the recommended minimum for serious uses. This is a bit impractical to do manually.

1.1k · December 10, 2003

Averaging is really a feel good thing and if it feels good using it then by all means, now what it means now that's different.

Still better means to ensuring a Wp has the best confidence and integrity possible.

Cheers, Kerry.

December 10, 2003

No the Legend doesnt have an auto average feature. However as several folks have stated I don't believe that it is a useful feature either. I owned Garmin GPS III Plus which is a great unit and it did have waypoint averaging. I noticed the same thing that others have said... it didnt seem to make any difference. The waypoints that it was averaging didnt vary reading to reading. So unless you are averaging over different times of the day and maybe over multiple days I don't see much point. I feel making sure that you have a clear view (as possible) of the birds and a good lock are more important. see ya, jeff'

December 11, 2003

I started averaging waypoints on my GPSV about a year ago after placing a cache and coming back the next day because I forgot to put a pencil in the cache. I was a bit surprised when I couldn't find it. Turned out my coordinates were almost 100 feet off. (Cache was hidden in a wooden area with tall grass/weeds).

So I averaged the waypoint and hit the enter button when the "accuracy" indicator had dropped from 28 feet to 13 feet. I take that to mean the GPSr is simply reading more sattelites. which can't be a BAD thing. Maybe I'm wrong.

Anyway I walked out of the area and came back to the cache using the new averaged way point and walked right up to it.

I look at waypoint averaging as simply a way to double-check your readings on the off chance you had a bad lock or was getting signal bounce on the first one. Watching the accuracy meter rise or fall reassures me that the unit is actually seeing the birds and using them to calculate the waypoint.

Many times when under heavy tree cover a GPSr can loose it's lock but fail to report it for several seconds. I'm sure most of you have had a unit tell you a cache is forty feet ahead of you and as you continue walking it suddenly does a 180 telling you the cache is NOW 40 feet behind you.

Waypoint averaging in my opinion is simply a tool that helps me rule out bogus coordinates. Since averaging is automated on my unit and it only takes a few seconds I don't really consider it a waste of time.

Edited December 11, 2003 by jollybgood

Garmin Legend – Waypoint Averaging?

Recommended Posts

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Join the conversation