+u.rusty Posted October 16, 2010 Share Posted October 16, 2010 I am surprised by the number of caches I find that are within arms length of ground zero, as indicated by my GPSr. I am curious what methods are commonly used to determine ground zero for your cache. Here is a list of a few methods that might work: Using waymark averaging, assuming that your GPSr has that feature. Choosing coordinates then finding the spot with your GPSr. (Obviously, this would only work under certain circumstances, such as just off the side of a trail in the field.) Recording the coordinates several times and averaging. What method do you use? Quote Link to comment
Andronicus Posted October 16, 2010 Share Posted October 16, 2010 I use GCzII. It does live waypoint averaging. I do 10 averages, and save the coordinates in my phone's notebook. Then I walk away from the cache and back, and do it all again. After 5 times of that, I do an average on all the coordinates in the notebook. It there is a outlier, I ignore it. If there is 2 or more outliers, I do more averages. Quote Link to comment
Clan Riffster Posted October 16, 2010 Share Posted October 16, 2010 My method is similar to what Andronicus uses, with the exception that I don't use the averaging feature on either of my Garmins. Early in my caching career, I used averaging a lot, and I discovered through the found logs that my coords were consistently off by a margin great enough to cause concern. I can't speak to what averaging does inside the device, as I am clueless regarding what the software does. In watching my GPSr screen during the process, I see that the estimated error range gets lower, as the device sits, which should give me better results, but in practice, it sure doesn't seem to work out that way. I eventually abandoned the internal averaging supplied by the Garmin software and implemented a more manual process. Once I figure out where I want my cache, I'll snap a set of coords, then I'll walk away in several different directions, at least 100', following my arrow back to the cache, seeing how close it gets me. If I consistently get readings showing less than 10' off, I'll use those coords. If I get extensive differences, I'll snap more coords, and repeat the process. On rare occasion, I'll have to snap a whole bunch of coords, with none of them giving me satisfactory results. At that point, I'll keep the best three sets of coords, tossing out the others. Then I'll zoom in on my map, finding a spot that is central to the three remaining sets of coords, and I'll use those coords, posting a note on the cache page detailing my difficulties. Quote Link to comment
GOF and Bacall Posted October 16, 2010 Share Posted October 16, 2010 I set my Oregon 450 on the cache site and let it average The results are always within acceptable limits. Quote Link to comment
+dakboy Posted October 17, 2010 Share Posted October 17, 2010 From what I've read, averaging on a single visit with a consumer handheld doesn't give much of an advantage over letting the GPS "settle" for a few minutes and just recording the coords. Averaging across several visits over a number of days, however, may be advantageous, as you'll have a different satellite constellation each time. Quote Link to comment
+addisonbr Posted October 17, 2010 Share Posted October 17, 2010 Once I figure out where I want my cache, I'll snap a set of coords, then I'll walk away in several different directions, at least 100', following my arrow back to the cache, seeing how close it gets me. If I consistently get readings showing less than 10' off, I'll use those coords. If I get extensive differences, I'll snap more coords, and repeat the process. This is what I do. I find I get better readings than the automatic waypoint averaging that my handheld offers. I also live in an area with both (1) poor satellite reception and (2) very excellent satellite photos. So I use Google Maps a lot. Quote Link to comment
+dfx Posted October 17, 2010 Share Posted October 17, 2010 (edited) averaging. always gets me great results. it seems that it's mostly (only?) owners of garmin 60 series units who experience worse results with averaging than without. my theory is that those units have a broken averaging algorithm. i don't have any proof of that though. Edited October 17, 2010 by dfx Quote Link to comment
+Mr. Wilson & a Mt. Goat Posted October 17, 2010 Share Posted October 17, 2010 (edited) From what I've read, averaging on a single visit with a consumer handheld doesn't give much of an advantage over letting the GPS "settle" for a few minutes and just recording the coords. Averaging across several visits over a number of days, however, may be advantageous, as you'll have a different satellite constellation each time. True, it probably does get it more accurate. But how many people would actually do that on an everyday hide?.... Just saying.... ________________________________________ I usually average it out. If I have another GPS with me I'll get an average out of both. For urban hides I also like to check google maps. Oh ya, I forgot to mention, I let the GPS do it's average and that's what I use. Edited October 17, 2010 by Mr. Wilson & a Mt. Goat Quote Link to comment
+briansnat Posted October 17, 2010 Share Posted October 17, 2010 (edited) I make sure I have a good signal, let the unit settle for about a minute and mark my waypoint. That's all. No averaging, no walking back and forth to the site. I find averaging to be a waste of time. If you have good reception conditions, averaging won't make much of a difference. If you have poor conditions you will be averaging bad data. The only caches where I get consistent complaints about coordinates are the ones that I averaged. That's because I only average when I see I'm getting a lot of signal bounce. Edited October 17, 2010 by briansnat Quote Link to comment
+ZeLonewolf Posted October 17, 2010 Share Posted October 17, 2010 I have an iPhone, a Droid X, and a Garmin 62s. When I used the iPhone, I would just let it settle for 10 minutes or so, and the coords it took after that were usually surprisingly accurate. The one attempt I made at hiding with the Droid X was a total disaster, with coordinates over 70 feet off. Then I got the Garmin 62s. At first, I would just snap coords from GZ, so long as the "accuracy" reading was small. Complete and utter disaster. One of my hides was 30-50' off, and another was over 90 feet off! How the coords could be 90' off while the accuracy was listed as less than 20' is beyond me. I've found that the Garmin 62 is way less accurate than it reads for the first 5-10 minutes after you power it on. I've figured out a method that gives me dead-on-balls-accurate coordinates on every time. I set the unit down and use the waypoint average feature to record a waypoint. Usually takes a couple minutes. Then I navigate to the waypoint while still at GZ. If the distance is more than 10 feet, I retake them. Since I started doing that, all the accuracy problems went away. Quote Link to comment
+u.rusty Posted October 17, 2010 Author Share Posted October 17, 2010 Wow! I am very pleasantly surprised at the number and quality of the replies to my question. Thank you all for replying! Cache owners go to much more trouble than I thought to obtain accurate coordinates. I, for one, appreciate the work that cache owners go to in order to place a cache well. Those of us that are newbies (myself included) should take note. One of the common complains, it seems to me, involving new geocachers placing their first cache or caches, involves poor coordinates. Hopefully, those of us that are new (me) will learn some good methods here that should help improve our ability to determine good coordinates much the enjoyment of other geocachers. Averaging, by the GPSr or manually, appears to be a common method. Testing the coordinates by finding them yourself a few times after moving a distance away from ground zero appears to be a good way to confirm reasonable accuracy. Playing around with my Garmin GPSmap 60CSx, I seem to get good results just using waypoint averaging over about 30 seconds. But, I am going to try some of the other methods described in this thread to see how I can obtain better results. Quote Link to comment
+Harry Dolphin Posted October 17, 2010 Share Posted October 17, 2010 My QC department insists on the best coords possible. We use waypoint averaging (though I seldom see much of a change during that process - but don't consider it a waste of time to do it.) Then we go 60-100 feet off, and use the GPS to find the cache. Keep doing that until the GPS leads to the right spot. Then, unless the hide itself is meant to be evil, we leave a good hint. Quote Link to comment
+===sgb Posted October 17, 2010 Share Posted October 17, 2010 One method I sometimes use it to get a rough coord and then use google maps satellite view and goto the coords. If the arrow is in the wrong place, I tweak the coords until it's correct. Seems good to better than 1 metre in the UK. Quote Link to comment
+===sgb Posted October 17, 2010 Share Posted October 17, 2010 (edited) Somehow my post went twice. Can't find a delete post button. Edited October 17, 2010 by ===sgb Quote Link to comment
GOF and Bacall Posted October 17, 2010 Share Posted October 17, 2010 Somehow my post went twice. Can't find a delete post button. There is no delete post button. When the add reply times out it almost always has posted so don't retry. They have been promising to deal with the forum issues for a long time now. I am surprised at how difficult some folks make this. Almost everyone's gps has the ability to do waypoint averaging. Quote Link to comment
+dfx Posted October 17, 2010 Share Posted October 17, 2010 One method I sometimes use it to get a rough coord and then use google maps satellite view and goto the coords. If the arrow is in the wrong place, I tweak the coords until it's correct. Seems good to better than 1 metre in the UK. so essentially you're not using the GPS to get the coords, but google maps. you're wrong in your assumption that the google sat view has the same level of accuracy in all locations within a larger area, or even within a small area for that matter. Quote Link to comment
+deercreekth Posted October 17, 2010 Share Posted October 17, 2010 I use averaging on my PN-40 and let it take 150-200 readings. Then I'll walk away an navigate to the waypoint. Quote Link to comment
+addisonbr Posted October 17, 2010 Share Posted October 17, 2010 If you have poor conditions you will be averaging bad data. I think that unless the data are biased in a particular direction (like consistently showing you 50 feet too far south or something), averaging bad data should still get you better results in general than just taking one snapshot. One method I sometimes use it to get a rough coord and then use google maps satellite view and goto the coords. If the arrow is in the wrong place, I tweak the coords until it's correct. Seems good to better than 1 metre in the UK. so essentially you're not using the GPS to get the coords, but google maps. That's very close to how I do it; I've never much concerned myself with whether I'm using the GPS *or* satellite maps. I just figure I'm using both tools together to get the best results. In my area, this technique seems to be very accurate. As always, YMMV. Quote Link to comment
+SSO JOAT Posted October 17, 2010 Share Posted October 17, 2010 Make sure the GPSr has been turned on with a good, open-sky sat lock for at least 10-15 minutes. This gives it time down update the entire sat constellation data file, which changes daily. At the cache location, check your sat signal strength and position the unit for the best reception. Mark the location. If you have built-in averaging, turn that on. Hold the GPSr still and let it take about 10-12 readings. Stop averaging and look at the distance to waypoint number. If it is within 2-3 feet. Save the waypoint and you're done. If the distance is much further than that, restart the averaging and repeat. When averaging, one or two readings that have a larger error will kill the average. Thus just letting the unit average for a long time won't get rid of those errors, which will remain part of the averaged point. When you restart the averaging function, it dumps the previous and starts over. All the walking away and dancing stuff doesn't do much but delay your progress. If the cache is in a spot with poor coverage, you should test the coords by coming back several hours or a day later to verify that your recorded point is good. Walking away and coming right back is still using the same sat constellation, which will give you the same results. For that to work, you need to come back and recheck under a different sat sky. Quote Link to comment
+MaxiP Posted October 18, 2010 Share Posted October 18, 2010 Using a HTC Tattoo, I use the waypoint feature in my preferred SW and let the device get 50-70 readings that have little deviation. I then reset the waypoint and let it collect another 50 or so whilst I have a snack / drink. Like others, I'll walk a way a bit and 'navigate' back to the waypoint to be sure that they are as accurate as possible. So far, I've only had positive comments on the accuracy of the posted co-ordinates. Quote Link to comment
+sbell111 Posted October 18, 2010 Share Posted October 18, 2010 I make sure I have a good signal, let the unit settle for about a minute and mark my waypoint. That's all. No averaging, no walking back and forth to the site. I find averaging to be a waste of time. If you have good reception conditions, averaging won't make much of a difference. If you have poor conditions you will be averaging bad data. The only caches where I get consistent complaints about coordinates are the ones that I averaged. That's because I only average when I see I'm getting a lot of signal bounce. I'm with BS on this one. Quote Link to comment
+Castle Mischief Posted October 18, 2010 Share Posted October 18, 2010 I make sure I have a good signal, let the unit settle for about a minute and mark my waypoint. That's all. No averaging, no walking back and forth to the site. I find averaging to be a waste of time. If you have good reception conditions, averaging won't make much of a difference. If you have poor conditions you will be averaging bad data. The only caches where I get consistent complaints about coordinates are the ones that I averaged. That's because I only average when I see I'm getting a lot of signal bounce. I'm with BS on this one. Somebody mark the calander. Quote Link to comment
GPS-Hermit Posted October 18, 2010 Share Posted October 18, 2010 Here is a fun thing to do with a GPS to challenge all this averaging. Turn your GPS on and let it settle, Walk 200 feet after settling in a straight line and stop at a repeatable spot, mark immediately, now average 100 times and take that point Now just sit there for 15 minutes and let the GPS store points creating what I call a STOP CLuster in your track log. Download your tract log and saved points to a map program and compare the point you averaged to the stop cluster. If the one you averaged is in the middle of your stop cluster - I say it did pretty good. Don't expect the same results every time but it is pretty interesting what happens. The immediate save could be right on target as well or not. Try again the next day at a different time of day and see what happens. Fun learning it is. Pretty soon you'll be even more confused! One really rouge point can throw your average off. I think the GPS needs a standard D-V-A-shun Also - find a cache where the co-ords where almost perfect and let your GPS just sit on the cache and watch the distance as it sits and see how far it gets away from the cache. Looks like the cache is moving around like a poltergize! Quote Link to comment
GPS-Hermit Posted October 18, 2010 Share Posted October 18, 2010 Here is a fun thing to do with a GPS to challenge all this averaging. Turn your GPS on and let it settle, Walk 200 feet after settling in a straight line and stop at a repeatable spot, mark immediately, now average 100 times and take that point Now just sit there for 15 minutes and let the GPS store points creating what I call a STOP CLuster in your track log. Download your tract log and saved points to a map program and compare the point you averaged to the stop cluster. If the one you averaged is in the middle of your stop cluster - I say it did pretty good. Don't expect the same results every time but it is pretty interesting what happens. The immediate save could be right on target as well or not. Try again the next day at a different time of day and see what happens. Fun learning it is. Pretty soon you'll be even more confused! One really rouge point can throw your average off. I think the GPS needs a standard D-V-A-shun Also - find a cache where the co-ords where almost perfect and let your GPS just sit on the cache and watch the distance as it sits and see how far it gets away from the cache. Looks like the cache is moving around like a poltergize! Quote Link to comment
+Frank AZ Posted October 19, 2010 Share Posted October 19, 2010 I also live in an area with both (1) poor satellite reception and (2) very excellent satellite photos. So I use Google Maps a lot. Having found some of your caches today (or maybe more tommorow) I can agree completly for a challeneged area. In most cases at home, I set the garmin down and let it sit for 5 minutes to average, and I am good to go. I will then spot check with google maps satellite imagery to make sure it looks OK (and not way out of wack) Quote Link to comment
+dfx Posted October 19, 2010 Share Posted October 19, 2010 now average 100 times and take that pointNow just sit there for 15 minutes and let the GPS store points creating what I call a STOP CLuster in your track log. Download your tract log and saved points to a map program and compare the point you averaged to the stop cluster. If the one you averaged is in the middle of your stop cluster - I say it did pretty good. unfortunately the track log doesn't necessarily include every GPS sample the receiver has gotten over the duration of recording. in fact, it's likely to discard most if not all of the samples of there's no significant deviation from the last sample, so such a test wouldn't give the desired or expected results. however, there's plenty of averaging tests out there already, such as http://users.erols.com/dlwilson/gpsavg.htm or http://gpsinformation.us/gps60c/60C-Accuracy.html those pretty much prove that averaging per se is beneficial, but don't say anything about the particular averaging algorithms used in individual units, which could still be broken. Quote Link to comment
+Viajero Perdido Posted October 19, 2010 Share Posted October 19, 2010 I relax while my 60Cx takes a huuuge average, at least 500 readings. On one occasion I let it take 2500 readings because I had to work on the hide anyway, which took time. I've never heard a complaint about my coords using this technique. Quote Link to comment
+Don_J Posted October 23, 2010 Share Posted October 23, 2010 averaging. always gets me great results. it seems that it's mostly (only?) owners of garmin 60 series units who experience worse results with averaging than without. my theory is that those units have a broken averaging algorithm. i don't have any proof of that though. I've never had a problem averaging with my 60CS. I've learned to set the unit on the ground and walk away from it for a minute or two. If you hover over it trying to read the display, you are basically blocking half of the sky. Out of a 157 hides, only one turned out with bad coordinates. Quote Link to comment
+BaytownBert Posted October 23, 2010 Share Posted October 23, 2010 I set my Oregon 450 on the cache site and let it average The results are always within acceptable limits. I do this exact same thing, but then I walk away about 50 feet and seek the waypoint. If there is a clear enough landpoint, I search the coords in Google Earth for confirmation. In 52 hides the only time I was off was when I transposed the numbers wrong. Quote Link to comment
+SeekerOfTheWay Posted October 23, 2010 Share Posted October 23, 2010 I use my BlackBerry and Geocache Navigator. I put the app on compass mode, walk up to the cache quickly and take a screen shot of the coordinates. I do this three times and then use an average. The three screen shots are always pretty close regarding distance. My hide coordinates seem pretty accurate. Quote Link to comment
+TheFinleys Posted October 23, 2010 Share Posted October 23, 2010 I'm curious about the best practices for averaging a waypoint in poor conditions such as heavy tree cover. I was considering using the average waypoint function and walking back and forth on the trail ±10 feet at a consistent rate during the averaging. I was hoping this would give me better satellite coverage through the trees. I still will do the test like everyone has described. What does everyone think about this idea? Quote Link to comment
+===sgb Posted October 23, 2010 Share Posted October 23, 2010 Make sure the GPSr has been turned on with a good, open-sky sat lock for at least 10-15 minutes. This gives it time down update the entire sat constellation data file, which changes daily. I suspect my batteries would go flat in that 15 minutes! Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.