Garmin GPSmaps 64s and 60csx

my Garmin eTrex 30 has the same issue…

I have had very good luck with my 62s (and my eTrex before it) I have found the units get better with each generation, particularly in terms of satellite acquisition and holding a signal.

Curious as to how you determined that your GPS is off by as much as 25%. Mine is usually pretty much spot on compared to other GPS units and most maps. Where you may see problems is if you leave it on when you are in one location for a long period of time (such as lunch breaks or in camp) as it does wander and essentially think you are still moving.

I would affirm some of James' comments here… first, how do you figure 25%
Trail mileages are notoriously inaccurate.
about the only way I can think to calibrate accurately is to mark a section of road with your car and the walk it with your gps to compare.

I have not done that but I generally figure my eTrex 30 is about +10%
If I am concerned about accuracy (which I rarely am) I turn it off for long breaks
Like James indicated, it's the long breaks that will throw the mileage off more than anything.

Really, I don't get too compulsive about trail mileages as there is no way to measure them to total accuracy anyway… gps or otherwise.. they're all estimates…

billy

My Garmin 60Csx has the same issue

The Odometer can be +-10% off, 25% is unusual but maybe the worst

When I download a track and analyze it with Mapsource, I think it's more accurate

I think the odometer software is inaccurate

If you are stationary, for each reading there will be an error, because that's just how it (or anything) works. There will be a small cloud of points. The cloud can be bigger or smaller depending on satelites, interference from cliffs and trees,… If the odometer was adding the distance between these points, it would slowly increase even if you weren't moving.

I've noticed that usually, the odometer doesn't increase when I've stopped, so the odometer software must figure out the distance between points is small, and just ignore it. There must be a bug in how they do this that causes some errors.

Another thing is, even if you walk in a straight line, for example, the readings will zigzag around a little. The distance of this zigzag is a little bigger than the distance of a straight line. That's another error. That would affect both odometer and analyzed track.

Another thing is, if there is a switchback on a trail, and the gps doesn't take a point right at the corner, then the gps distance will be a little less.

Another thing is, occasionally it will calculate my position maybe 0.1 miles away, and it gradually gets back to the real position, so that would add 0.2 miles.

My conclusion is the gps is only so accurate. If it's within 10% that's pretty good. But it isn't really that important.

It does tell you where you are, and if you have a downloaded track of a trail, you can see where the trail is if you get off it. Or I can get the distance to some point I'm trying to reach.

I analyze tracks when I get back home. Delete most of the points to remove most of the gps error. Also remove points when I'm not moving or if I explore a viewpoint or campsite or something off the trail. That distance seems pretty accurate.

I have found Garmin customer support is pretty worthless. I've asked the same question and got a worthless response, like send in my unit at my cost and they'll check it out. They have the biggest share of the market so they don't care. I think they care more about cars than hikers because market is bigger.

My back cover broke. There's a flimsy plastic catch. They said they'de sell me a new one for $30 or something. I said this was a flimsy design and they should just send me a new one free. They said no. I just fixed mine with epoxy, a piece of aluminum, and a cut off nail. I am now giving them more than $30 of bad PR : )

One way to check it is to walk a straight line, like 1 mile. Like along a straight road. The real distance is the distance between the beginning and end. You can compare this to the odometer and the analyzed track. The error of the two endpoints is small. The small errors of each point of a track add up to be significant.

Another way to check it is to walk the same trail many times. For each one, analyze the track – remove error points and such. Compare them. Yeah, it doesn't give you absolute, but it will give you repeatability error. You can have an analyzed track stored, and then walk the trail again, to make sure you haven't edited out any switchbacks or anything.

I know it's being inaccurate just based off of my usage on several trips. I've done a 40 mile trip on popular, well-maintained and marked trail where at each signpost i could see my odometer overestimated more and more. At 10 miles it would read 12, at 20 it would read 25. By the time we were done it read 50 miles for a 40 mile trek. On a 10.5 mile day it read as high as 15 or 16 miles when i was recording points auto "less often" (still havent tried least often yet).

I'm basing my estimates of the "actual" mileage of of topo maps, trip reports, etc. and though i try to turn the gps off when im breaking for longer than 10 minutes and I don't always remember to do so. But still, even with a few hundred extra feet tagged onto the reading, it shouldn't be THAT bad!

When i load the tracks into basecamp, the tracks look pretty accurate and the distance in basecamp is much closer than the GPS reads, and even more accurate after cleaning up and filtering the tracks. But that doesn't do me much good on the trail.

It's very important to me to know how far i've gone, how fast i'm going, and how long it could take us to get to where we're going. a $300 dedicated GPS device should be able to tell me more than looking at a map and ballparking an estimate.

It's also pretty confusing that there's 3 different data readings for the distance traveled.

There's the Odometer, the trip odometer, and the track distance. I assume track distance is what i should be reading, but i have read the Odometer up until now because that's been the default.

Also i'm considering changing the track record interval to distance and using something like a 1/20th of a mile. Surely that would help with the "wandering" when taking a break?

I think part of the problem isn't the "wandering" when taking a break, but that the software incorrectly tries to correct for this, so it's a software bug

But, what difference does it make if the odometer says 50 miles after a 40 mile hike?

It's important to know how far it is to the next campsite or stream or whatever. Even with a 25% error, if a campsite is 4 miles but your odometer says it's 5 miles, it's not that big a deal. Usually it's more like 10% or less.

I agree it should work as well as possible, and someone at Garmin should be into this, figure it out, and fix it.

I wonder if DeLorme is any better.

Not defending Garmin or saying 25% is acceptable…. but:

"I know it's being inaccurate just based off of my usage on several trips. I've done a 40 mile trip on popular, well-maintained and marked trail where at each signpost i could see my odometer overestimated more and more."

The milages on sign posts are only estimates. It would be a very rare trail that someone went out with a measuring wheel to get a truely accurate distance. Some of the old trail signs in Yosemite can be 50% off !!! I have found some signs 10% off and some 50% and some 30% off in Yosemite… If I had been trying to compare those mileages to a GPS I would have been very confused and if I assumed the signs were right I would have been very upset with my GPS…

"I'm basing my estimates of the "actual" mileage of of topo maps,
trip reports, etc."

The mileages on topo maps are also estimates. No one goes out in the filed and uses a surveyor's wheel to measure them… they're estimates. Same for trip reports…

You get used to it.

Billy

Don't always blame the GPS! Just sometimes. There are many sources of 'error'. What you need to rememebr is that many of the 'features' on a GPS are add-ons which the programmers threw in because there was memory and processing power available. Then marketing hyped these 'features' up.

1) If you are near cliffs etc you can get reflections. These will really screw the readings up. When I was reviewing the SPOT I the GPS in it put me on the wrong side of a very big inlet. This was due to the cliffs above me, and is a physics problem.

2) With some GPS units (older ones?), the software would assume the first reading was right for some time even if subsequent readings were quite different. So a bad reading at the start could put your track out of position for many miles. Yeah, that WAS a software problem, but modern units have mostly corrected it.

3) If the weather is changing the propagation path from satellite to you can effectively change a bit, so your position drifts. You can see this when you are quite sure you are stationary. Physics problem.

4) If you are under heavy tree cover the signal can also get upset. This normally appears as a loss of signal. Physics problem.

5) Distance errors are subject to several quite different problems.

5.1) The track shown on a map, even a topo map, can quite easily be just plain wrong. You can see this with Google Earth when it overlays the topo track on the satellite photo, and you can see the real track off to one side.

5.2) The real track on the topo map may not lie on the 'official' route. Often the bureaucrats drew tracks on topos with little regard to reality, so the track makers just went where they wanted. Gullies, cliffs … This can create interesting legal problems. (The Council road past my place went THROUGH my property in TWO places in the early days, rather than around it. I made the Council move the road.)

5.3) Estimating tracks distances from a map is hopeless. The walked route can easily be 25% greater due to wiggles on the ground. Even an odometer wheel can get distances a bit wrong – although not as much.

I think you may be expecting too much accuracy.

Cheers

Steve, if you read all of the replies that you got here, you will notice that they mostly agree. Your expectations for GPS distance accuracy are high.

As you probably know, GPS distance measurements along a trail happen through a sum of individual small distance measurements. Let's say that the GPS receiver takes a position fix every second, and between the start of one second and the end of one second is a distance of six feet. The receiver might try to display to you that your distance traveled is six feet and your velocity is six feet per second. However, you may have been weaving back and forth in a zigzag pattern, so the actual distance that you traveled was eight feet. You see right away that the sampling interval (in time) might make a big difference in what it displays as distance traveled. There is no single sampling interval that gets the best results for all conditions. However, note that military GPS receivers have a much quicker sample rate, so they get better accuracy.

Take another example. If you start to go into a tunnel, and the receiver gets a position fix outside the entry, and then if you travel through the tunnel and exit on the far side, the receiver will get a position fix somewhere after that exit. Now, should it record the missing tunnel period as distance traveled, or not? This is where the scientific accuracy of the receiver must be separated from the heuristics of what to display. Some users would argue that the _apparent_ tunnel length must be added to the total distance. Some users would argue that the tunnel distance should not be added because we really don't know where we were in there. This is why some leading companies like Garmin offer different profiles for users in some receiver products. Different profiles allow the receiver to make different assumptions about what the user expects for the display.

Let's take another example. Suppose that I have my GPS receiver running while I carry it up a steep mountain ascent. Suppose that the horizontal distance covered is one mile and suppose that the vertical elevation gained is a mile. What should the receiver display for total distance covered? Should it be the hypotenuse, about 1.4 miles? Should it be 1 mile? My point is that everybody has different expectations.

Some forest service guy may have designated one trail segment as being ten miles long. What does that mean? You really don't know how he was measuring it. Without knowing that, you really don't know how much faith to put into that ten miles.

One big factor to consider is multipath interference. The signals from space are supposed to come down and hit your receiver antenna directly. Suppose that they bounce off a smooth rock cliff on the opposite side of the canyon, then hit your receiver antenna. That can create position inaccuracy of all kinds, and that generally results in false distance being added in. There isn't much of a way to avoid that except to use a heavy expensive GPS antenna. What I do is to watch some statistics on my receiver. Some receivers show VDOP and HDOP. This is the vertical dilution of precision and the horizontal dilution of precision. Basically, it allows the user to see how much faith to put into the accuracy of specific position fixes. If I see VDOP and HDOP numbers going to hell, then I know not to put too much faith that the trail segment was exactly ten miles. On the other hand, if the VDOP and HDOP stay good, then I can put more faith in the ten mile measurement.

Another interesting experiment is to walk the ten-mile trail segment, say north to south, and keep your receiver in a good view of the sky the whole way. Now walk the same trail segment a day later, going south to north. You know that it is the same ten miles, but you may see a difference in what the receiver displays. That may be from a different view of the sky, which means using different GPS satellite signals.

Just about any receiver can give you inaccuracies when you are using only the normal internal antenna and you have it oriented in a less-than-perfect angle, like pointed ahead on the trail. Having it pointed up at the sky with no obstructions around it will get you better results.

I was climbing Mount Washington NH for the first time with a couple of feet of snow on the ground. Due to cold, I had the receiver in my inside parka pocket to keep the display warm. I had an external GPS antenna mounted on top of my head, between my inner hat and outer hat, so there was a coaxial cable snaking down my neck. That allowed me to keep a good antenna view of the sky, and I kept very high accuracy for the whole day. I know that it sounds crazy, but it gets good results.

I was climbing a big peak, almost 7000 meters high. All the way going up the mountain, I was getting squirrelly readings depending on which way I faced, the time of day, and which side of the mountain I was on. Once I reached the summit, it was perfect, and the displayed elevation was only one meter off the official elevation. I had had "view of the sky" problems.

So, you just have to use these things a bit and you will gradually get a feel for accuracy.

–B.G.–

I suppose… but then why are the tracks much closer to these reported distances when i've loaded them into basecamp and cleaned them up? That 15 mile day just turned into 10.7 a mere .2 miles off the maps reported distance.

I've traveled in all sorts of terrain with this GPS, it's a consistnent problem regardless of if i was hiking in a canyon or forest or open desert or beach, and the points it records ARE accurate and consistently so once i take them home and even before i clean them up. I also try to carry my gps at the top of my pack, always out in the open, with the antennae facing up and on the side of my pack farthest away from cliff walls.

Sounds like a software issue to me, and it sounds like one that hasn't changed in 2 generations of the device.

Makes me think i should try picking up a cheap android phone to try modding the gps antennae. If i could have a reliable and powerful antennae on an android there would be almost no reason to use garmin. One of the main reasons i got a standalone gps was because i can't rely on my phones gps to get a fix out in the thick of it or at the bottom of a canyon.

"Some of the old trail signs in Yosemite can be 50% off !!! "

Billy, when I first started seeing those in Yosemite, I was confused. On some specific trails, I figured it out. They had stuck the trail signs along the right trail in the wrong order. One would say that we are eight miles to the finish. Then the next one would say that we are five miles to the finish. Then the next one would say that we are six miles to the finish. All that, even though we were heading the right direction on the right trail.

–B.G.–

"If i could have a reliable and powerful antennae on an android there would be almost no reason to use garmin."

Steve, that might be part of the problem, but you are not going to like the full solution.

The best GPS antenna solution to eliminate multipath is called a choke-ring antenna. Go ahead and buy one of those, but then you have to have a GPS receiver with a so-called external antenna port. Oh, by the way, a choke-ring antenna tends to be pretty big.

Have you ever seen a land surveyor out doing accurate measurements? He will have a vertical rod that he sits on the ground surface, and the rod sticks up about six feet. On top there is a "flying saucer" shaped antenna. That is a choke-ring. The rod gets it up high enough that he is not shielding it with his body. The actual antenna design is quite complicated, but it gets excellent results. They get used a lot in geophysical monitoring stations. These are measuring the seismic movement of the Earth, and they measure things down to millimeters. Expensive and heavy, but excellent.

My point is that you are not going to get professional-grade results from a consumer-grade receiver and antenna unless you have perfect conditions.

–B.G.–

>It's very important to me to know how far i've gone, how fast i'm going, and how long it could
>take us to get to where we're going. a $300 dedicated GPS device should be able to tell me
>more than looking at a map and ballparking an estimate.

Why do you think that, because it is a technological device, because it cost $300. GPS units are no more or less accurate than a map and compass, just more convenient. If you are ballparking an estimate on a map, then maybe so, but if you are really reading your map, keeping track of where you are, using your compass, and possibly an altimeter you will be every bit as accurate either way.

The distance log on your GPS is a fun feature, but the real utility of the tool is to use the GPS location. Combined with a map in the unit, or a paper map you should be able to get an accurate idea of how far you have to go.

If you are just using the log, and your guidebook says camp is 10 miles away, and you are at 5 on the log, who knows where the error is, the GPS, the guidebook? I hope you aren't basing your complaint on trail signs which are notoriously inaccurate in most areas.

I wouldn't need THE most powerful antennae. Just something on par with a garmin oregon (which i dont think uses a quad-helix?) or if it's possible mod a quad-helix onto the devices gps reciever. I've seen a couple videos of people doing diy gps improvements with copper wire/foil with sexy results!

James explained it.

I used to help my customers with a system that they had just paid $50K for, but the GPS portion was worth no more than $10K or $20K.

We figured out how to get almost-perfect position accuracy. Nail the antenna down in one spot and average the results for four hours. Now reset the power and take it to the second position and average there for four hours. I kid you not.

–B.G.–

Steve, I'm not sure if you are trying to understand all of this.

A quad-helix was quite common as an external GPS antenna about twenty years ago and up to about five years ago. I've used literally hundreds of them. They get good results. Unfortunately, by modern standards they are bulky and use more power. You intend to carry that around with you? Interesting. Most of them are either cone-shaped or egg-shaped and about the size of a coffee mug.

I purchased my first GPS receiver in 1994, and it had a quad-helix. Although it was good, it was not completely immune to multipath.

–B.G.–

@bob

I'm trying of course. I just dont think you guys are understanding the reasoning for my questions and curiosity.

I'm not asking for a 100% accurate reading from the device. 10-20 feet of accuracy is plenty for me, and i understand that means for every reading the gps takes it could be adding 20 feet of inaccuracy every 10 seconds or so for hours of hiking.

I also understand that any receiver is limited by physical and environmental conditions and that trail and map reported distances are also inaccurate to a certain degree.

That's not really what the thread was about, because all things considered, the GPSmaps 64s does a REALLY good job of being accurate enough when it gets its readings on the move, it just doesnt seem to REPORT that information in an accurate way.

So my reason for starting the thread was more to ask if this is a problem with this recent GPSmaps iteration, a problem with the line of GPS, or a problem with Garmin's firmware design. Sounds like it's the latter. If i can't get it to report more accurately by tweaking settings, that makes it less useful to me than my phone's gps which was in no way intended for backpacking and did not cost me $300.

The GPSmaps series all use quadhelix antennae. But it sounds like from what you're saying its probably not something you can just solder on to a phone's built-in gps like you can with a copper wire.

"10-20 feet of accuracy is plenty for me, and i understand that means for every reading the gps takes it could be adding 20 feet of inaccuracy every 10 seconds or so for hours of hiking."

The issue is that the 10-20 feet of accuracy tends to be rather random, and not all 10-20 feet in one direction.

It is entirely possible for the GPS receiver to display a distance of 9 miles when the trail was really 10 miles.

The whole user secret is in humanly recognizing the bad effects around you so that you know how much faith to put into the display accuracy. Similarly, if you get into some slot canyon in Utah, you may not be able to get any kind of position fix at all, and not a great deal can be done about it. You can get a good position fix on one side and then do some dead reckoning in the canyon.

Like I stated, some GPS receivers have user profiles that allow the user to choose a different set of heuristics and assumptions to be used internally. Apparently that doesn't interest you. The issue there is that most big GPS companies like Garmin invest a lot of engineering dollars in their heuristics, whether good or bad. They typically do not explain very closely how they work or why.

The other thing to watch out for is a microwave radio transmitter site. These are often used by telephone companies, and they can be identified by a 2-, 3-, or 4-foot diameter parabolic dish antenna on top of a tower. Sometimes they splatter out so much microwave interference that they will halt GPS reception (1.57GHz)

–B.G.–

"I just dont think you guys are understanding the reasoning for my questions and curiosity."

Like Roger stated, often it is a physics problem.

–B.G.–

"but then why are the tracks much closer to these reported distances when i've loaded them into basecamp and cleaned them up?…Sounds like a software issue to me, and it sounds like one that hasn't changed in 2 generations of the device."

Exactly! That's my logic too.

Pisses me off that Garmin doesn't care

But, the GPS is still very useful. The odometer just isn't as accurate as it could be.

"It is entirely possible for the GPS receiver to display a distance of 9 miles when the trail was really 10 miles."

Right, but i WISH my gps was only 10% inaccurate in ideal conditions, and the essence of what i was getting at is: is 25-50% inaccuracy really acceptable for a $300 GPS in 2014 and has that been the norm since forever?

I'm assuming user error for now, but most of what i'm hearing is pointing dangerously close to "wasted $150 on the new model when the 2 gens old one works just as well if not better".

This is my first GPS remember, and I would've expected noticable improvement in both accuracy and heuristics the past decade or two and it seems like lots of people report 10% inaccuracy in older devices.

I'm definitely interested in different heuristics profiles, but I've already looked through the device's settings multiple times and the only thing that comes close are profiles for automatic routing (which VERY few maps are routable, yet it can tell me the vertical distance on any given point on the topo?! UGH) and that has nothing to do with recorded tracks, only routes.

" I'm assuming user error for now, but most of what i'm hearing is pointing dangerously close to "wasted $150 on the new model when the 2 gens old one works just as well if not better". "

Yes, I think that is about right.

I'm sorry that you purchased a model based on what they wanted to sell you, and not what was the best tool for your purpose.

I currently own four GPS receivers. The one that I like the most is one that I purchased in 1997. It has no map database and the screen is monochrome. The newer receivers spend a lot of effort looking good, but the accuracy is no better.

Too many users spend their effort complaining about a lack of accuracy, but they are never willing to spend any effort learning how to extract better accuracy through practice.

One place to start is to look up the local survey monuments for your area. USGS has a listing of them. Then go find one of them and see how your receiver's display compares to the official lat/long/alt. Note that in some areas that are geologically active, the official monument is sliding in time. I have one in my neighborhood, so anytime I have doubts about one receiver or another, I can go find the monument and analyze the numbers.

–B.G.–

Personally, I've never seen the utility of using a GPS odometer for any kind of trip planning. I always either have a paper or digital map and know how many 'finger widths' = a mile… and just step-off and count finger widths to know how far I have to go or how far I've come. That is about + or – 20% accurate which has always been good enough for me.

I do understand about buying an expensive toy and not having everything on it work right.

But that's the way consumer grad GPSs are… or at least the Garmin ones I've owned.

But, again, it's really no big deal as there is so much variation in the trail mileage estimates (maps, signs, guide books, etc. all being estimates) and my walking pace and terrain that any kind of ETA is unknowable exactly anyway. Besides, I go out backpacking to get away for the compulsions of schedule, knowing things exactly, and other brainiack things anyway.

Still, the most important thing about a GPS is the ability to get my location and that's what I bought it for and it works great for that.

Billy