lalittle Posted August 20, 2010 Share Posted August 20, 2010 Is there any way to set the track "record method" to a few feet on a Garmin 62st? The scale is in miles, and I can't figure out how to change it. I'd like to record some trails with more detail, and the "auto" mode, even with the "most often" interval setting, is still more coarse than I'd like. Using miles, the smallest distance I can set is 0000.01 miles, which is 52.8 feet. That's way to high for any accuracy, which makes me think that I'm missing something. Ironically, I can set it ridiculous far in the OTHER direction -- i.e. I can set it to 9999.99 miles, which is just silly. I can set the time to very small increments (down to 1 second), but I prefer to use distance since it's considerably more efficient on memory. Thanks, Larry Quote Link to comment
+ryan3295 Posted August 20, 2010 Share Posted August 20, 2010 Best you can get would be to set the recording method to time and the interval to one second. Doing some math...if you averaged walking speed was 3 mph this would give you a track log point every 4.4 feet. However, the track log would be full after about 2.5 hours. Quote Link to comment
lalittle Posted August 20, 2010 Author Share Posted August 20, 2010 Best you can get would be to set the recording method to time and the interval to one second. Doing some math...if you averaged walking speed was 3 mph this would give you a track log point every 4.4 feet. However, the track log would be full after about 2.5 hours. So there is really no way to set a recording interval of less than 52.8 feet, but I can set it to 9999.99 miles? That just seems utterly, UTTERLY ridiculous to me -- like an actual mistake. Why offer such a uselessly high number, but nowhere near a small enough number? I feel like I'm misunderstanding something here. Even the automatic/normal setting records a LOT more often that every 52 feet. Regarding the time interval, that's exactly why I don't want to use time -- I don't want constant track points if I'm moving REALLY slowly since this will just eat up memory. Using a distance of around 6 feet, however, would create a more detailed track then the "most often" auto setting, but it would still last a lot longer than 2.5 hours. I just can't believe that 52.8 feet is the smallest distance I can set it to. Thanks again, Larry Quote Link to comment
+Redwoods Mtn Biker Posted August 20, 2010 Share Posted August 20, 2010 How about 0.01 yards?... http://gpstracklog.com/2010/05/hands-on-wi...1#comment-75266 Quote Link to comment
+Prime Suspect Posted August 21, 2010 Share Posted August 21, 2010 Best you can get would be to set the recording method to time and the interval to one second. Doing some math...if you averaged walking speed was 3 mph this would give you a track log point every 4.4 feet. However, the track log would be full after about 2.5 hours. So there is really no way to set a recording interval of less than 52.8 feet, but I can set it to 9999.99 miles? That just seems utterly, UTTERLY ridiculous to me -- like an actual mistake. Why offer such a uselessly high number, but nowhere near a small enough number? I feel like I'm misunderstanding something here. Even the automatic/normal setting records a LOT more often that every 52 feet. Regarding the time interval, that's exactly why I don't want to use time -- I don't want constant track points if I'm moving REALLY slowly since this will just eat up memory. Using a distance of around 6 feet, however, would create a more detailed track then the "most often" auto setting, but it would still last a lot longer than 2.5 hours. I just can't believe that 52.8 feet is the smallest distance I can set it to. Thanks again, Larry Is the unit value not changeable? You can't change miles to feet or meters? Quote Link to comment
lalittle Posted August 21, 2010 Author Share Posted August 21, 2010 Is the unit value not changeable? You can't change miles to feet or meters? It's not changeable on the screen where you choose the track log interval (which is where you would intuitively look for it), but thanks to "Redwoods Mtn Biker's" post above it was pointed out that if you go to the "Units" setup page and change the unit type to "yards," you can now go back to the "Track" setup page it will show yards for the interval instead of miles. It's a bit annoying that it's "yards" instead of "feet" (who uses "yards" these days?) but you can at least get a distance down to a usable level. You can then go back to the "Unit type" and change it back to "Statute," and the track log distance will remain at it's current setting. Note, however, that the interval will say "0000.00 miles," which means you can't really tell what it's set to without changing the unit type back to yards, and THEN check the interval distance. In other words: - Change the unit type for the gps to "yards." - Go to the "record method," set it to "distance," and make the distance 2 yards (or whatever you prefer.) - Go back to the Unit Type setting and change it back to "Statute." You'll now have a track recording interval of 6 feet, but the unit will still display feet and miles. Thanks again for the help with this, Larry Quote Link to comment
Grasscatcher Posted August 22, 2010 Share Posted August 22, 2010 Ahhh...so many experts but still so little information....... The unit maximum for logging trackpoints is (1) per second of time,.... so, set to "Auto" /Most Often, or "Time" and One/second, the distance between adjacent points is controlled strictly by the speed you are traveling. Consider ......if you had the unit set to log a point every 10 ft , as soon as you got over 6.XX mph, the distance between adjacent points would exceed the "10 ft setting" due to your traveling more than 10 ft in the previous/last second. The "Auto" /Most Often method is the most accurate/trackpoint efficient method. If you are traveling "straight" it will log fewer points. When the unit detects a change in direction, it will log more points to describe the actual path more closely. (but still only up to a max of 1/sec) Quote Link to comment
+Red90 Posted August 23, 2010 Share Posted August 23, 2010 The "Auto" /Most Often method is the most accurate/trackpoint efficient method. Yep.... You will not get any "better" data taking more points. It will look different, but the data will not be any better than this method. Quote Link to comment
lalittle Posted August 23, 2010 Author Share Posted August 23, 2010 The "Auto" /Most Often method is the most accurate/trackpoint efficient method. Yep.... You will not get any "better" data taking more points. It will look different, but the data will not be any better than this method. We're apparently not talking about the same thing as the earlier posts, because the "Auto / Most often" setting does not create enough track points for my needs. Note that we're not talking about the accuracy of the data of any single track point. We're talking about the shape of the track as a whole, which becomes more accurate as more points are used. The idea being discussed in this thread is to end up with more points along the track, which gives a more accurate picture (i.e. a "better" representation) of the actual track that was taken. This means that curved turns end up looking like actual curves and not sharp points. As stated above, in my particular tests, the "Auto / Most often" setting simply does not put enough points on the track. Complex trails end up having their smaller turns turned into sharp points, or sometimes cut off completely. The resulting track does not look like the trail at high magnifications. When using short time or distance intervals (6 feet or less in my opinion), the track becomes a much closer representation of the actual trail. Larry Quote Link to comment
lalittle Posted August 23, 2010 Author Share Posted August 23, 2010 Ahhh...so many experts but still so little information....... I'm not sure what this means, because the answer to the original question of this thread was indeed provided, and the steps discussed work exactly as stated. Nobody claimed to be an "expert," but they did answer the question. The "Auto" /Most Often method is the most accurate/trackpoint efficient method. If you are traveling "straight" it will log fewer points. When the unit detects a change in direction, it will log more points to describe the actual path more closely. (but still only up to a max of 1/sec) Yes, the auto method is more efficient, but efficiency was not the main point of the original question. The problem with the "Auto / Most often" setting is that at high magnifications, it does not provide enough points to create an accurate picture of the trail. Curves are turned into a few points, and short, quick turns are sometimes completely cut off. Whether or not the auto mode will work for you is completely dependant on how close you want the track to match the trail at higher magnifications. I found that when using an interval of 6 feet (i.e. 2 yards), the resulting track gives me a far better picture of the trail compared to the "Auto / Most often" setting. This is of course due to the specific trails I'm hiking/biking, the speed at which I'm moving, and the magnifications I like to use. It's absolutely not as efficient as the auto setting, but the additional points on curves result in tracks that represent the trail more closely. Note that any given track point is not any more or less accurate -- we're talking about the shape of the overall track, which benefits from the additional points. Larry Quote Link to comment
xyzee Posted August 23, 2010 Share Posted August 23, 2010 Last winter I tried all of the Track settings, to see which showed the curved drive around my Mobile the best, when zoomed in to 20 ft. Both the 1 sec and 1 yd were the best to get a smooth curve. Also, on the path out back to the weather box, where I had to walk around, and up close to, the storage unit, to avoid the drifts, it showed every detail of the turns and even where I slipped off of the icy path. Back to the other selections when the detail stuff isn't needed. Quote Link to comment
Grasscatcher Posted August 23, 2010 Share Posted August 23, 2010 Most Often=Up to and including the maximum logging rate of the unit. ie 1/sec Auto=Unit determines the logging rate automatically, as needed (determined by unit calculated direction traveled) to accurately describe actual path traveled. A straight line path can be more accurately described by only two points than by adding multiple equally spaced points along the line.....but very few hiking paths are perfectly straight lines. But again....it all hinges on the speed traveled that determines the distance between points, whether they be logged at a rate of one / second "manually set" or logged at 1 / second in auto/most often, and therefore the "smoothness" of the track described. Higher magnifications ? I can zoom in to where I can see only the empty space between adjacent trackpoints, but that proves absolutely nothing. Use UTM coordinates and then trigonometry to calculate the actual distance between adjacent trackpoint coordinates logged and I think what you learn will prove to be a real education. Quote Link to comment
lalittle Posted August 23, 2010 Author Share Posted August 23, 2010 Most Often=Up to and including the maximum logging rate of the unit. ie 1/sec Exactly, but "up to" does not mean that it will actually create points that often. It just means that it "can" if it determines the need to. The "issue" is that the "Auto/Most often's" algorithm does not determine that there is a "need" to create enough points for some people's tastes. This results in more crudely drawn paths compared to using short time or distance intervals. Auto=Unit determines the logging rate automatically, as needed (determined by unit calculated direction traveled) to accurately describe actual path traveled. Again, this whole thread was started because the auto/most often setting did not describe the path as accurately as I wanted. Short distance intervals, however, do. But again....it all hinges on the speed traveled that determines the distance between points, whether they be logged at a rate of one / second "manually set" or logged at 1 / second in auto/most often, and therefore the "smoothness" of the track described. The auto setting, however, usually does not create points every 1 second even in situations where I would like it to. The algorithm for determining how many points are "necessary" does not create enough points to give me the detail I'm looking for. Higher magnifications ? I can zoom in to where I can see only the empty space between adjacent trackpoints, but that proves absolutely nothing. I'm not sure what you're saying here, or what you think I'm trying to "prove." By "higher magnifications" I'm just talking about the fact that as you zoom in, a track starts to look more and more jagged. Conversely, as more track points are used, the track looks more smooth. The point is simply that the further you zoom into a track, the more points are necessary to keep the track looking smooth. In other words, if I use 10 points to describe a circle, it will look generally smooth if I'm zoomed out far enough, but if I instead use 50 points, I can zoom in much further before it stops looking smooth. This is what I meant when I talked about "higher magnification. On the trails I'm hiking, short distance intervals (2 yards or less) result in tracks with more points on the curves than the "most often" setting creates, so the trails end up looking much smoother at higher magnifications. Use UTM coordinates and then trigonometry to calculate the actual distance between adjacent trackpoint coordinates logged and I think what you learn will prove to be a real education. I'm getting the impression that we're just not on the same page here, because I'm honestly not clear what it is you think I don't understand. The original points of this discussion are relatively simple: 1) More track points create smoother, more accurate tracks, and 2) the "auto/most often" setting does not create as many track points on tight curves as using a short distance interval does. The question of how to get the unit to use a short distance interval was already answered by Redwoods Mtn Biker (see his post above), and my tests (and xyzee's as well) clearly show that this works to create smoother tracks. If you want to test what I'm talking about here, just set the unit to "auto/most often" and walk a short distance with some tight turns, then set the unit to a distance interval of 1 yard (or even less if you want to really see the difference) and walk the exact same path. You'll find that the track created with a 1 yard interval is a much closer representation of the path since more points were used, and therefore more detail was preserved. Larry Quote Link to comment
Grasscatcher Posted August 24, 2010 Share Posted August 24, 2010 The main problem is that you don't seem to understand that you really ARE NOT getting what you THINK you are getting. See if this makes sense.....Normal walking speed=3mph At 3 mph that is 5280 X 3=15,840 ft /hr....then divide that by 3600 to get ft / second of time which = 4.4 ft/sec. Set at 1/sec and traveling at 3 mph walking speed, your unit will log the maximum number of points it is capable of and that will be once every 4.4 ft. ( HINT: 4.4 ft is further than 1 yd ! ) Traveling faster than 3 mph will result in TP being logged at an even greater distance apart, or you could change your method of traveling to crawling to get the points closer together. So setting your unit to log at interval of 1 yd or .01 yard falls in the category of ..........DUH ! Then next you will view your tracks displayed over maps (USGS quads) created with mapping tolerances of +/- 40 ft. Then for the next exercise in futility you need to follow the exact same single track path multiple times and on different days to check the repeatability of your unit and methods. What you thought you knew "'taint necessarily so" Quote Link to comment
lalittle Posted August 24, 2010 Author Share Posted August 24, 2010 The main problem is that you don't seem to understand that you really ARE NOT getting what you THINK you are getting. I think the problem here is that there is a misunderstanding of what it is that you say I "think I am getting." What I am saying is EASILY demonstrated -- you just need to do the test as I outlined it. I'm not sure what it is that you think I'm not "getting," but THAT is where the misunderstanding is occurring. Both xyzee and myself have already proven the point I am making by actually testing it. If you do the test I outlined, you'll see EXACTLY what I'm talking about, which is precisely what I "think I am getting." I'll outline what I'm saying one step at a time in the hopes that it will shine a light on the misunderstanding here: 1) Assuming a similarly even distribution of track points, using more points to describe a path will result in a more accurate path -- i.e. describing a circle with 50 points creates a more accurate representation of the same circle described with only 10 points. I don't think that you have any disagreement with this first point given that it is not dealing with the specifics of the gps. The idea of more points = more accurate track, however, is at the heart of the issue here. Just to be totally clear, we're assuming the same general accuracy for all the individual points. 2) When hiking, using the "Auto/Most often" ends up creating FEWER POINTS along curved sections of a trail than using a short interval such as 1 or 2 yards. I'm still not sure, but I think this may be where the misunderstanding is taking place. Do you disagree with this statement? If so, then you should test it. You'll find that using a short interval of 1 or 2 yards will create more points along the same path compared to using the "Auto/Most often" setting. 3) Since creating more points along the same section of a path results in a more accurate representation of the path, and since using an interval of 1 or 2 yards creates more points than using the "Auto/Most often" setting, then it follows that using a 1 yard interval will create a more accurate representation of the path. This is simply a conclusion using the transitive property on the first two points above. It's also easily verifiable in practice by walking the same section of path using both track logging methods. Which of these do you disagree with? If you don't disagree with ANY of these 3 points, then there is no argument because these three points are the entirety of what I'm saying. If you disagree with one or more of these statements, please tell me which one. See if this makes sense.....Normal walking speed=3mph At 3 mph that is 5280 X 3=15,840 ft /hr....then divide that by 3600 to get ft / second of time which = 4.4 ft/sec. Set at 1/sec and traveling at 3 mph walking speed, your unit will log the maximum number of points it is capable of and that will be once every 4.4 ft. ( HINT: 4.4 ft is further than 1 yd ! ) Traveling faster than 3 mph will result in TP being logged at an even greater distance apart, or you could change your method of traveling to crawling to get the points closer together. I don't disagree with this at all, but none of it applies to the point I'm making, which leads me to believe that there is some basic misunderstanding of what I was saying earlier in this thread. So setting your unit to log at interval of 1 yd or .01 yard falls in the category of ..........DUH ! I have no idea what you're trying to say here. Are you under the impression that using the "Auto/most often" will always log every 1 second? Maybe this is where the misunderstanding is happening. I can absolutely confirm that when set to "Auto/Most often," the 62 definitely does NOT log a point every second. There is an algorithm at work that decides how often and how close to log points, and this algorithm does not log that often except in specific circumstances. On the trail, at normal hiking speeds, I get points that are WAY over 4.4 feet apart even when I'm travelling slower than 4.4 ft per second. Because of this, curves get turned into points or squares, and small, tight maneuvers get completely cut off. The important point, however, is that this DOES NOT HAPPEN when using a 1 or 2 yard interval on the same path. Then next you will view your tracks displayed over maps (USGS quads) created with mapping tolerances of +/- 40 ft. This is another possible place the misunderstanding is occurring, because mapping tolerances are irrelevant to what I'm talking about, which is the overall SHAPE of the path, NOT its placement on the map. In other words, I don't care if the path itself is a bit off of where it should be on the map, I just want to be the correct overall shape. Then for the next exercise in futility you need to follow the exact same single track path multiple times and on different days to check the repeatability of your unit and methods. Again, I just don't know what you're getting at here. Where is the futility? What I said above works, and it's very easy to test and confirm, so what about this is "futile"? I set the unit to a distance interval of 1 or 2 yards, and I get more points along turns compared to the "Auto/Most often" setting. This gives me a more accurate track shape which more closely matches the shape of the trail. This is all 100% verifiable by looking at sat images of the trail, and comparing them to the track that was created. What you thought you knew "'taint necessarily so" I know what I can test and verify, and I've tested and verified what I've said above. If you think I'm not "correct" in what I'm saying, than either I haven't explained it clearly (i.e. we're not talking about the same thing), or you haven't tested it. Don't take my word for it, just test it yourself. Larry Quote Link to comment
+user13371 Posted August 24, 2010 Share Posted August 24, 2010 What I'm getting out of this discussion is that if you asked the GPS to record 1 point per second, you're gonna get the most precise track that the GPS is capable of recording -- because that's really the maximum/fastest rate it can record points. If you set it to any "Auto" mode it's gonna try to economize on the number of points and you WILL lose some detail. Have I got that wrong? Quote Link to comment
+gpsblake Posted August 24, 2010 Share Posted August 24, 2010 (edited) The idea of more points = more accurate track, however, is at the heart of the issue here. Disagree with that. More points are not more accurate. Proof: Set your GPS down and walk away for an hour with it set to record every second. Then look at the track it made. You'll see your GPS did all sorts of "walking" around. Especially if the GPS is in a dense environment like the deep woods. Another example. You are in the deep woods and set your GPS on a spot that is N25 00.000 W81 00.000.. You will notice it will bounce around say from N24 59.995 to W80 59.995 to N25 00.005 W81 00.005. But your GPS did not move. But the tracklog will show that it has moved. Basically you want the unit to record every 4 feet while the unit itself is nowhere that accurate in the first place. If GPS units were more accurate, your point that more points = more accuracy would be correct. Edited August 24, 2010 by gpsblake Quote Link to comment
+ryan3295 Posted August 24, 2010 Share Posted August 24, 2010 The main problem is that you don't seem to understand that you really ARE NOT getting what you THINK you are getting. See if this makes sense.....Normal walking speed=3mph At 3 mph that is 5280 X 3=15,840 ft /hr....then divide that by 3600 to get ft / second of time which = 4.4 ft/sec. Set at 1/sec and traveling at 3 mph walking speed, your unit will log the maximum number of points it is capable of and that will be once every 4.4 ft. ( HINT: 4.4 ft is further than 1 yd ! ) Traveling faster than 3 mph will result in TP being logged at an even greater distance apart, or you could change your method of traveling to crawling to get the points closer together. Isn't that the same math I did in the beginning? Thanks for restating my point. Quote Link to comment
+ryan3295 Posted August 24, 2010 Share Posted August 24, 2010 What I'm getting out of this discussion is that if you asked the GPS to record 1 point per second, you're gonna get the most precise track that the GPS is capable of recording -- because that's really the maximum/fastest rate it can record points. If you set it to any "Auto" mode it's gonna try to economize on the number of points and you WILL lose some detail. Have I got that wrong? Nope, especially going back to OP's question. Quote Link to comment
+user13371 Posted August 24, 2010 Share Posted August 24, 2010 Yes, but the OP doesn't want to use time-based tracking because it'll eat up memory/points while standing still Quote Link to comment
lalittle Posted August 25, 2010 Author Share Posted August 25, 2010 The idea of more points = more accurate track, however, is at the heart of the issue here. Disagree with that. More points are not more accurate. We're talking about two different things here. You're factoring in poor reception issues, which is a completely separate issue. When I talk about "more points = more accuracy," I'm talking about a basic concept, which is that if you describe a circle with 50 points, it creates a more accurate representation than if you describe the same circle with only 10 points. If we assume a certain "basic level of accuracy" without gross errors due to poor reception, then more points will create a more detailed drawing of the trail. This is usually the case on the Southern California trails I'm talking about -- I rarely have any significant reception problems, and my tracks are always more accurate when more points are created . Proof: Set your GPS down and walk away for an hour with it set to record every second. Then look at the track it made. You'll see your GPS did all sorts of "walking" around. Especially if the GPS is in a dense environment like the deep woods. As stated above, I've already tested this on numerous trails, and can very clearly see that the "auto / most often" setting results in tracks that are far more crude than when I instead use a short distance interval of 2 yards or less. The short interval allows me to see curved sections of trail that were completely "averaged out" with the auto setting. That said, I would argue that a track that has lots of "walking around" can still be a more accurate overall representation of the trail compared to a track that shows points or squares for smooth turns, or that completely cuts off certain curved sections. In other words, I can generally tell by looking at a track when errors have occurred, and I can factor this in. I cannot, however, tell when a straight section of track is actually a section with several small curves. Basically you want the unit to record every 4 feet while the unit itself is nowhere that accurate in the first place. In my experience, the gps is actually accurate enough such that more track points give you a more accurate overall "shape" of the trail. The whole thing might be a bit off on the actual map or sat image, but the shape is usually very accurate, and I find that looking at the overall shape of the track is a much more reliable way to see where I am. I personally find that 2 yards seems to be a pretty good sweet spot to create an accurate picture of the trail without eating up too many track points (which is why I don't like the time intervals -- they create points that are not needed.) If GPS units were more accurate, your point that more points = more accuracy would be correct. Again, I'm just going by what I'm actually seeing on the unit. I'm sure it's largely location dependant, but in SoCal, the tracks I get with somewhere around a 2 yard interval are CLEARLY more accurate than those that I get with the "auto / most often" setting. Larry 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.