Results 1 to 18 of 18
  1. #1
    Just hit it with speed
    Reputation: GFisher2001's Avatar
    Join Date
    Mar 2006
    Posts
    440

    Edge vs Vista HCx v 60CSx

    My original post was a little to wordy.....

    I've heard from multiple sources that the EDGE computer has difficulty getting accurate trip distance and speed on tight twisty singletrack with tree cover when riding at "race" speed. Comparing distance from an edge computer to a lap at a recent bike race edge users were getting upto a mile difference. Course was 7 miles (trusted) vs Edge GPS users getting 6 miles.

    Is the 60CSx or the Vista HCx any better or more accurate in tight twisty singletrack with tree cover in calculating distance and speed while riding?

    I will be pulling the trigger on the 60CSx or HCx once I get some feedback. Anyone have any suggestions for websites that i can get free trail info to upload to my GPS for trails in New England (NH, ME, MA, VT)? Is motionbased the best?

    Thanks in advance!
    Last edited by GFisher2001; 05-17-2009 at 06:31 PM.
    Don't hate on the minivan!!!!!!!

  2. #2
    saddlemeat
    Reputation: bsieb's Avatar
    Join Date
    Aug 2003
    Posts
    2,487
    I can't answer your question directly, but I do know that the sampling rate affects the distance calculation. I have observed that a friend's Edge 305 consistently matched my well calibrated cycle computer in the past. I have also observed that a race course layed out with a well calibrated (i.e. the average of 3 marked and measured roll outs at riding tire pressure with geared up rider onboard entered for wheel diameter) cycle computer consistently shows up as up to 10% shorter on the racer's gps units. Then consider that a well calibrated 29er will show about 7% less distance than a similar 26er. Turns out distance is a pretty subjective thing if it's not measured in a straight line.
    A Useful Bear is a handy thing.

  3. #3
    MTC rep for LPCPCAC
    Reputation: bankerboy's Avatar
    Join Date
    Oct 2006
    Posts
    1,064
    Part of the problem with distance discrepancies, as I understand it, is most GPSs cannot translate changes in elevation into miles traveled. Units usually measure point to point, horizontally only. Even if you have a point every couple of inches, you would miss out on mileage.
    Think about 2 points 1 meter apart. Lets assume you have the perfect GPS. If you move from one point to another, it would read 1 meter. Now keeping the same distance for the points but put one of the ground and another on your roof. Vertically they are 20 feet apart, but horizontally, they are still one meter apart. Most GPSs out there can see both the elevation change and the distance traveled. Unfortunately, they cannot combine them to come up with the 20+ feet you actually traveled to cover the 1 meter in distance horizontally.
    So if you rides are hilly, you can expect some discrepancies. I rely on my computer for closer distance measuring but even that is not fool proof.

    As far as recommendations, I love my 60CS (CSx's older brother), I have been able add all sorts of maps for unknown destinations and it is easy to read in daylight. I keep it mounted to my stem. No needs for guides on new rides if I don't want one. For maps, I have had good luck with MotionBased in the past but have heard there may be some problems and big changes in the works. For topo maps www.gpsfiledepot.com You can get all sorts of states topo maps down to the 20 ft increments in some cases. These are fully supported by the 60CSx. You can always pay a small fortune for the Garmin versions, but I was less than impressed about their 100K topo. Check out the free sites.

    Whatever you decide, good luck!
    Apathy will get you exactly what you deserve

  4. #4
    Flyin Pig
    Reputation: IvanLasston's Avatar
    Join Date
    Jul 2006
    Posts
    159
    Quote Originally Posted by bankerboy
    Vertically they are 20 feet apart, but horizontally, they are still one meter apart.
    I believe this conversion rate is what caused the mars orbiter to become the mars impactor.

    http://www.cnn.com/TECH/space/9909/30/mars.metric.02/

    On a serious note - don't forget most Garmin Edge have 2 data recording modes. Garmin has locked down the pdf manual so you can't cut and paste out of it anymore but there is a smart record which is the default - and isn't as accurate, and the every second record which means you can only grab 4.5 hours of riding. These numbers are from the Edge 705 manual. The Smart recording marks where you change direction, speed or heart rate. So you may want to look at the Edge again and set it to accurate data recording before making a choice.

  5. #5
    Just hit it with speed
    Reputation: GFisher2001's Avatar
    Join Date
    Mar 2006
    Posts
    440
    Wow I went from thinking my post was going to end up in no man's land to three very very informative posts. Love me some MTBR.

    Think I'm going to pull the trigger on a 60CSx.

    Now for tracking down some trails.

    Motionbased seems a little difficult to figure out. Everytime I search for trails all it shows me is Anchorage Alaska Trails!!!!
    Don't hate on the minivan!!!!!!!

  6. #6
    MTC rep for LPCPCAC
    Reputation: bankerboy's Avatar
    Join Date
    Oct 2006
    Posts
    1,064
    Motionbased isn't too hard to locate trails. Use the second tab at the top of the screen called "quick filters" or something like that. Add each filter seperate from the next i.e. State name is one filter and city or county is another. Then click on run report. This will narrow it down to at least the area you are interested in. I have used it this way when traveling to new areas. The down side is not everybody uses the same names for a trail so you might have to click on many links to find what you want. Once you have done that and found your potential ride site, you can import it to google earth to get a good idea of the topography. Then just choose to export the gps file, use mapsource and import it to the 60CSx.
    Another thing you can do is find forums for the area you are interested in and post the "where to ride" question. At the very least somebody will give you a trail name you can search for on MB or give you a web site that has info.
    One other possibility is Google the trail names or park names. You will be surprised just how much info is out there.
    It may not sound like it now but half the fun is the treasure hunt that leads to the discovery of a new area for you.

    **EDIT***

    On MotionBased, the second tab at the top is "TrailNetwork" That is the tab you want to research your trails. Below that are the Search, Quick Filter, and Advanced Search links. Click on the Quick Filter link and follow the directions above.
    Last edited by bankerboy; 05-18-2009 at 08:22 PM.
    Apathy will get you exactly what you deserve

  7. #7
    MTC rep for LPCPCAC
    Reputation: bankerboy's Avatar
    Join Date
    Oct 2006
    Posts
    1,064

    I also forgot....

    Geoladers

    http://www.geoladders.com/overview.php

    The trick to finding some of the "off the grid" rides is to use the dropdown menu at the top of the page. Scroll all the way to the end of the ride list and select "view personal routes".

    You now have a lot to sort though. See you next year
    Apathy will get you exactly what you deserve

  8. #8
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    1,140
    Quote Originally Posted by IvanLasston
    I believe this conversion rate is what caused the mars orbiter to become the mars impactor.

    http://www.cnn.com/TECH/space/9909/30/mars.metric.02/
    I thought those guys were spending way too much time riding the trails at deer creek on those extra long lunch breaks.
    http://ludb.clui.org/ex/i/CO3165/
    http://www.dirtworld.com/trails/trail.asp?id=1062

  9. #9
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    11,794
    Yep, the two big reasons are sampling rate and the inability of the GPS to take elevation changes into account. Some software will do the elevation change calculations after the fact, but the resolution on most elevation data is not good enough to really pin it down. Usually, 20m is standard, but some of the new, fancy satellites can get 10m elevation. That data's not cheap, though.

    Add to the fact you're talking about race speed. The faster you go, the more tight corners will get 'cut off' due to sampling rate. When I'm gpsing a trail for the first time, I walk it because slower walking speeds + fast sampling rate usually equate to more accurate data. I tell you, volunteer crews HATE this part because most of them just want to strap the GPS to the bike and ride. But it really does make a better end product, especially if you take routes of the same trail from different people on different days, you can get a bit of an 'average' between them and hand-sketch the trail on the map.

  10. #10
    Two Tired
    Reputation: CheapWhine's Avatar
    Join Date
    Dec 2005
    Posts
    368
    Altitude change and sampling frequency error have been mentioned as causes of error, but I am not so sure. Personally, I am not too concerned about getting "true" distance accurately to three decimal places since I ride for fun, but I suppose some people are interested in this.

    I think people will agree that altitude measurement based on either satellite signals or barometric pressure is 100% accurate. What is the real impact of this inaccuracy on distance? If you do the math, it is pretty small. If you consider a 6% grade over the length of your entire ride (this is in the reasonable range), there seems to be a 0.18% difference between horizontal distance and the distance considering the grade.

    Sampling every second may improve accuracy over sampling every 4-8 seconds in the "smart" sample mode, but I doubt that really makes the difference. I noticed that my smart sampling seems to vary from 1 to about 10 seconds, which seems reasonable for different trails, speeds or other factors.

    I think the error is associated with the limited precision of the satellite signals. You are able to calculate your position within X feet of your true position. Any turns within this range are likely to be missed or filtered out by the GPS unit. Turns include all sorts of "normal riding" like weaving on the trail while you climb, turns around small obstacles, narrow twists on the trail and just following the best trail line. I think all of this error accumulates and shows up as the discrepancy reported.

    By the way, it is good to have you (NateHawk) posting again!
    Let the good times roll.
    trailroller.com

  11. #11
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    11,794
    Quote Originally Posted by CheapWhine
    Altitude change and sampling frequency error have been mentioned as causes of error, but I am not so sure. Personally, I am not too concerned about getting "true" distance accurately to three decimal places since I ride for fun, but I suppose some people are interested in this.
    The differences appear to be on the order of miles. I have no quibbles if my computer says 7.2 and my buddy's says 7.4.

    I think people will agree that altitude measurement based on either satellite signals or barometric pressure is 100% accurate. What is the real impact of this inaccuracy on distance? If you do the math, it is pretty small. If you consider a 6% grade over the length of your entire ride (this is in the reasonable range), there seems to be a 0.18% difference between horizontal distance and the distance considering the grade.
    Yes, the inaccuracy inherent in altitude measurement with your handheld GPS is mostly minor, but it can be huge, depending on conditions.

    Sampling every second may improve accuracy over sampling every 4-8 seconds in the "smart" sample mode, but I doubt that really makes the difference. I noticed that my smart sampling seems to vary from 1 to about 10 seconds, which seems reasonable for different trails, speeds or other factors.
    Sometimes, sampling once every second is not fast enough. I know I've been on a few fast, twisty trails where not all of the twists appeared on my final map. The twistier the trail, the more prominent this error, and the straighter the trail, the less prominent this error will be.

    I think the error is associated with the limited precision of the satellite signals. You are able to calculate your position within X feet of your true position. Any turns within this range are likely to be missed or filtered out by the GPS unit. Turns include all sorts of "normal riding" like weaving on the trail while you climb, turns around small obstacles, narrow twists on the trail and just following the best trail line. I think all of this error accumulates and shows up as the discrepancy reported.
    Yes, this is yet another source of error on GPS signals. This one is pretty obvious, though, and I generally don't bring it up unless I have to. But some of that does come back to sampling frequency and altitude measurements. We know from basic trig that the hypotenuse of the triangle is longer than the base, but if you look at them, they both cover the same horizontal distance. If you take the height measurement out of the equation, you end up with the shorter base length. ALL consumer GPS receivers (not sure about professional surveyor grade models) leave out the height measurement. They just don't factor it in, so the reported distance will always be less than the actual. This has nothing to do with inaccuracies inherent in altitude measurement.

    By the way, it is good to have you (NateHawk) posting again!
    Thanks. It's good to be back in the land of the living.

  12. #12
    Two Tired
    Reputation: CheapWhine's Avatar
    Join Date
    Dec 2005
    Posts
    368
    Interesting points, but I still disagree. You mentioned simple geometry, so let's explore this for a moment. If you consider a 6% grade, this translates to a vertical incline/decline of about 3.4 degrees. This is a pretty small angle, which explains the 0.18% difference between the horizontal and the "hypotenuse" in the vertical direction. This is definitely a discrepancy, but it does not seem to be a major one.

    Now think about the triangle in the horizontal plane. A 3.4 degree deviation from your "straight" course would still result in a 0.18% difference in track length. Now think about tire marks on a wet trail or it the dust. They wander all over the place, especially going up hill. The deviation is a lot more than 3.4 degrees. A 45 degree turn (perhaps a tight switch back or zigzagging up a hill) would result in a 41% increase in track length. This looks like a much more likely source of error.

    I don't know, but I suspect, that the GPS units internally do some smoothing to reduce jaggies that would exist without it. Without this smoothing, every track would be jagged due to inherent (and intentional) positional error in the GPS signal. But applying the smoothing eliminates the effect of the "horizontal hypotenuse". If you think about this happening the entire length of the trail, this could result in a pretty significant accumulated error.
    Let the good times roll.
    trailroller.com

  13. #13
    Scott in Tucson
    Reputation: Krein's Avatar
    Join Date
    Jul 2004
    Posts
    1,313
    I agree with one thing here, and that is that it's good to see NateHawk back here posting! Welcome back, dude.

    Now, I agree with CheapWhine, the distance lost due to not measuring the hypotenuse is very small. So small, that other sources of GPS error (positional and sampling) far outweigh it.

    I've thought about adding 'terrain distance' to TopoFusion, and had it in there for a while. I loaded up my most ridiculous rides (trust me, I have a few that made me bleed out of my eyes), you know, ones full of climbing, steep hike-a-bikes, etc. I was wholly unimpressed by the difference between linear and terrain distance. I never even got it to add a mile, even on massive (100 mile+) rides.

    I think sampling frequency is a major culprit for GPS measuring low mileage, depending on how twisty the trail is.
    Author of TopoFusion GPS Software. MTB+backpacking = bikepacking.net. Ride Diary.

  14. #14
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    11,794
    Quote Originally Posted by CheapWhine
    Interesting points, but I still disagree. You mentioned simple geometry, so let's explore this for a moment. If you consider a 6% grade, this translates to a vertical incline/decline of about 3.4 degrees. This is a pretty small angle, which explains the 0.18% difference between the horizontal and the "hypotenuse" in the vertical direction. This is definitely a discrepancy, but it does not seem to be a major one.

    Now think about the triangle in the horizontal plane. A 3.4 degree deviation from your "straight" course would still result in a 0.18% difference in track length. Now think about tire marks on a wet trail or it the dust. They wander all over the place, especially going up hill. The deviation is a lot more than 3.4 degrees. A 45 degree turn (perhaps a tight switch back or zigzagging up a hill) would result in a 41% increase in track length. This looks like a much more likely source of error.

    I don't know, but I suspect, that the GPS units internally do some smoothing to reduce jaggies that would exist without it. Without this smoothing, every track would be jagged due to inherent (and intentional) positional error in the GPS signal. But applying the smoothing eliminates the effect of the "horizontal hypotenuse". If you think about this happening the entire length of the trail, this could result in a pretty significant accumulated error.
    Note I didn't point to one source of error. All of those sources of error occur every time you use the GPS. How much each source contributes depends on the trail. In the midwest, the discrepancy between the horizontal distance and the dist+alt will be negligible for the most part. But in big mtns where flat trail is the luxury, not the norm, such a source of error becomes more significant. I dunno bout you, but doing trailwork over the years, I've measured many stretches of trail well over 10% grade...some over 30%. Now granted that doesn't always translate to an overall high average for a 10+ mile ride, but some trails do have an inordinate amount of excessively steep trail. It happens.

    CheapWhine mentions weaving on the trail. That's a possible source of error, but it still comes to sampling frequency. If the GPS took positions every 0.5 or -.25 second, it might pick that up better. GPSes do often do smoothing, but it depends on how you use the GPS. I try to avoid the smoothing features, as it's more likely to cut off switchbacks and such that way.

    Maybe you're on a trail and you're using a GPS to record it. Those error sources are always there, no matter what. On a climb, maybe the terrain increases in contribution while the others are small (S facing slope, very steep, no canopy, good sat geometry). On the downhill (N slope, poorer reception, more canopy), maybe the error is greater due to higher speed/sampling frequency and a spottier sat reception. Then on the next climb, on an E facing slope (still good reception, more canopy than S slope, but the time of day has poorer sat geometry, and it's gentle), the terrain error is not a significant factor, but reception is a big deal. They all play a role, but at different times and in different amounts. You can't discount any of them.

    Even one some might call insignificant is still a notable source of error and can even be measured for any given point on the trail (doing so for the whole ride would be an exercise in tedium).

  15. #15
    Scott in Tucson
    Reputation: Krein's Avatar
    Join Date
    Jul 2004
    Posts
    1,313
    Agreed that you can't discount any of them. I just don't feel that looking at the terrain distance gives you that much better an estimate of the total distance covered. I ride big mountains and have never noticed a significant difference when I add it in (whether with GPS or DEM elevations).

    It all contributes, for sure, and it is one error source that can be more easily quantified, so for people obsessed with distance it is worth worrying about. I have a few users who are this way, for sure (mostly trail runners).
    Author of TopoFusion GPS Software. MTB+backpacking = bikepacking.net. Ride Diary.

  16. #16
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    1,140
    I love for one of you bright geo engineers to come up with an algorithm for mtb ease of distance or something like that to factor into calculations.
    10 miles of narly singletrac 12% grade vs ten miles of smooth 12% trail…... as we all know how useless distance and speed often are anyway except for comparable stats for frequency on the same route. Btw, Topofusion is awesome !
    Anyone got any good comments on the original thread : Edge vs Vista HCx v 60CSx.

  17. #17
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    11,794
    Quote Originally Posted by ghawk
    I love for one of you bright geo engineers to come up with an algorithm for mtb ease of distance or something like that to factor into calculations.
    10 miles of narly singletrac 12% grade vs ten miles of smooth 12% trail…... as we all know how useless distance and speed often are anyway except for comparable stats for frequency on the same route. Btw, Topofusion is awesome !
    Anyone got any good comments on the original thread : Edge vs Vista HCx v 60CSx.
    I'm no math guy...haha. The algorithm stuff would be beyond me. But my inclination is that creating one to more accurately estimate the data based on trail type would be more challenging than it's worth.

    As for choosing the gps, it all depends what it'll be used for. If you actually want a nav tool, more than just a bike training computer, then throw out the edge. Between the vista and 60, you have to decide what button layout you prefer.

  18. #18
    mtbr member
    Reputation:
    Join Date
    Oct 2007
    Posts
    147
    I just don't trust GPSs as odometers. It is akin to measuring for drapes using lasers from outer space. I have a regular cheapo computer on my bike that I use for the odometer functions, and the Edge for all of the other stuff, like tracing routes, recording the ride and climb, heart rate, cadence, bla bla. If you really want an accurate distance measure use a regular computer.

    Also, each recorded point is what, +/- 10 ft in the xy plane? That error is going to add up over a course no matter which unit you are using, especially one with a lot of switchbacks.

    Keep in mind that I'm new to all of this GPS hocus pocus. Just wanted to offer my 2 cents.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •