GPS Need Not Apply.............Oh Alright

[NATURAL ANKLING]

Hi,
In the early years I some how managed to get drafted into orienteering, maps and stuff, little did I know it was grooming for torture by the elements.
Ten Tors on dartmoor, you must be mad but just do it several times, fail the 55 miler by walking 53 of the 55 then being crashed out, failure on time
In the training we did Map Mile's was always expressed, as the crow fly's.
So on taking up another torture Marathons, seemed a good idea at the time All my training mostly alone was measured acurately as I could on a map.
You dont really want to put yourself though all this pain to find the distance you have run in training is short
So to cycling, late August this year made two attempts at riding for 24 hours non stop, another one of my good idea's
One week apart both were failures to stay awake after about 18 - 19 hours with no sleep I know now I need to take forced rests to have any chance.
I also like to keep my cycle computer accurate as possible, I have done this by pessimisticly setting a 700C 37 tyre 75 psi to two centermetres less than the circumferance for wheel input, if you measure the radius of tyre with wieght on bike this I.M.O. will give a better result.
I do not have access to GPS etc, so I am left with physicaly mearuring on a map. To my surprise on a 40 mile ride my computer is only 0.1 of a mile long
Then I decided to use Bikemap.net, and plot my course, I was desperate to get an acurate reading as I lost some of my recolection of where I had cycled after the ride, easily done I.M.O.
My computer (basic 15 year old design) read 295.9 miles and on plotting with Bikemap.net manually it came to 294.6 miles wow
Originally I could only make it 265 miles old way, so I was to put it mildly, considering I simply could not remember where I had been on the looped course, which I varied. guted and disapointed. Now I am ok
Q1 Who non GPS comp users set their wheel imputs on the low side To allow for actual tyre radius.
Q2 Who nows if Bikemap.net or anyother systems alow for actual distance up hill and all (corrected)
Q3 Who nows if GPS run systems apps whatever are gradient corrected
T.I.A. Cheers,

[Mick F]

I can answer regarding my Garmin 705.

It has a rear wheel magnet that measures speed and distance and the system is automatically calibrated using GPS fixes. That way, the wheel circumference setting is continually modified as tyre pressure variations and deformation occur during a ride. Also road surface variations are taken into consideration. By using this system, the GPS signal can vary too and everything is as self-checking as possible. If the GPS signal disappeared temporarily like under trees or in a tunnel, the system could still cope.

Distance and speed displayed - and recorded - are as near as dammit correct.

[Ayesha]

One wise old sage once said to me, "Two towns are not going to get closer together because you have a more accurate milometer".

Magnet bike computers measure the distance by wheel revolutions.
GPS units measure a series of straight lines between log points.
BikeHike et al accumulate calculated distances between spot points from an enormous database; and are between contour lines and junctions.
Map measuring wheels measure distance as if the world was flat.

None are accurate. None are absolute.

[Ayesha]

I can answer regarding my Garmin 705.

It has a rear wheel magnet that measures speed and distance and the system is automatically calibrated using GPS fixes. That way, the wheel circumference setting is continually modified as tyre pressure variations and deformation occur during a ride. Also road surface variations are taken into consideration. By using this system, the GPS signal can vary too and everything is as self-checking as possible. If the GPS signal disappeared temporarily like under trees or in a tunnel, the system could still cope.

Distance and speed displayed - and recorded - are as near as dammit correct.

Now HOW do you know the "correct" distance?

[Mick F]

Oh good grief!
Here we go again.

Where are accurate distances marked out?
Dunno, but if you go on a dual carriageway with your bike, and count the 100mtr posts, you can see if your computer is correct.

I had a long thread last autumn on this very subject. I went up to the A30 on three occasions to carry out different experiments.

[Ayesha]

The vast majority of tour and long distance cyclists research the distance of a ride to decide how much to eat before and during the ride.
Using a figure of 50 kCals per mile to maintain bodyweight, a 1% inaccuracy in 100 miles will not be a ‘fatal’ error.

The little blue pointers on the AA road atlas are good enough for me, and I’m maintaining my rotund with ease.


To give an example.
When someone in the pub asks “How far have you been today?”, I reply “We went to Evesham.” He then gets an idea of how flippant I am about 0.1 of a km on my bike computer.

[mattsccm]

As I have to input some measurement into the computer i feel that it might as well be accurate (within certain limits) Thus I measure the distance of one rotation with a paint blob on the tyre and I sit on the bike to give the compression as in riding. I reckon thats near enough as I only round mileage to the nearest tenth anyway.
Getting perfect mileage won't happen. A recent ride had 4 GPS units give a 1% differnence over 90+ miles.
I am happy with that.
To get more accurate you would need to have to tow a very small trundle wheel type measurer.

[NATURAL ANKLING]

Hi,
Thanks for your answers, more non GPS answers including is there any correction FOR going up and down hills on GPS and do the WEB based sites as they so often qoute vertical distance Use any type of correction for hills
You all know how to use your GPS but do you know how the system works
GPS by default can only go on the lowest resolution it has, so yes draws in what appears to be straight lines on the web picture, this might be the resolution of the graphics or the input from the GPS itself
At one time satalite cameras had a resoulution of one sq metre, that is it could only give one colour / shade for every one square metre on the ground
Despite that films and peoples perception was that you could see into peoples garden, and what they do.
I have no doupt that its better today, but it was still recently that mapping the british isles was done with a camera on plane and not a satalite camera
I am not nit picking on 0.1 of mile etc which I am more than happy with, but the correction to hills.
Does it exist on systems or is this just to much infomation for computers to handle at source
Paper maps were until recently still copys of old maps errors and all.
Paper maps include a error for the curve of earth and magnetic and true north, and is quoted on map.
Sorry to be a bit finicky alot of us still cant afford GPS run computers.
Map miles as the crow flies, wheel driven computers distance as you actually travel on the tarmac depending on the input.
On say dartmoor the hills would have more than what most would say as acceptable an error on differences between different systems.
Map miles is map miles at the end of the day, thanks again, just ona wonder..................
Edited: To put it in anut shell do GPS run systems measure flat miles and do they have an option for gradients as the gps system on flat miles would always be different from old wheel input systems with no external input

[meic]

Here is a good one we had earlier.

[iviehoff]

What level of precision do you input your wheel dimensions to your computer? Mine has numbers of the order of 200 in increments of 1, (wheel circumference to nearest cm) so any measurement by that computer was subject to an error range of 0.5% simply because of the coarseness of the calibration. So for such a computer, you cannot say whether you are 0.1 km long over 40 km, because there is an inherent uncertainty of 0.2 km over such a distance.

Maps are imperfect representations of reality, and such is the width of roads shown on a 1:50,000 maps you could easily be 100m away from where a point is in reality. Google maps are much finer, but uncertainty over the exact line a bicycle takes on the road, around roundabouts, hill slopes, etc, I guess are likely to be of the same order as the 0.5% just mentioned. Then there is the uncertainty of the algorithm used by the computer.

I would say that for these reasons actually measuring a route with a bicycle computer was likely to be more accurate than measuring it off a map, even google maps, provided you have calibrated your computer well. If you want the most accurate measurements from your bike computer, it would appear appropriate to measure an exact km with a tape measure or the like from close to your front door, and use it regularly to monitor the calibration on your computer - remember that 0.5% of a km is just 5m, so a computer that indicates to the metre would be best for this, otherwise you'll have to note exactly how far beyond the start and end ticks your computer ticks over to the next 10m, if your computer only gives you 10m increments.

Someone suggests in that other thread that bike computers over-measure distances because bicycles weave left and right - but that ought to be corrected for in callibrating against a measured mile.

They also talk about GPS systems measuring your location at discrete intervals. A potential problem with this, not mentioned, is that if you are measuring positions at intervals of about 2 seconds, then a bicycle will only move a short distance in 2 seconds, usually less than 20m. But the accuracy of GPS measurement is at least a metre. So you are adding up a lot of small numbers each of which has an error of about 5-10% in it, and I think that is potentially a problem. The other thread suggests undermeasurement by approximating each short segment by a straight line. But if you have an error of a metre or two, some of that error will be left/right off the actual line of travel, and you get an entirely spurious weaving from left to right, because of the line uncertainty in the GPS measurement, far larger than a bicycle's own weaving. It is also possible that each of these locations is only recorded to the nearest metre, which again would be a large proportionate error. So it's far from obvious to me that the GPS is more accurate than the bicycle computer. It maybe that the GPS understands this issue and corrects for it.

(btw - set your computer to km, you know it makes sense.)

[meic]

What level of precision do you input your wheel dimensions to your computer? Mine has numbers of the order of 200 in increments of 1,

I agree entirely with the point that you are making here and I only buy speedometers which are calibrated to the mm rather than the cm. Though I have not been able to precisely calibrate it right to the millimetre, I have got past the centimetre level.
I dont think I can get to .1% but I can better .5% by repeated correction over a regular set journey.

Unfortunately I can get quite precise but I dont have a standard so that I can make my mile the same as a standard mile.

[Mark1978]

My GPS is slightly different from my cycle computer, the GPS is usually less than the computer. Probably about 0.05 miles per mile, but then some times the error doesn't compound so it must be in agreement, most of them time I just accept there is a slight error on both sides.

[meic]

My GPS reads lower than my speedometer but it is of the order of just under 1%.

I think that at 5% your speedometer is over reading as the discrepancies between the different methods are not that great.
Unless you ride in a lot of shade and tunnels.

[Mick F]

It maybe that the GPS understands this issue and corrects for it.

Like I said above:

I can answer regarding my Garmin 705.

It has a rear wheel magnet that measures speed and distance and the system is automatically calibrated using GPS fixes. That way, the wheel circumference setting is continually modified as tyre pressure variations and deformation occur during a ride. Also road surface variations are taken into consideration. By using this system, the GPS signal can vary too and everything is as self-checking as possible. If the GPS signal disappeared temporarily like under trees or in a tunnel, the system could still cope.

Distance and speed displayed - and recorded - are as near as dammit correct.

Some folk disable the auto function and use only the wheel magnet. 705's can input a wheel circumference down to the millimetre, but you need to know the figure! Mine is set to Auto, and if I were to look at the recorded wheel circumference from time to time, I would see that it varies.

Why?
Because different road surfaces compress the tyre more, powering up hills deflects the tyre whilst freewheeling down hills doesn't. Tyre pressures vary too - temperature varies it of course; tyres are cold when you set out, and warm at the end of a ride.

With the unit set to Manual, and you know a set distance, it's possible to select a wheel circumference to make the readout correct.

When I supported nine LEJOGers in May 2007, none of them had GPS of any description, just Cateyes and Sigmas etc. Hardly a day went by without a heated discussion as to the distance they'd covered despite all nine having done exactly the same! By the time they'd finished at JOG, there were errors of tens of miles.

[iviehoff]

I just had a quick discussion with a positioning expert, and he averred that because of the positioning uncertainties, which are actually even worse than I suggested, a distance travelled measured for a bicycle in motion by GPS would be pretty poor unless - as many of them actually do - they correlate the position measured to a map. So for such a GPS, the distance measured by GPS essentially the same as the distance for a specified route on a map as assessed by a computerised mapping program.

Those people who say that they go further as measured by bike than by GPS, (assuming the bike computer is properly calibrated) I suspect this is because their GPS is putting them on a map and giving them an "idealised" route as represented on a map, whereas the computer is measuring a true path along the surface of the earth that the bicycle actually moves, including going around potholes, mini-roundabouts and parked cars.