I currently have a Garmin 755T and I am very happy with it. However, I would like to use a netbook and Garmin Mobile PC, but so far, I am a bit dissatisfied with the accuracy I am getting on Mobile PC vs the accuracy of the Nuvi. With the Nuvi, the tracking is smooth, and as I see street signs at 30 - 40 MPH, the Nuvi shows accurately that I am approaching and passing each street. Mobile PC is not as accurate as I can pass a street and it takes a few seconds for Mobile PC to show that the street has been passed. Also, the Nuvi tracks my speed near perfect, but Mobile PC, lags a few seconds behind. When I stop, the Nuvi speedometer goes to zero quickly. Mobile PC on the other hand rarely makes it to zero before I am moving again. I am currently using an Ambicom BT-GPS, which has a Sirf III chipset. After researching, it seems that I need a 5HZ GPS, like the I.trek M7 for more fluid movement, but I wonder if someone could tell me if it will track more like the Nuvi does or if you just can't get that kind of tracking on a PC/GPS unit. If I was traveling and had not made a predetermined route, but I was looking for a certain street, and lets just say that there were no street signs available to me, I would probably be ok with the NUVI as the tracking for street ahead is almost perfect. I would be in trouble with Mobile PC as I would already pass the street before Mobile PC would show that I actually passed it. So, I am looking for a solution with Garmin Mobile PC that would be near equivalent to the Nuvi. Will a 5HZ GPS accomplish this? I am not really open to other mapping programs as I like Garmins routing the best, but I would like to go with the netbook as a mobile wifi and Garmin mapping solution. Thanks!

You didn't say what computer you are using to do the testing with? How fast is the CPU and how much RAM?

I've heard these problems described with at least a couple of other programs but this is the first time I've heard it with Mobile PC.

...ken...

It is unlikely an issue caused by Garmin Mobile PC software, or an issue of your SiRFstarIII GPS receiver.

5Hz is nice to have, and it does indeed make the map movement more fluid, but even with 1Hz unit your Garmin Mobile PC should work just fine.

I have used Garmin Mobile PC a lot on an ASUS UMPC with a crummy Celeron 900 chip 1.2G memory and multiple things running in the background to support the device. I had problems but map updating was not one of them. It is easily the match for my Nuvi 750 in this respect.

Could it be due to where you are placing your GPS receiver? GMPC will show the satellites being used if you click on the satellite symbol in the top right hand corner of the screen and choose 'GPS Info'. BTW your 755 will do the same thing if you touch the icon in the top left for about 3 seconds.

Terry

Okay, the original post prompted me to go out and spend a big chunk of the afternoon with my laptop and a variety of GPS receivers in the car. My arsenal included the following:

Mobile PC v5.00.50
iNav iGuidance2009
Mappoint 2009

Pharos 500 USB
Globalsat BU-353 USB
i.Trek M7 Bluetooth
Garmin GPS10x Bluetooth
Garmin eTrex Legend HCx handheld connected by USB

(I could not persuade my Navation 168 USB to work, no way. Tried three different cables and couldn't get data from it.)

I don't have a Nuvi so to do a similar visual comparison as the original poster and get my own reactions to the visual differences I used my Garmin iQue 3600A which is very smooth to view and very positive on speed changes.

In a nutshell, all three PC nav programs do what the original poster described.

The apparent jerkiness of the moving map is a function of the difference in screen size. On the small screen of a handheld or other personal navigation device the map doesn't move very far even when you are travelling at high speeds, because the screen is so small. However, on the large laptop screen the movement is very apparent and looks quite jumpy by comparison. The faster you go the jumpier it looks because when the screen updates with the new position each second, you have moved a fair piece down the road.

To do a fair comparison with the handheld you would need to hold it about six inches in front of your nose, and then you would see the same relative jump with each position change.

Regarding the speed display keeping up, part of that is simply the lack of synchronization between the handheld and the laptop. That is, the receivers will be sending new position information as much as nearly a second apart. It also takes your eyes a good portion of a second to track from one screen to the other.

I can see what the original poster was talking about because with most of the receivers I tested I saw the same lag when speeding up and slowing down, with the laptop not showing the speed increases or decreases as fast as my iQue.

Here's the thing, it did it with Mobile PC, Mappoint and iGuidance. All lagged my handheld a small but noticable amount.

I did discover two very interesting things. Both demonstrated that it's all in the receivers.

First, with my eTrex Legend connected, I was able to set the eTrex close enough to the laptop that I could take in both displays in a single glance. The speeds tracked nearly perfectly between them. With both of them displaying speeds to the tenth of a kilometer (xx.x kph), there was never a variation of significance. That means the positional information was being received, processed and displayed by Mobile PC on the laptop pretty much as fast as it was being processed and displayed in the eTrex. In both cases there was still a small apparent lag behind my iQue 3600A both increasing and decreasing speed.

Second, with my Garmin GPS10x Bluetooth receiver connected, there was no apparent lag between the speed displayed by Mobile PC, iGuidance or Mappoint versus my iQue 3600A when increasing and decreasing speed.

To put it into perspective, my iQue 3600A also has some lag in the speed display. When I have stopped completely, it takes between one and two seconds for the display to settle to Zero. And it takes one to two seconds of movement for the speed to start climbing when I start up again. This is to be expected when you understand how a navigation program or personal navigation device works.

In contrast, the laptop programs all had an apparent lag of perhaps one additional second when connected to everything except the GPS10x receiver.

Part of that apparent differential could be overcome by killing the tenth display. iGuidance displays in full MPH or KPH (no tenths displayed). This has the side effect of making it appear both more responsive and more stable.

Bottom line is that if you want the same sort of apparent responsiveness of the Nuvi, it's all in the receiver so spend the bucks on a Garmin GPS10x Bluetooth receiver to use with Mobile PC. It's big bucks.

And it really does not buy you anything in navigational usability.

That's really the bottom line. All of the receivers worked well. None had anything over any of the others to justify any significant price differences.

All of the navigation programs have the same display jump when they update the screen for the new location every second and all look jumpy if you compare them side-by-side with the display of a small personal nav device.

The only thing I didn't test was cranking up the refresh rate on the i.Trek M7 to 5 Hz. I don't have the necessary software installed on my laptop. I am sure that it would have supported Marvin's and Terry's experience that it smooths out the display because you can't move far enough in a fifth of a second at any sane and rational speed to cause big jumps on the screen.

Truly, the simplest and cheapest solution to the original poster's problem is:

Option A: turn off the Nuvi when you're using the laptop, or
Option B: use the Nuvi and forget the laptop for nav.

If you only use one or the other you won't notice those little differences.

Two more things I want to mention.

I really love Mobile PC's ability to discover any type of GPS .. Bluetooth, true USB, and even very high numbered fake COM ports from a USB-Serial driver. To get iGuidance and Mappoint to work I had to use GPSGate. That way they always saw a COM port and it was always a nice low number.

iGuidance has the nicest navigation display of the three. By a wee bit over Mobile PC and ... well, Mappoint/Streets&Trips just sucks in that department, especially if you don't have a route active. In my judgement.

...ken...

Very interesting tests Ken and I can see how a larger screen would cause larger jumps between samples. However, this should not affect how soon GMPC identifies the next street as this is not a function of how far apart they are on the display. If that were the case, increasing the scale would make it work better as the streets would appear further apart.

Lately, I have had to use the speedometer as mine is being flakey. I note that both the Nuvi and GMPC update at about the same rate. I am not concerned with when they go to zero as I can usually tell when I stop.The disadvantage of the Nuvi is it doesn't show the speed in navigation mode; GMPC does.

I agree that in landscape mode, iNav presents a slightly better picture of the map as it wastes less space on menus and borders. In portrait mode this is not an issue. Now if iNav was just a little more usable and had the same amount of POIs for my area, I might give it more time.

Terry

The issue is not with the netbook (a Dell Mini 9, with 1 GB of memory) as I have also tested it with a 2.2 GHZ core 2 duo notebook with 3 GB of memory. The placement of the bluetooth GPS receiver was actually more closer to the front of the car than the Nuvi (it was sitting on the dash as close as the windshield would allow and had a better "view" of the sky). The Nuvi was on a windshield mount about a foot behind the bluetooth unit, but nevertheless, I do not believe that the signal is the issue here as both units were receiving very similar signals.

My issue is not because the screen is larger and just shows the 1 second updates clearer. I feel fairly confident that the NUVI is not only given the GPS data on a more frequent update than one second, but also that is uses its proprietary software to do the things that it does so well. Although I don't recall testing this, I feel confident that if I pull up to a road at 25 MPH and stop, the car icon on Mobile PC will not stop as soon as I stop, but actually catch up to the accurate placement of where I am and then stop. The Nuvi does this immediately when I stop so the placement is just about perfect as soon as I stop. The speedometer (which the NUVI 755T has on the main map screen) is far more accurate than the one on Mobile PC's main map screen because again, it does not seem to play catch up like Mobile PC does.

If you were testing these units on the interstate at a constant speed, both the NUVI and Mobile PC would track speed fine. However, if you increases speed 2 MPH for a second or two and then back down 2 MPH, the NUVI would change almost identically to the car speedometer. Mobile PC would lag and would not update the changes until a few seconds after they occured. I have tried a different USB/GPS unit with Mobile PC and it's just as bad. Even so, I feel that the issue is with the receiver as I have yet to try the Garmin receiver or a 5HZ receiver. I would be very curious if someone could test with a Nuvi 7XX series unit against Mobile PC and a Garmin GPS10X and see if they react the same as Ken in Regina implied but was not able to test against a Nuvi.

I would also be curious if the Garmin GPS10X is a 5HZ receiver or what are the exact modifications to that particular unit that would make it work so much better with Mobile PC? Maybe because of a Garmin specific protocol vs NMEA?

Please do not misunderstand me and think that I am trashing Mobile PC. Mobile Pc would okay for predetermined routes as the warnings are enough for you to understand when to turn, etc. However, I do not like the fact that the data from the GPS lags on the screen, unlike the NUVI 755T. I would prefer to use GMPC over the Nuvi, but I would also like GMPC to update on the screen so that when I pass Main ST in my car, GMPC shows that I just passed Main ST. The way it is now, I can pass Main St in my car and it takes GMPC a good 3-5 seconds to show that I have passed Main St. Maybe this is the norm for most units, but the Nuvi does not do this. When you pass Main St, the Nuvi shows that you are passing Main St. Thanks for your time.

P.S. I don't know why my paragraphs are all running together after I posted this, but they keep getting joined together when I save changes.....sorry

edit: I am trying to add the paragraphs again, this time with javascript on and using the "Go Advance" button, "Preview Post", then posting.....it worked on my post below.....so we shall see.

All my previous on-road usage of GMPC was with a Garmin 10x. I didn't realize that would make a difference until Ken mentioned it and will try some real world testing with some of the other units I have.

Yes, it was very difficult to read your post with no paragraph breaks.

Terry

I'll bet you're cutting and pasting into the forum from a word processor on your PC. The carriage control characters (CR, LF -- equivalent of the Enter key) are not translating, or aren't even in the source text. If that's the case, try a different editor (perhaps a true text editor like Notepad) or use the forum's online editor. Or compose offline, cut/paste, do a preview and fix the paragraph breaks with the online editor.

My previous post was rather long so I'll repeat that the lags I noticed in speed changes and vehicle icon location on the map relative to intersections were not specific to Mobile PC. They were consistent to all three nav programs that I used. That's why I said I'm confident it's a function of the receivers.

The slight lag in position of the vehicle icon relative to intersecting streets that I noticed in my testing was as minor as the speedometer display lag. It was insignificant to any safe kind of navigational use.

With my iQue 3600, when my car is fully in the intersection, the cross street almost exactly intersects the centre of the vehicle icon. With all of the PC nav programs, with all of the receivers, the nose of the vehicle icon was generally just touching the intersection as I was fully in the intersection. The lag was never more than a vehicle length.

The Garmin GPS10x performed precisely the same on all three PC nav programs so Garmin clearly has not done any tweaking specific to using it with Mobile PC.

Finally, an observation....

The only time the visual lag on the laptop screen will affect your ability to navigate is if you are trying to use it for visual navigation rather than audio guided navigation.

In my opinion, that's simply not something you should be attempting to do with either your Nuvi mounted in line of sight and you surely shouldn't be trying to do it with a laptop that cannot even be placed in line of sight in most cases.

And if you insist on trying to use these devices for visual navigation, a lag of a carlength, or even a couple, isn't an issue. In the first place, it's consistent so you know exactly how much to you need to "lead". In the second, the faster you go, the bigger buffer it gives you to make an upcoming turn.

...ken...

I know my memory is failing and I used the Nuvi just yesterday but isn't the following true. In non-navigation mode, both Nuvi and GMPC (and iQue/ XT) show the upcoming street in a banner above the screen. I assume that is what we are discussing.

Terry

I am not coping and pasting. I am typing everything in the quick reply section below. I do have javascript on now, so maybe that will help - we shall see.

No, I am not testing the "upcoming street" notification in non-navigation mode, which I think Garmin was no longer allowed to use on the Nuvi I have due to copyright problems (if memory serves correctly). I am testing a problem with Garmin Mobile PC in that it seems to lag in both speedometer readings as well as the map display. Once you are moving at a constant rate of speed, I do not notice the speedometer reading accuracy problem so much, but I still notice that GMPC does not (with my hardware and test) track oncoming intersecting roads accuratly. It takes a few seconds for the GMPC map display to show that I am passing a intersecting road, when in fact my car has already passed it. The Nuvi 755T (and 750 that I've tried) do not exhibit this issue so blatantly.

Ken in Regina - either you and I are testing differently, or misunderstanding each other in how we are testing, or our results are simply much different. My results with GMPC do not leave the car icon touching an intersecting road when I pass it in my car. I have not tried stopping at the intersecting road to see if it "catches" up (which I think it will) because that was not my issue, but when passing intersecting roads, GMPC is a few seconds behind the Nuvi in updating the car icon on the map screen.
Since you do not have a Nuvi to test with, I can conclude that you are not aware of the differences (if there actually are or are not) between the tracking performance of GMPC and the Nuvi 755T. You certainly have a right to your opinion about what I should or shouldn't do and what I should or should not expect with my GPS system. Even so, I am still interested in the answer to my question from others who may be able to provide it. I don't think you will be able to help me answer it since you do not have a Nuvi 7XX to compare. If you can get access to a Nuvi 7XX that you can compare with, I would be greatful for the test results.

Thanks to all for your time and help. Again, I am just wondering what differences in tracking that there might be between GMPC and the NUVI 7XX series and what it would take to make GMPC perform like the NUVI 7XX series. I can certainly keep the NUVI and use it for my particular application, but I would prefer GMPC.

After reading through the post again, I just noticed that my Mobile PC version is 5.00.60 and Ken in Regina's is 5.00.50 - I wonder if that could be the difference between our test results?

If I am completely misunderstanding your problem, please forgive me, please ignore the following and try to help me understand what the problem really is.

When I drive through an intersection, my iQue 3600A shows the vehicle icon smack in the centre of the intersection exactly as my vehicle is smack in the centre of the intersection.

Said another way, as I drive through an intersection, the line of the intersecting street/road almost exactly bisects the centre of the vehicle icon on the display as the actual intersecting street/road is off my left and right shoulders.

No stops involved here. Just observing as I drive through an intersection to try and determine any lag in the display of the vehicle position. There is no apparent lag in the vehicle icon display on my iQue 3600.

Can you describe, with a similar example, how your Nuvi is more accurate than that, in terms of where the vehicle icon is displayed on the map versus where your vehicle really is, please?

When I use this same observation with the PC nav programs as I drive through an intersection, I see that all have a lag in the display. The front tip of the vehicle icon appears to be just entering the intersection, or just slightly before entering, at the point where my car is passing through the centre of the intersection.

I use that description because the only place I can tell, visually, how much the vehicle icon display lags my actual vehicle position is at intersections. Using this method of observation it appears that the vehicle icon in the PC nav programs lags my real position on the road by roughly a vehicle length.

These observations were done at city street speeds, typically 50kph speed zones. Faster speeds would, of course, show a larger visual lag of the vehicle icon.

Are you seeing a bigger lag than that with Mobile PC? You have not given us an example that allows us to quantify just how big a lag you see with it. I have tried to do that as best I can. I would be interested in a comparison using the same metric I did above ... as you drive through an intersection, where is the vehicle icon being displayed on your Nuvi and on your laptop versus where your vehicle really is.

As far as the lag in the speedo display, there is a guarantee of 1 second in any device that has a GPS reading refresh rate of once per second (1 Hz). I'm seeing almost exactly that amount of lag on my iQue 3600A. If you are seeing less than a 1 second lag on your Nuvi, that can only mean that it's refreshing the readings more frequently.

I have no idea why some of the receivers I tested -- all except the GPS10x -- produce a little bit longer lag on the speedo display. For some reason there seems to be about an additional second lag. It's close enough that it's surely no more than that. Trying to guage it visually is difficult and the difference is small for any sort of manual measurement, like a stopwatch.

I went into this testing thinking that the explanation of any lag would be the latency (delays) of having the GPS receiver external to the device ... having to wait for the position data to go through the cable and the driver software before the PC nav programs even see it. I was persuaded otherwise by the results from my eTrex Legend (virtually identical speeds on the built in display of the eTrex and the PC nav programs, suggesting no significant lag due to the connection length) and the GPS10x (no lag versus my iQue).

As I said at the start, if I'm completely misunderstanding the problem please forgive me. If it's simply a difference in the amount of the lag you are seeing versus the amount of lag I'm seeing, perhaps some specific examples that would help us understand the magnitude of the lag would allow us to help with some testing and perhaps troubleshooting if that seems warranted.

...ken...

I doubt it. None of the listed changes to .60g are related.

It would be helpful if you could give us an idea of what your test results are with some sort of examples similar to what I've described so we can visualize what's happening and compare. For instance, when you drive through an intersection at city street speeds, where is the vehicle icon being displayed relative to where your car really is as it pass through the dead centre of the intersection?

If it sounds like our test results are significantly different I'll be happy to update my laptop to Mobile PC 5.00.60 and retest with a couple of receivers.

...ken...

Thanks Ken... No need to apologize to me as everyone has a different way of thinking and communicating. I believe my Nuvi responds how you are describing your iQue 3600 responds - There seems to be no obvious lag on the visual display of the units car icon movement. In other words...the car icon on the map mirros very closely the actual movement of the physical car while going through an intersection. I am happy with that performance!

The lag on GMPC is much more obivous to me than what you are describing. If two intersections are close enough together, I believe I could pass the first street and get to the second street before GMPC would show that I was at the first street. I will take GMPC out tomorrow and test at 50 MPH and count the seconds it takes for the car icon to reach a street intersection that I have already passed.

Thanks for your help!

Just for the record, I'm in Canada. My testing was at 50 KPH. That translates to 30 MPH when I get south of the 49th. Sometimes slower in traffic. I would guess that my speeds through intersections on surface streets wouldn't be much more than 20 - 25 MPH.

If I remember correctly, at highway speeds of 50 - 65 MPH there was about 1/2 to almost one full icon's length of space between the vehicle icon and a crossroad just as I was passing it.

...ken...

I did a test today at between 25 and 30 MPH. At that speed, when I cross an intersecting road, it takes around 2-3 seconds for GMPC to show that the car icon is crossing the same intersection. The car icon movements on the map are jerky so that most of the times you can't really see it cross the intersection. It's more like you see it approaching the intersection, then with the next update it is just passed the intersection. Even with the lag in updates, it does not normally pass intersections smoothly.

I've also noticed that sometimes (when traveling much slower), the car icon will move forward and then backwards, even though I am continually moving forwards. When moving at around 25 MPH and coming to a stop at an intersection, I was correct in that GMPC continues to update the car icon on the screen until it catches up to my physical location at the intersection. This takes between 2-3 seconds.

My previous test were at between 35 - 40 MPH, so the lag with GMPC was even greater as I would come up to intersections.

The speedometer lag when traveling from 0 to 25 MPH is about the same with approximately a 2 second lag.

My Nuvi is much more fluid than this and that is what I prefer for my application. However, I find it difficult to believe that it is impossible for GMPC to reproduce what the NUVI can do, with of course the appropriate hardware and any needed tweaks to the software.

Thanks!

I use Garmin Mobile PC daily, and honestly I don't know exactly what the lag is. I just don't pay any attention to it. It has never been an issue. I would expect a software like this, while connected to a 1 Hz GPS receiver, to lag 1 or 2 seconds. If you connect it to a 5Hz unit, the map movement will be considerably smoother, and the lag will be shorter.

I use a 1Hz unit, as most people do. But I did test GMPC with a 5Hz GPS receiver, and if you don't like what you see now, I trust you would like the 5 Hz experience.

5Hz GPS receivers are not that expensive anymore. You can likely buy one for around $50 to $60.

Hey magnumpi,

That's really terrible. It's clearly way worse than I'm seeing so I can understand why you are so irritated by the comparison to your Nuvi.

Mine is like Marvin's, not nearly enough to be any sort of an issue at all.

Based on my testing I would be willing to bet that simply trying a better GPS receiver would make a difference. Marvin and I, and tcassidy, all use pretty much the same receivers and we all see pretty much the same thing. At least I assume Terry is because he hasn't been whining about it. None of us has anything nearly like what you are seeing.

The back and forth deal (position seems to move backwards at some points) is why I think it's your receiver. It sounds like it's having trouble getting good positional data.

My testing supports that the PC nav program will simply display the positional data it is given.

In particular, when I connected my eTrex Legend to use as the receiver it allowed me to see the internal display of the eTrex at the same time as I could see what Mobile PC was displaying (that would be just like if you could connect your Nuvi as the input to Mobile PC). Speeds and icon movement were in virtual lockstep. So you can be pretty comfortable that whatever Mobile PC is showing you is just what the receiver is sending it, not something Mobile PC is making up on its own.

As Marvin said, the i.Trek M7 is a pretty fine receiver. It's not expensive. And it can connect with Bluetooth or USB, your choice. Even at 1Hz I find it good enough that I have no urge to install the management software to change the update frequency to 5Hz. ...... Perhaps I should try it out. Maybe I'll like it so much I'll not want to use any other receiver. Ya think???

...ken...

Ken, you should try to switch the unit to 5Hz. The difference is huge. Yet, I'm happy with 1Hz .

I do recall reading on another forum of a couple of people experiencing a lag of 10, even 20(!), seconds with BU-353 + iGuidance. BU-353 is a popular and reliable USB GPS receiver, which I use, too, but I have never experienced this issue. In both cases, if I recall correctly, the issue was elliminated by resetting the GPS receiver to factory settings. I'm not sure if we are talking about the same problem in this thread though. I doubt it.

Marvin, now that we talk about it I would like to try the M7 at 5Hz. Can you email me the GPSInfo program? I looked all over the Globalsat site and I can't find a download for it. I thought it would be included with the Vista drivers but it wasn't. That package just installs the Vista drivers but no GPSInfo.exe.

...ken...

I think Marvin is referring to changing the refresh rate on the iTrek M7 which is not a GlobalSat product. I will see if I can find the program (I know I've seen it before).

Terry

Ken,

I think this is the link you want.

http://www.itrekgps.com/Bluetooth%20GPS/DAC023AF-5D53-4399-BDD3-47A6C4354BAD.html

Terry

DOH!! I can't believe I did that ... looked high and low on the Globalsat site for it. I guess it's no surprise I didn't find it, eh??

Thanks for catching that, and for the link.

...ken...

Thanks everyone for the participation and insightful information.

Ken, I think I will await your test of the I.Trek at 5HZ to determine if that is better suited with GMPC or if the Garmin GPS10X is better.

Now that Terry has so kindly got me pointed in the right direction and I've got the configuration utility for the M7 I'll see if I can give 'er a go tomorrow. Sometimes a slap in the head is more useful than a pocketful of GPSs to reset directional stability......

I've updated the laptop to 5.00.60g and I was going to ask you which receivers you wanted me to use but I guess you've already answered that. I'll use the GPS10x and the M7.

...ken...

Thank you Ken..... While we await your test, what is everyone's feeling on the I-Blue 737 verses the i.Trek M7. I've read in another forum, that there was some positives and negatives for both, but I was wondering which would suite my application better should the I.trek M7 test better than the Garmin GPS10x. Here is the forum I am referring to:

http://www.gpspassion.com/forumsen/topic.asp?TOPIC_ID=105784&whichpage=6

I believe it's the third post by Doug99 that offers the comparative test results.

Thanks for that link. Your answer about the "lag" we've been talking about, in both the speed display and the vehicle icon position, is explained very nicely in Doug99's comparison posting:

He explains, further down, that this data filtering time constant also affects the kind of "jumpiness" that many people see on their nav displays when they are sitting still. That is, with a short filtering constant, the i737 will show no movement in the vehicle icon (or track file) when sitting still, whereas receivers like the M7 will show quite a bit of movement even when the vehicle is sitting still.

When doing static testing in the house with VisualGPS, most of my receivers will show my house moving as fast as 2 or 3 kph at times, and I've seen it travel as far as the far side of my neighbour's lot!

Good find. Thanks. It's something I understood but I've been scratching my head trying to figure out how to articulate it without writing another novel. Doug99 did an excellent job of a short and concise explanation.

...ken...

The signicance of knots is important too as vehicles and houses rarely wander around when it is visually obvious they have stopped.

Terry

Since I would not only like accuracy as I pass an intersection at say 30 - 35 MPH, but also when I am approaching an intersection and slowing down to say 5 - 10 MPH, it seems that the i.Trek M7 would perform better than the I-Blue I737 for my needs; or am I misunderstanding the test results?

Though we don't have the test results back yet comparing the Garmin GPS10X to the i.Trek running 5 HZ, I have a feeling that the i.Trek will perform better for my desired results. With that said, it looks like Semons is out of stock on new M7's and only have an open box version. I called and was told that the open box units are returns which have been tested and work perfectly. My only concerns with purchasing an open box would be the battery and possibility that somone tried a hardware hack on it, failed with it, and then returned the unit. Otherwise, I am not sure why someone would be returning an M7 with all of the positive things I have heard about it so far.

I won't be making any purchase though until the comparitive test results come back because with this purchase, I want to buy the GPS that will work best for my needs.

Thanks!

Okay, the test is done. The results are in.

And the winner is..........

? ? ? ? ? ? ?

DamnedIfIKnow........


Readers Digest Version:

There are differences but I don't think they are significant.


Long, Windy, Detailed Version:

If you care about lint-picking, hair-splitting details, read on. If you just want to get a functional laptop navigation system going, decide whether you want Bluetooth or USB, check out a few reviews and then buy one with a positive review that you can get at a decent price. Or buy one of the nav programs that has a receiver bundled with it.

The Test Setup:

I have an Acer Aspire 15" laptop with 1.7GHz dual-core CPU and 1GB of RAM.

Nav software:

Garmin Mobile PC v5.00.60g set to 2D mode and the default vehicle icon.

GPS Receivers:

Garmin GPS10x - Bluetooth, 1Hz data frequency
Globalsat BU-353 - USB, 1Hz data frequency
i.Trek M7 combo Bluetooth/USB - tested in Bluetooth mode at 5Hz data

Now here's the thing ... I used the default vehicle position icon in Mobile PC. It's a big blue wedge-shaped thing, emphasis on big.

...... No. I mean B-I-G!!

With the zoom level set at 200m (650ft) the vehicle icon is about a block and a half long.

That makes the testing about like doing brain surgery with a chainsaw. Or as my engineer friend is fond of saying: Like picking fly poop out of pepper with boxing gloves on.

When you are driving a vehicle that's about 20 ft long, how the heck do you decide which part of that block and a half long icon is supposed to represent the vehicle? The very front portion? The center section? How much of it .. half of it? A tenth of it?

You see my dilemma?

Example: When sitting stopped at an intersection I could be three cars back from the intersection and the tip of the icon would still be touching the intersection.

All of that said, there are visible differences.

Speed lag:

All of the receivers have some lag. It can't work any other way. The receiver first has to detect that there's movement before it can send the data to the laptop nav program. So there's going to be some lag.

The GPS10x is the tightest. Consistently less than two seconds to start showing movement or drop to zero after stopping.

The BU-353 was next with more variability. It ranged from 2 to 5 seconds, mostly 2 to 3 seconds.

The i.Trek M7 was third. It ranged from 3 to 5 seconds and was quite variable in that range.

I have to caution that this was not stopwatch measurement. It was the closest I could get with "1-steamboats, 2-steamboats".

Vehicle Icon Position Lag:

Remember the size of that giant icon and take the rest of this with a few grains of salt.

When stopped at an intersection, all three receivers would provide positional data that had the tip of the icon sticking into the intersection a little, even with three cars between me and the intersection.

I wasn't able to see any significant difference in the position of the icon in relation to the intersecting street at surface street speeds of 50kph (30mph) versus local freeway speeds of 80kph (50mph).

As I passed through the intersections on surface streets or passed under the overpasses of cross streets on the freeway:

- the M7 would have perhaps 10% of the icon into the intersection,
- the GPS10x would have perhaps 15% of the icon into the intersection,
- the BU-353 would have perhaps 20% of the icon into the intersection.

At the highest speeds - 100 to 120kph - the tightest position had the very tip of the icon just poking into the line of the overpass of a cross street and the absolute worst-case lag had the tip of the icon just visibly short of the overpass.

If I make the unsubstantiated assumption that the center of the icon should represent the centre of my vehicle (thus putting the front of the icon a good three or four car lengths in front of my vehicle) that means they all show a "lag" of two or three car lengths, with about a car length difference among them.

Map Redraw Smoothness:

All of the receivers have their quirks.

The GPS10x and BU-353 both show 1 second jumps, just as you would expect. Those jumps are bigger at higher speeds, just as you would expect.

The M7 was sending data five times more often than the other two so you would expect a smoother redraw because the map is only moving a fifth as far on the screen. And that is the case. ... Sort of...

The M7 is a victim of its own sensitivity. With the increased data frequency it has a lot of "jiggle". In addition to smaller, but more frequent, redraws to show the forward motion, there is also a lot of sideways movement as the vehicle icon jiggles from side to side on the road.

The result is that the big jumps of the 1Hz receivers struck me as providing a more "stable" view than the jittery feel I got from the M7 at 5Hz.

Of course this is all subjective.

In both the speed "lag" and the vehicle icon position "lag" I found absolutely nothing that would bother me in using any of these receivers.

When I'm accelerating or stopping, I'm so involved in the task of operating the vehicle that I truly don't care what the speedometer is doing. I literally never look at it at those times. I use the speedometer to give me feedback when I'm moving at a relatively constant speed to make sure I'm travelling at a safe or legal speed. I have no interest in my rate of increase or decrease in speed, so I don't care whether there is lag or not.

When used as a speedometer is used by the vast majority of us, there is nothing to choose among these receivers. They are all spot on when comparing them to the known error in my vehicle's speedo.

The differences in the vehicle position icon between the three receivers is also nothing that concerns me. Even if I was trying to use the laptop for visual navigation, there is too little difference among the receivers to recommend or condemn any of them. That vehicle position icon is just so damn big! that the worst-case lag from any of the receivers was never more than 50% of the length of the icon itself.

When you combine that with the fact that the actual accuracy of any consumer-grade GPS receiver is never going to be much better than a car length or so at best, this simply isn't an issue.

The map movement was something that is notable. I definitely found a personal preference for the big, but stable, jumps of the 1Hz receiver. As soon as I get done writing this history novel I'm going to change the data update frequency to 1Hz on the i.Trek M7. Whenever I looked at the screen with the M7 connected, I was uncomfortable with the busy, jittery look.

This brings me right back to my original contention that half or more of the issue of perceived map redraw smoothness is with the big screen on a laptop when compared to a small handheld navigation device. I have used the M7 many times with Mobile XT on my Palm T|X. Mobile XT is virtually identical to Mobile PC. The M7 has always been in 5Hz mode. I have never noticed this jittery display characteristic on the much smaller screen of the Palm.

So, that's my report. I'm not sure it resolves or clarifies anything. It is what it is. I hope most people will find it comforting that unless you stick your head in the sand and buy some really cheap no-name piece of junk, you'll have a device that will serve the purpose admirably for normal vehicle navigation.

In my opinion.

...ken...

After my testing I can promise you one thing, categorically. You will never get the sort of visual responsiveness from Mobile PC or any other PC nav software that you do from the Nuvi or any other personal nav device, regardless of which receiver you choose. Your choice of receiver will have a small impact but it won't turn that big laptop into a Nuvi.

Jonas's explanation is correct. I'll explain later. Right now my carpal tunnel is telling me that that last post was enough for today.

EDIT: Ooops, I see you've deleted the post I'm replying to. Oh well, I'll leave this for now. If you think it's irrelevant please let me know and I can delete it, too.

...ken...

Ken, big thanks for taking the time to test all this, and for sharing the results with us ! I also have been under the impression for a long time now that differences between GPS receivers performance are insignificant for my own purpose, which is GPS road navigation.

P.S. But I wonder if you would find the 5Hz receiver performance more visually pleasing while in 3D (your test was done in 2D ).

Thank you Ken for taking the time to do that extensive test.

I still want to disagree that the screen size is causing the difference I see in updates and accuracy on the NUVI vs GMPC, but without being able to see the Nuvi on a larger screen, I can't be certain.

What I can do is tighten up the resolution on my netbook and also change the window size on GMPC as small as it will go in order to get a size closer to that of the Nuvi. When doing this, I can get GMPC to be about 25% - 50% larger (estimation) than the Nuvi screen. I'll try tracking with it this size and see if there is any difference. Keep in mind, the screen on my Dell is considered an 8.9", so the smaller the screen, the closer we can get to a Nuvi comparison.

I use 3D mode on GMPC and my car icon seems smaller than what you describe. I would describe my car icon on GMPC as being equally porpotional in size with the Nuvi's (when using the screen size as the porpotional difference); this is from memory and little use in GMPC, so don't quote me on that. However, the car icon's porpotion to the map is dependant on the zoom level (as you noted) and I think that's the difference between mine and yours.

As I said earlier, I have an Ambicom BT-GPS (revision 2)

When I purchased it, I thought that it would be a decent Sirf III GPS device. With the results I receive with GMPC verses what you receive with several different GPS units, I can only assume that either the Ambicom receiver is the difference or we conduct and conlude our test differently.

My original contention was that GMPC was not equal in performance to that of a Nuvi. Of course, I now accept that one would need to at least use a GPS receiver that was equal to or greater than the GPS10X for this expectation.

My gut is now telling me that with the Nuvi, Garmin has complete control over the software and hardware, and I am afraid that if I test GMPC with a better receiver, I am still not going to get the same performance I have with the Nuvi.

For most people, Ken is right, the lag doesn't matter. For my application, I want the performance that the Nuvi gives. If I can't get the performance of the Nuvi on the PC, then I will continue using the Nuvi. It seems to me though that if Garmin developed a hardware/software solution (like they have) for the PC, that they could achieve the performance of a Nuvi if that was the goal.

Now I need to decide whether to buy a better GPS in order to try and at least achieve the better results you seem to get Ken; or should I conclude that the results I am receiving are already as good as yours and I simply view them differently than you do. I think I might be happy enough with the results you get Ken...

Ken, I am a slow typer, so I posted before your test came in and then deleted it after reading your post as it seemed somewhat irrevelant at that point. This time I was smart and reloaded the page in another window, so I can actually answer your follow-up. The earlier post I deleted referred to a post in a different forum regardin how the Garmin GPS may use Doppler for speed indication, which I thought was interesting:

http://forum.brighthand.com/showthread.php?t=239562

I have carpal tunnel too, but I think mine lags more than yours.... :-)

Thanks again!

First and most important point, our testing is definitely apples and oranges.

You cannot visually assess the positional accuracy in 3D mode. The perspective is wrong to begin with and is constantly changing. You cannot effectively gauge, with your eyes, the absolute position of the icon in relation to any fixed point on the screen, like a cross street, with any reliability or repeatability. That's why I tested in 2D.

No comparison of our individual results to this point is possible. You need to do a run in 2D mode for my information to be meaningful to you. Then you can directly compare what I described in my previous post to what you see on your screen.

Second, and equally important, I repeat that you will never duplicate the apparent visual responsiveness of the Nuvi on a laptop. Not even on a small screen netbook, although it will be closer. Not with any of the five receivers I've tested.

Let me remind you of the test I did with my eTrex Legend HCX - one of the most accurate receivers on the market today. When I connected it to my laptop as a receiver for Mobile PC, I could tell from the exact synchronicity of the speedo displays on both machines that the PC was getting the data from the eTrex in as close to realtime as makes no difference. But the difference in the smoothness of the display movement was no comparison. The map movement on the little screen of the eTrex looked as smooth as silk. You already know what the map movement looks like on a laptop screen with any receiver that is sending data once per second.

You also read what I said about the tradeoff in display movement with the 5Hz data frequency of the M7. The jittery-ness of the M7 at high frequency that shows up on the laptop screen does not show up in Mobile XT on the screen of my Palm, which is the same size as a standard Nuvi display. The display on the Palm is smooth.

Third, your definition of performance, so far, seems to be focused almost exlusively on the apparent visual performance that you see on the laptop screen versus the Nuvi. I can assure you that the meaningful performance - the actual positional accuracy - of any of the receivers I've tested with the PC nav programs so far is the equal of the Nuvi. That includes the Pharos 500 that came bundled with my Streets&Trips 2008 (which I paid less than half for the whole hardware/software bundle than what I paid for the GPS10x).

I've used both of the Bluetooth receivers with a golf program (Intelligolf) on my Palm on the golf course where my yardage on a golf shot matters very much to me. Both are very effective. They give me yardage as accurately as I can make a shot. That is, they will always allow me to select the correct club.

I appreciate that the visual performance is the part that's most in your face, especially when in a moving vehicle. I just want it on the record for those following this testing that the issues we are discussing are no reflection on the positional accuracy of any of these receivers. Rather, it's the visible side effects of the delay between receiving the satellite data and passing position information to the PC and how often, per second, they pass the data to the PC. My testing shows that this does vary from receiver to receiver and affects the visual display accordingly.

Finally, back to the first point. I sure would appreciate it, just for my own information if you could duplicate my testing with your receiver. I would be interested in knowing how much of the perceived difference is between 3D and 2D visualization and how much, if any, might be a function of a receiver that's inferior to the ones I tested. I'm not assuming yours is. I'm sufficiently anal that I really want to know and only an apples to apples test will tell us that.

So, with that in mind could you please do the following tests in 2D mode with the default vehicle icon (the big blue wedge) and with the zoom level as close to 200m (656ft) as you can get (you could just switch to metric mode for the test .. would save me having to go back out and test in statute)?

1. When stopped at an intersection, where is the tip of the icon in relation to the line for the cross street:
a) when you are the first car in line and
b) when you are the fourth car in line?

2. When driving at surface street speeds (20 - 30mph), when you are in the middle of the intersection .. your shoulders are about lined up with the centre line of the cross street .. where does the line of the cross street hit the icon in relation to the tip of it? Or if the icon hasn't made it to the intersection yet, how much space is there between the line of the cross street and the tip of the icon?

3. Same as #2 at 50mph.

4. Same as #2 at 60-65mph.

I found that for tests #3 and #4 it's best to do it on a local freeway that has lots of overpasses from cross streets.

For safety it's best if you can get a copilot to do the looking for you. And if all tests are done at low traffic times.

I'm looking forward to the results to see if there's actually any difference between your receiver and the ones I've tested.

...ken...

Ken, I appreciate your reply.

I will try to do the test as you have requested. I am not sure I can do it tomorrow as I have a full day of work, but I will do so as soon as possible. I'll have a driver for my test as I always do.

Are you still stating that the reason the visual display on the Nuvi (even in 3D mode) mirrors the physical location of my car more accurately than GMPC is because of the size of the screen - or are you also stating that the difference is between the hardware and software of the Nuvi being so closely integrated with each other, which causes it to perform "visually" better than a PC setup?

For my application, it's all about being accurate visually on the screen, which I would argue is more accurate period, but that's is my very uneducated opinion in this field.

I do not doubt that most any GPS receiver bundled with some software will do what is is supposed to do within the current limitiations that have been put on them, but I also feel that with the proprietary hardware and software integration that Garmin has used in the Nuvi (and I'm sure other series), that they can visually appear to be (and in my opinion will be) more accurate for most applications.

On a slightly different topic: Even if I do these new test and the positional 1 second update accuracy is more close to your results, I still don't understand why Garmin doesn't offer the same "Nuvi" experience for a PC. If the only issue is screen size and or the time it takes to get the GPS data into the PC, it could be easily resolved by coding the visual display to constantly position the car icon where it should be, taking into account the known current data, the data update interval and the time it takes to get the info into the PC. There are GPS units with dead reckoning features and I imagine you could use a similar if not the same technic to accomplish a more pleasing and accurate visual display in all modes. Clearly, if it turns out that the 2D option in GMPC is more accurate than the 3D option, then there is a difference between the Nuvi 3D mode and GMPC 3D mode because the 3D mode on the Nuvi mirrors my cars physical location fairly accurately. Okay, I am done ranting.... carry on! :-)

Thanks again!

Hey magnumpi,

I'm going to wait until I see the result of your 2D test before I answer your last post. At least a couple of the questions are speculation on your part without any actual data. To answer them would require me to speculate on your speculations (how many pinheads can dance on an angel). So let's wait until the hard apples-to-apples test data is in. At that point I promise I'll address them all. But I'll be able to address them in the context of comparable data.

One thing I have to respond to, though, is that I am not saying nor am I implying that 2D mode is more "accurate" than 3D mode on anything. Not on the Nuvi nor on Mobile PC.

What I said is that you cannot accurately do a visual assessment of the absolute position of the vehicle icon in relation to a fixed point on the screen, like a cross street, in 3D mode. You can't. The angles are wrong and they change almost constantly as you approach and leave those fixed points. That's not a reflection of accuracy between the two viewing modes. It's simply the fact that the views are so completely different from each other.

When viewing in 2D mode there are no angles involved. You are always looking straight "down" on the vehicle icon and the objects near it.

You'll see what I mean when you try it.

I also need to emphasize that this is not intended to be a recommendation that 2D mode is the "Right" viewing mode to use for anything other than this specific testing.

...ken...

Could you elaborate on what "your application" is, and how you intend to use Garmin Mobile PC? I personally just use it for driving, and it has never ever even occurred to me that there was any issue with accuracy. For driving purposes the accuracy is just perfect.

Do you actually drive while Garmin Mobile PC is running, or are you a passenger observing the laptop screen?

I finished earlier than I thought today, so here are the test results:

First, Marvin, for this particular application, I am a passenger - I agree with you in that the accuracy of GMPC is perfectly fine for a predetermined route, which is what 99.9% of people probably use it for. I was simply trying to determine if the Nuvi gave positional and speed data more accurately / more often than GMPC and with my test today, I have proven to myself that it does give positional data more often and more accurately; whether that information is true data from the GPS or some of it is assumed by the Nuvi software to be more visually pleasing, it works for me!

Ken, I did the test you requested and the results pretty much mirrored yours (no real obvious lag). I know, that contradicts what I just told Marvin, but here is why. I mentioned earlier that we may be testing and or viewing results differently and indeed we were. The problem I have with your test is not 2D mode, but the magnification level you used. I do not understand how one could gauge accuracy when using a 200 Meter (or 650 feet) zoom level on GMPC. At that zoom level, the car actually grows to be extremely big at about 650 feet long (estimation). The car icon of course doesn't increase in size, but because the maps viewing area is increased, the actual area of the map is compressed, the car becomes larger in porpotion to the amount zoomed out on the map by default. This zoom level is obviously not a good representation of your cars actual physical location and is proven in your test result; it is the very reason that the car icon is always sticking out in front of the road you are intersecting (even when you are behind two or three cars) and also the reason the car icon never really lags behind intersecting roads that you cross.

At the closest zoom level (30 meters), the car is roughly 30 meters long or almost 100 feet. That's still much larger than the typlical car on the road, but with Garmin's proportions on GMPC and the Nuvi, the closest zoom level will be the most accurate to your GPS physical location, which can be proven by taking your GPS to an intersecting road and stopping, then viewing the car icon at this zoom level. In GMPC, you will most likely see that the car icon comes right up to the edge of the road which your are intersecting (and stopped), but will lag behind intersections that you cross at higher rates of speed. If you move very slowly, say under 15 MPH, GMPC seems to track fine at this appropriate zoom level. The larger you increase your speed, the larger lag you will see as you cross intersections. Just make sure you are on the 30 meter zoom level (zoomed as tight as it will go).

I had multiple helpers today so the Nuvi 755T was tested at the same exact time, under 2D and the closest zoom level available (also 30 meters). The Nuvi is much more accurate at higher speeds (the car icon almost mirrors the cars physical location) up to 50-55 MPH. However, at above 55 MPH or so, even the Nuvi starts to lag a bit, but still no where near as bad as GMPC.

I also tested GMPC with a much smaller window and it still clearly shows the 1 second update lags....jump....jump....jump. The Nuvi on the other hand, even when held 6 inches from our eyes (three people viewed it) was much smoother with its updates. If someone told me that the Nuvi was running at 5HZ, I would believe it because the updates seem at least to be 5 times as much as GMPC.

I don't know why my speedometer test showed lags earlier because today, the Ambicom BT-GPS / GMPC was on par and even better on some starts than the Nuvi.

Lastly, GMPC and the NUVI have an autozoom feature. I haven't really tested it on GMPC, but it works great on the Nuvi. This feature zooms out at higher speeds and zooms in at lower speeds, which gives a better visual simulation of your cars physical location.

I still might be convinced to try a 5 HZ GPS and I believe the visual updates would be more pleasing in GMPC, but I doubt it would stop what I will call, "the intersection lag" of GMPC at higher rates of speed.

Ken, if you have any free time, try a test at the closest zoom level on GMPC and I bet you will now see the lag I have been talking about.

Thanks again for everyone's time - I don't take it lightly that you have done it to help me answer my question!

Thanks. That's what I needed to know. Apples to apples. Everything else will fall from that.

No need. I already know it. Even without knowing or seeing it previously, it's completely predictable if you understand even a little bit about scaling on the computer screen (I'm a computer professional). I don't mean that to be as arrogant as it sounds. If you go back and read my description of the test results again you will see that I covered that possibility when I talked about the size of the icon relative to the car size and observed that the "lag" was probably one to four car lengths in reality, depending on speed. My apologies if I didn't articulate it as well as I might have.

My selection of 200m zoom level had no special thinking behind it. It's simply the level that I normally operate at around town. It was as good a place to do the test as any.

Your concern seemed to be whether there was an actual problem with either Mobile PC or your receiver. That was the basis for the testing I did, and asked you to do, so we both had results that could be compared properly. We have those results and they demonstrate three things:

1. There's nothing wrong with your Mobile PC installation.

2. There's nothing wrong with your receiver.

1 and 2 are all I wanted to establish. The rest is personal preference. There's nothing I can do about that. Either Mobile PC, when it is functioning properly with a receiver that is also functioning properly, meets your needs or it doesn't. Only you can make that decision.

Out of the testing you get a third bonus item.

3. All five of the receivers I tested with produce essentially the same results you are seeing with yours. They represent a range from one of the least expensive to one of the most expensive in the consumer-grade category. Now you have that information without having to spend a bunch of money.

There is an outside chance that you might find a receiver with 5Hz capability more pleasing to your eye. I did not find that to be the case for me but we all have different preferences. If you want, I can do a test with the M7 at 5Hz to verify what the visible intersection lag is at 30m zoom.

You're welcome. It's interesting questions like this, questions that need both testing and discussion, that keep this interesting and add to the knowledge base on the forum.

...ken...

Ken, I would be curious of the lag with the 5HZ receiver at the closest zoom level. No hurry, just curious.

Also, if anyone else has other answer as to why there is so much less lag on the Nuvi and why it seems to update more often, I would be interested in that information.

Thanks!

Not sure about the Nuvi, as I'm just a PC kind of a guy, but some GPS software uses predictive algorithm. It guesses where you will be in a couple of seconds, and it will display that info, instead of your actual GPS location. The advantage is that this may be more visually pleasing to some users, and the disadvantage is that it may show your location incorrectly for a couple of seconds at a point when you decide not to follow the plotted route.

I still think it's mostly a display issue (size and resolution). I can't imagine why Garmin would spend the time and expense, and ongoing software management issues, to do a specially optimized version of the nav software just for the Nuvi. Especially when most people simply don't care.

It's not likely to be a predictive algorithm because we're seeing this when doing visual navigation, not with a route active. Without an active route, there's no basis for prediction.

It might be faster location data refresh (more than once per second) but I don't think so.

As was explained in a link posted by magnumpi to a post from another site, all GPS receivers do some sanity checking on the position data before passing it to the nav application. That simply means that the GPS receiver is sampling the satellite data at a far higher rate than even the 5Hz (5 times per second) of some receivers. It analyzes that data and tries to pick out the true from the false (many of the signals it's receiving might be multipath reflections rather than the direct satellite signals).

There is a fixed period of time that the GPS receiver will analyze the data before deciding on the next location to send to the nav software. This is the "time constant" that was refered to in that link. You can think of it sort of like an "averaging" period (although it's way more complex and sophisticated than simple position averaging).

It's this delay from the time constant filtering for the receiver's best guess at the correct location that is causing the "lag" that we've been discussing.

It's called a time constant for a reason. It's programmed in and it will always exist. Whether you send the data from the receiver to the nav software once per second or five times per second, each of those position calculations went through the same time constant filter. If it's a two second time constant, that's the "lag" you'll see ... if you can see a lag at all.

And it will exist whether the receiver is external to a laptop or internal to a personal nav device like a Nuvi. It's all about doing the best to filter out spurious data. It's far more critical with today's more sensitive receivers because they are capable of operating almost entirely on multipath reflections without any clear view of the satellites at all. That's why my eTrex Legend and my two Bluetooth receivers can operate in my basement office.

You can predict all of this quite effectively based on one observation. Start off from a dead stop and time the delay until the speed display gives the first result. For a sanity check, come to a stop and time the delay until the speed display hits zero. Do this a few times to get a typical time that takes the extremes out of the picture.

You will likely find that this delay is in the two to four second range. That's the time constant delay of the receiver you are looking at. From there it's simple arithmetic.

At any speed, you will travel a specific distance each second. At 60mph it's about 88 feet per second. If the time constant delay observed above is two seconds, the "lag" we've been discussing will be 176 ft. In other words, at 60mph the vehicle icon on the screen will be trailing your actual position on the planet Earth by 176 ft.

The next part of the issue is how accurately your nav software is capabile of positioning the vehicle icon on the picture of the map and how accurately the vehicle icon represents your vehicle.

If you zoom in to the very tightest zoom level, where the size of the vehicle icon isn't more than twice the size of your actual vehicle, you will see a reasonably accurate representation of this. That zoom level is useful for a visual inspection of the display accuracy and the effects of the position delay. But a zoom level that tight is not going to be very useful at anything faster than 20 or 30mph even on a large laptop screen. It will be completely useless on small personal navigation device screens at higher speeds.

That's my opinion, anyway. For whatever it's worth. It's one of those questions that can be endlessly fascinating for those of us who are sufficiently anal-lytical. But it's the sort of thing that most people don't, and shouldn't, care anything about because it does not affect any normal uses of these devices.

Y'er welcome. Y'all have a fine weekend.

...ken...

Ken, I have to disagree again with you about the screen size being the differences between GMPC and the Nuvi, but I will agree to disagree. I believe that you've hinted at the fact that the hardware in the Nuvi might have something to do with the performance of it vs GMPC, but nevertheless......

I still think that the integration of the hardware and software by Garmin in the Nuvi is the key difference in the performance of each. I highly doubt that the software code in GMPC is verbatim of the Nuvi, not only because of the performance and feature differences, but simply because I believe Garmin ported their software from their proprietary language to a language that can be ran by a PC.

No one is certain what operating system Garmin is using in the Nuvi's or if there even is one as such, but we do know that it is proprietary to Garmin. Since their proprietary software and hardware do exactly what they chose for it to do, I am fairly certain than when porting to the PC, they had to make changes to accommodate the requirements set by the language they chose to port to. Even so, I still believe that if Garmin wanted to, they could make GMPC perform exactly like the Nuvi. They may not want to though, and here is a possibility of why....

Garmin realizes that any skilled programmer could decompile their software and see what makes it tick. Do any of Garmin's competition in the PC/GPS market do better with this lag issue? I'll bet if they did and it made a difference to the end user, that you would soon see the equal results in a GMPC update. I've heard that Iguidance has a smoother update than once per second. I bet that feature will be in GMPC if they find that it is needed to please the end user and keep up with the competition. I would think that they could protect anything in their code by patent and copyright, but I don't know. I believe Garmin has the best routing algorithm of any mapping software and that seems to be used in GMPC, so maybe this, "possibility of why" that I have proposed is pure nonsense.

My point of bringing up this discussion was to find out whether or not the Nuvi was more accurate than GMPC. The answer I found is that it is indeed. Ken says that the accuracy level is only visual and that may be the case, but I say that by testing both GMPC and the NUVI, we easily prove that it does not have the lag that GMPC does. As I said earlier, the Nuvi does have a slight lag above around 55 MPH or so and it probably grows with speeds higher than that, but nothing compared to the lag that GMPC has at the same speeds.

A GPS system is suppose to let you know where it is (within the limitations imposed). If the GPS software is programmed to place a car icon on the map by not only using the GPS data, but also by first processing a vast amount of variables/calculations, and the software is able to accomplish a visual result with more accuracy than just the positional data on its own, then I say it is more accurate. Even if the accuracy level is not based on the final absolute position given by the last GPS data alone.

As Ken has explained, GPS's these days use software to evaluate the data that they are receiving. Before the GPS data is passed on to the navigation software, the data is processed and sometimes the data that is given is a best "guess" of possibilities. If the GPS did not filter and calculate a best guess, the results would most likely not be the same and probably not be as accurate as they are. Ken, would you hold the position that this filtering would not necessarily make the GPS more accurate, but more visually pleasing? If you would, I would have to disagree.

In older version of GPS software, like Delorme Street Atlas, the GPS data would be shown on the map without any "snap to road" feature. Based on my physical location in the car, when the GPS data was off, the car icon would indicate that I was say, 15 feet off a road. The software could have been coded to correct this GPS data error, map error, or a combination of both and provided much more accurate information to me the end user.

Hypothetically, if the GPS data shows that I am 15 feet off of a road and traveling at a speed of 25 MPH, the software could assume that I am in a car (I am traveling at 25 MPH, so I am probably not on foot). First assumption is that I am in a car and it happens to be correct. Since the map data shows that there are no other parallel roads within 50 feet, the software could make a guess that I am traveling on the road which is only 15 feet off of the positional data. The second assumption is also correct. Is it more accurate based on the GPS data, no! Is it more accurately visually, yes! Is the snap to road feature more accurate for the end user most of the time, I would say yes! This is exactly what I am talking about with the Nuvi VS GMPC. If the results to the end user of the GPS software/hardware combined offer more accuracy based on an assumptions (I should add that the assumptions are based on data and calculated reasoning to back them up), then it is more accurate. The GPS product performance results is my measurement of accuracy, not the data coming from the GPS alone.

Can assumption based data sometimes be wrong, absolutely! In my hypothetical, I could have been using an off-road vehicle, traveling on an unmapped road. In this case, the snap to road feature should be able to be turned off for those occasions. With software like this, there are a vast amount of possible variables which could be analyzed and accounted for in order to make the accuracy level increase.

Ken, I agree that the closest zoom level is pretty useless for high speeds. That's why I use 3D mode and autozoom. With this feature, the zoom level changes according to my speed. The Nuvi does this very smooth and shows little to no lag at most speeds. GMPC does this as well, but the lag is clearly there at above 10-15 MPH and grows the faster you go.

I also agree with Ken and Marvin that GMPC will do fine for predetermined routes. However, there is no getting around the fact that because of the software differences, GMPC has a visual lag on screen where the Nuvi does not. This is not based on screen sizes, but the hardware/software differences between the Nuvi and GMPC. I did not start this discussion to prove that the Nuvi was better than GMPC. I did not mean to make anyone feel that they needed to defend it because you don't. I only wanted to determine if it was as accurate as the Nuvi, based on my set of accuracy standards. If anyone from Garmin wants to chime in on why it seems to be to me, I would love for you to share your proprietary secrets with us.

As are all of my post, just my opinions, based on my thinking with very limited filtering calculations and a lot of lag.

Thanks!

Just read through my post and corrected some things to make my points clearer.

I also want to add that the tone of the post should not be taken as an attack on Ken or his opinions as that is not my intention. I'm certain that Ken knows more about GPS stuff than I do. I just wanted to be clear that I am only expressing my opinions, which happen to oppose some of his. By the way Ken, if the 5 HZ GPS unit test with no lag in GMPC under the tightest zoomed level, I'll buy one, but I'll bet that it won't (a bet placed without the shaking of hands does not count). :-)

Not iGuidance 2009. That was covered in my first round of testing. I not only tested the different receivers with Mobile PC. As I described in the first testing report, I also tested with iGuidance 2009 and Mappoint 2009. All three nav programs had the same 1 second map redraw jumps corresponding to the 1Hz data transfers. On that first run the i.Trek M7 was still in 1Hz mode when running Bluetooth. I can't recall if iGuidance and Mappoint will even accept the 5Hz update rates.

The speedo display and the map display are based on the same position data being passed from the receiver to the nav software.

When you reported the results of your testing to duplicate mine you said that the lag in the speedo display when starting and stopping was very similar between the Nuvi and Mobile PC. I seem to recall that you even expressed some surprise at the discovery. Is that true or was it a typo and you meant that they were very different?

If the delay in the speedo display is similar in both, then the filtering lag, as I explained in my last post, is similar in both.

That means the delay in position information provided by the receiver to the nav software is similar in both.

That means that any visual differences in the map display are just that: visual.

There is only one way you can establish the actual physical real-world accuracy of the Nuvi and of your GPS receiver (Mobile PC has no accuracy because it cannot directly receive GPS satellite signals. It simply goes by the position data passed to it by any external receiver that is attached.) You have to find something with a known precise geographic location, place the receivers on it, let them sit there and average the position for awhile, then compare the latitude/longitude reading they give you to the known latitude/longitude position.

There are geodetic monuments all over North America that can be used for such purposes. (There is even a hobby activity where people try to find and log as many of these monuments and markers as they can, similar to the sport of geocaching.)

Any other test to declare whether one receiver is "more accurate" than another is not a test. It's mostly speculation based on, and extrapolated from, some other observation(s) that may or may not be relevant.

Our testing shows that the Nuvi and the receiver you are using both produce similar, but not identical, "accuracy" relative to each other but it says nothing about how accurate either one of them really is.

You can do another test to compare their relative positional accuracy. Just set them both in a spot outside with a good view of the sky, sit back and have a couple of drinks or read a chapter or two in a good book, then see what their readings are. It still won't tell you how accurate they are in any absolute sense but it will tell you if they differ much on where they think they are.

That's a much better test of their relative accuracy than anything we've done so far.

No. And I'm most discouraged that, after all the typing and testing I've done in this thread, you are still able to think I might. Clearly I have not done a good job. {...heavy sigh...}

I've never taken it that way. I have two simple goals.

1. Help determine whether your Mobile PC installation and GPS receiver are working normally.
2. Help us all understand the technology and its limitations a little better.

Understood. If I didn't believe it was a rational discussion I would simply cease to participate.

By the way, I am ignoring your dissertation about the value of an accurate representation of the vehicle location on the map display. I did read it. I don't disagree with you. Clearly it's important to you. For discussion purposes it's of little interest to me. For all of my purposes, all of my GPS devices provide me with better than sufficient information representation on their screens, large and small. I always welcome any functional improvements in the technology but the current state of the art satisfies me.

To put it into appropriate context, there are, frankly, too many big inaccuracies in the data provided by all manufacturers for me to get very excited about any minor inaccuracies in location representation on the map display. For instance, if the address database contains a location of a particular address that is off by a block or more, I just can't get exercised about a couple car lengths error in the display of the vehicle icon. Especially when there isn't a consumer-grade receiver on the market that will give you real positional accuracy any better than a car length.

I will report on the results of 5Hz data transfer at a tight zoom as soon as I get a chance to test it.

...ken...

Okay, I took Mobile PC for a ride with the i.Trek M7 at 5Hz data rate and zoomed Mobile PC to 30m.

In a word: "Gawdawful!!"

Of course the 30m zoom level just magnifies everything. So any jumps and jiggles are bigger jumps and jiggles. The 5Hz refresh rate just makes them happen faster.

If the jumps at 1Hz annoy you at 30m zoom, I promise you'll positively hate the action of a 5Hz positional refresh rate. You mentioned that you didn't like the reverse motion that you see occasionally with the 1Hz refresh at slower speeds. Well, at the 30m zoom you can see it even at speeds as high as 50mph, even with the refresh at 5Hz.

You raved about the Goodness of "snap to roads" in your previous note (an excellent example of intentional programmed compensation, by the way). Well, it isn't such a great thing at 5Hz refresh to have to watch the vehicle being constantly snatched sideways back onto the road when you're travelling down a road that isn't located perfectly. It's a feature I like and it's not a big issue at what I consider a more usable zoom level of 200m or 300m or more. But it's simply unusable (for me) with snap to roads on, 30m zoom and 5Hz refresh. Really really bad.... (kiddies, don't try this with a hangover!)

Okay, now that I've got that off my chest....

At 30m the correlation between the speedo display delay, starting and stopping, and the positional delay in the vehicle icon placement on the map screen is clear and undeniable.

With the M7, the average delay is 3-steamboats (make of that what you will). That is:

- the delay from a dead stop for the speedo to display its first reading is about 3 seconds.

- the delay when coming to a complete stop for the speedo to drop to zero is about 3 seconds.

- when going through an intersection at surface street speeds of 15 to 20mph it takes the vehicle icon about 3 seconds to "catch up".

- when travelling under an underpass at highway speeds of 50 to 65mph it takes the vehicle icon about 3 seconds to "catch up".

That's what I've been trying to explain. It's all in the filtering time constant delay. If it's three seconds, all motion in the nav program will always be three seconds "behind" reality. That applies to the speedo and the map display equally.

With the 5Hz refresh rates, the only difference is that there will be more "jumps" of the map repositioning to cover the three seconds distance.

Let me remind that in my original testing I saw the same positional difference when I was running the M7 at 1Hz. And I saw that positional difference on Mobile PC, Streets&Trips and iGuidance. I think in my previous report I said it was 3 to 5 seconds because it's more difficult to get an accurate number of "steamboats" with that giant vehicle icon at 200m than at 30m, as you pointed out.

To figure out what that is in terms of positional "inaccuracy" of the vehicle icon on the map display, all you need to do is the simple arithmetic to calculate what distance you'll travel in three seconds at whatever speed you're doing.

The zoom level makes no difference in the reality of this. It does, obviously, make a difference in the visibility of it.

Oh yeah, static positional accuracy is dead on. That is, when sitting stopped as the first car in line at the intersection, the vehcile icon tip is just poking the edge of the cross street.

So the only real choice between the i.Trek M7 and the Garmin GPS10x is approximately 1 second of filtering delay. As observed by me and my "steamboats", the 10x is about 2 and the M7 is about 3. It's not a difference that is relevant to me (it means Mobile PC will display the vehicle icon 2 seconds behind reality instead of 3), but I mention it for your decision-making.

I hope that gives you the information you needed. Now that the testing is done, the M7 is going straight back to a 1Hz refresh. The testing has shown me that 5Hz is not my cup of tea.

Be well.

...ken...

Yes but did you test 4, or 3, or 2 Hz?

Terry

What would you want me to test for at the in-between rates?

- Do you think there would be a difference in the filtering time constant delay at 4, 3, or 2 when there was no difference between 1 and 5Hz?

Since the delay is a function of the internal receiver processing and is the same at the extremes of the refresh rate range, I see no reason to assume it would be any different at one or more of the intermediate rates.

- Or do you think there might be some kind of improved viewing at one of the other refresh rates?

I'm currently operating on the basis that if it's okay at 1Hz and hugely annoying at 5Hz, it will be successively less annoying at intermediate rates down to only mildly annoying at 2Hz. Or maybe it will be as good at 2Hz as at 1Hz. But why would I care about that?

...ken...

Ken, thanks for the test results. I figured that GMPC would not perform any better wth a 5HZ GPS. The problem is not with the GPS. By the way, you were correct about Iguidance not being any smoother on GPS updates, the article I was thinking of was for ALK CoPilot Live Laptop 11 done by Marvin. As I said earlier, I will agree to disagree with you on the accuracy issue. My terms of accuracy and your terms of accuracy are much different - maybe it's just semantics, but I don't think we can come to an agreement.

I did run across some proof that seems to back up what I have been saying about the Nuvi 755T. According an article by Gpsmagazine.com:

Garmin nuvi 755T Review (GPSmagazine.com)

Garmin states that the Nuvi 755T update rate is 10HZ. It is more smooth and accurate than even past Nuvi's (something I was really not aware of before reading that). As I thought, it seems the combination of the software/hardware in this unit is what allows it to perform so much better than GMPC.

I tested a Garmin Nuvi 750 just to see how different it was and the update rate on it looks to be around once per second like GMPC. However, when at the closest zoom level and going around 45 - 50 MPH, it does not lag like GMPC does. It lags maybe a half car link when passing intersections at that speed, where as GMPC has a 2-3 second lag at the same speed.

I stand by this assertion - that for a GPS package to be considered accurate, the "tracking" icon must show your actual physical location (not only when stationary, but when moving as well). If it does this perfectly at all speeds, then it is indeed perfectly accurate, no matter how the accuracy was achieved. The less accurate it tracks visually, the less accurate it is. Whether this makes a difference to the end user or not is really up to each individual end user.

In my test, even the lag of GMPC does not seem to affect the ability of the software to assist the end user properly on a pre-determined route. However, the lag in the software is greater than a Nuvi 750 and Nuvi 755T, therefore, I consider it less accurate and that answers my question.

I will end my participation on this topic by thanking everyone again for their time and help to answer my questions.

Sure we can. We just need to be specific so others know what we are referring to. Can you say "Display accuracy"? And I'll say "Positional accuracy".

If that's true, it explains a smooth-looking display movement, especially on a small low-resolution screen.

GMPC has no lag. I already described the test that proved that -- the eTrex Legend connected so I could watch the speedo and map displays on the Legend and the laptop simultaneously. The displays were completely synced.

Mobile PC displays exactly, and only, what the GPS receiver connected to it sends it, exactly and only when it sends it.

We're getting close. If you could just say "The less accurate it tracks visually, the less visually accurate it is." we could have complete agreement. It's all about being sufficiently specific to avoid confusing others.

Absolutely. We agree totally on this. That's why it matters that we be specific when we are discussing and describing things. So users can make their own decision about whether that particular characteristic will likely matter to them or not.

I apologize to you and to everyone else for not letting this drop. I wanted to, I really did. I spent a lot of time trying to decide whether to answer. I really didn't want to keep rehashing this. But many people read these discussions, and will continue to do so long after you and I are satisfied and move on. It's important to make sure that things are reported accurately.

What convinced me that clarification is required is that I've already had two private messages in the last few days from people who have read this discussion expressing a concern that Mobile PC was inaccurate and would cause them problems by taking them somewhere they didn't want to go or by giving them seriously inaccurate positional information.

So, one last time to clarify for those who, up to this point, may have received the impression that Mobile PC is inaccurate, it's not. It displays what the GPS receiver sends it very accurately. If the receiver has a filtering algorithm in it that creates a delay in when it sends positional data to Mobile PC, Mobile PC will display that delay accurately. This is typically in the range of 2 to 3 seconds.

At a stop, there's no inaccuracy in the display.

At slow speeds the display of that delay is nearly undetectable.

At highway speeds the delay will be obvious if you are looking for it.

There is no speed at which this will cause Mobile PC to give bad routing instructions.

This applies to all laptop navigation software. I have personally observed this characteristic with Mobile PC, Streets&Trips, Mappoint, iGuidance, Street Atlas and MapFactor.

...ken...

If you want Garmin Mobile PC to refresh its screen ten times per second, then just buy a 10 Hz GPS receiver for your laptop.

That's true Ken, and thanks for taking the time to do the tests and write your replies. A person who is new to GPS, and who reads this discussion, could easily get the wrong impression there is something wrong with Garmin Mobile PC. No, GMPC is perfectly fine. I love using it on a large laptop screen. The view is really gorgeous in 3D as well as in 2D. On a large screen we may notice things we wouldn't have noticed on a small screen. Most of us use 1Hz GPS units with our laptops, but for pretty much the same price people can buy GPS units that refresh GPS information 5 or even 10 times per second. I used a 5Hz GPS receiver with GMPC for a while, and I liked the smoother map movement better. I suspect 10Hz would improve this even more. But these are all so minor, almost irrelevant, issues we are talking about, if the purpose of GMPC is to get us from A to B.

I'm actually back to using a plain 1Hz USB GPS unit now. The 5Hz was nice to show to my passengers, perhaps it impressed them more, but my primary purpose of the software is navigation, and for that purpose 1Hz is just fine - or it may actually be better, as Ken mentioned, since it is possible that after a while 5Hz could become annoying.

Ken, in your test with eTrex legend, what GMPC version was used? 5.00.50 original with GpsGate or without it, or the patched?

The first time was with 5.00.50. I can't recall if it was the patched version and I can't check now because I updated it to 5.00.60 for the last rounds of testing.

I have to use GPSGate with the eTrex Legend if I'm going to use it with Mobile PC. If I don't and Mobile PC sees the eTrex directly it refuses to use it. It can tell it's an eTrex Legend HCX but declares it's an invalid unit ID, says it can't validate the device and ignores it after that.

Based on beeps and a couple of messages on the eTrex, Mobile PC is sending something to it but it's not obvious what, but it appears it's not happy with the response (or lack of response) from the eTrex. It's obvious that it could use the eTrex directly if it wanted to because it starts to display the signals on the satellite info screen before it decides it's not going to use it.

It's strange because it will use my Garmin GPS10x directly. It will even show me the remaining battery charge of the 10x and let me set a couple of its settings, like WAAS and automatic power control. I have no idea why it will use one Garmin device directly and not the other.

...ken...

Okay, just out of curiousity I put the patched version of 5.00.50 back in place and tried it with the eTrex Legend HCx. It behaves exactly the same way as the unpatched versions. It refuses to use the Legend. It recognizes the device, complete with the details of model number and unit ID. It even starts to display the signal strength bars from it momentarily until it figures out that it's not whatever it's looking for. Then it stops displaying them and declares that it's an invalid unit.

I don't have a GPS20x to test with but I'm willing to bet that the 10x and 20x will work just fine with the software-only version of Mobile PC and it will refuse to work with any other Garmin models, unless you use something like GPSGate to hide the fact that they are Garmin devices.

...ken...

The patched works with Vista HCx, Legend Cx, 60CSx, and Nuvi 2xx 7xx.
I didn't try Legend HCx, but is strange that didn't work.

I had 5.00.60 installed. I just moved ProITM.exe to that directory and ran it. Is it possible that there are DLL files from the .60 installation causing a problem? Right now the patched .exe behaves exactly the same as the standard version. I suppose I can uninstall and reinstall from the DVD, upgrade to .50 and try the patched version again if you think the .60 installation might be a problem.

...ken...

Certainly from my experience, downgrading without starting again is an iffy situation. I think it has to do with the GarminDevice.xml file though.

Terry

Okay, I removed Mobile PC, reinstalled from the DVD (5.00.30), ran the 5.00.50 update, then moved the patched version to the GarminMobilePC folder and ran it.

Same problem. It identifies a "eTrex Legend HCx software version 2.70", Unit ID: Unknown and at the top of the GPS Info page it says "Invalid GPS". I have checked my Legend and it definitely has a Unit ID.

One thing I should mention is that I have not activated this one. I tried but it has not been six months since I last activated it on my other desktop so it won't activate. If I hit the Disable button so I can force it to start the search over, it says that I need to connect a Garmin GPS or else I need to activate the software.

...ken...

It looks like it must be related to the fact that it's not activated because I just tried the GPS10x and it refuses to use it, too. Same deal,

"Could Not Authenticate Garmin GPS Hardware"

But it lists the device, "GPS10x Software v2.20, BT v2.70"
GPS Unit ID: Unknown

And on the GPS Info page "Invalid GPS".

I guess I'll have to dig the laptop back out and try it on an activated system.

...ken...

Now this is bizarre. I just left Mobile PC (patched version) running while I entered that last message so I could make sure I got the messages right. When I was done with the message I noticed that the satellite constellation on the GPS Info page was moving and the signal strength bars are behaving just like you would expect from a connected GPS. When I went back to the front page and hit the map, it's showing my current location.

The GPS Info page still says "Invalid GPS" at the top of the page and the location information, accuracy, speed and elevation fields all remain blank. But the silly thing is clearly reading and using the GPS data from the GPS10x.

So I'm going to go back and see if it will actually use the Legend even though it declares it's invalid.

More to come......

...ken...

Yep, same thing with the Legend. I hooked up and after Mobile PC was done complaining that it couldn't validate it if I wait for awhile the satellite constellation starts to move, showing that data is coming in. When I view the map it is at the correct location, even though at the top of the page it says "Searching for satellites".

So now I'll have to take the laptop out for a spin again, only this time I'll give it some time to see if it will use the Legend to navigate even though it says it won't.

Garmin, what a stupid way to write a program.

...ken...

Last post on this tonite, I promise.

So, out of curiousity I fired up the unpatched version of 5.00.50 and it behaves exactly the same way. It swears it can't validate the GPS but if I view the map it is showing the correct location and it's "wiggling" around the way it always does because of the weak signals I get down here in my basement office.

It looks like I get the same behaviour no matter whether I use the patched or unpatched version.

...ken...

You can't set up a route while it is 'Searching for Satellites'.

Terry