I've seen plenty of posts regarding the inaccuracy of circuits in some
motor racing sims, Hungary in F1 2001 springs to mind, and I've noted
software publishers proudly boasting that they use GPS data to create
their versions of the tracks.
My knowledge of the GPS system led me to doubt that the system was
accurate enough to create totally authentic rendition of a track so I
did some research and found the following quote
"Typical normal accuracy (one-sigma) for welldesigned civil equipment
under nominal operating conditions with SA off should be about 10 m
horizontal and 13 m vertical."
For those not yet cursed with Napoleon's measuring system (metric),
that's about 32 feet horizontal and 42 feet vertical.
Now, for navigation, to find a harbour or an airport or to get you
home when hiking, that's more than accurate enough but I feel an error
of that size would make modelling a race circuit accurately very
unlikely.
For further info on GPS accuracy, check out the links below
http://www.ngs.noaa.gov/FGCS/info/sans_SA/compare/ERLA.htm
http://www.edu-observatory.org/gps/check_accuracy.html
Peter
Peter Holt <[email hidden]> wrote:
: My knowledge of the GPS system led me to doubt that the system was
: accurate enough to create totally authentic rendition of a track so I
: did some research and found the following quote
:
: "Typical normal accuracy (one-sigma) for welldesigned civil equipment
: under nominal operating conditions with SA off should be about 10 m
: horizontal and 13 m vertical."
Yes, that's the standard normal accuracy for a GPS receiver you
buy from your local store cheaply. But I doubt they used those
cheap ones for the tracks, because you can rent/buy better ones
if you have the funds.
Differential GPS is the way to go for better accuracy. With good
equipment you can get <1 meter accuracy.
For a track layout, you don't need the absolute position, so it's
possible to use more than one GPS receiver to accurately get the
relative position between the receivers.
TTT
On 9 Jun 2002 14:24:29 GMT, Tauno Taipaleenmaki
<[email hidden]> wrote:
I'll repeat part of my quote, "welldesigned civil equipment". As to
differential GPS and other methods of improving the accuracy, I have
no evidence that any of that was used and given the errors in tracks,
every reason to believe that they weren't.
Further, I would argue that 1 meter accuracy is still not sufficent,
if you have a 1 meter error in where the apex of a corner is located,
you would alter the nature of the bend completely.
> For a track layout, you don't need the absolute position, so it's
> possible to use more than one GPS receiver to accurately get the
> relative position between the receivers.
How does this work?
Peter
Quoted text (9 lines)
> > Yes, that's the standard normal accuracy for a GPS receiver you > buy from your local store cheaply. But I doubt they used those > cheap ones for the tracks, because you can rent/buy better ones > if you have the funds. > > Differential GPS is the way to go for better accuracy. With good > equipment you can get <1 meter accuracy. >
> "Typical normal accuracy (one-sigma) for welldesigned civil equipment
> under nominal operating conditions with SA off should be about 10 m
> horizontal and 13 m vertical."
I've used GPS a great deal in Gliding -and specifically gps data logging
during competition gliding.
I once took a GPS reciever and a data logger in a car and drove all around
the airfield perimiter track, each side of the runway and all service
roads - then used this data to plot a trace.
Using a large ord. survey map ( 1:2500 ) and arial photographs - we
determined that the absoloute accuracy was less than 10 metres - and the
relative accuracy was a great deal less than a metre. ( i.e. it positioned
one end of a runway relative to the other end with an accuracy of about 75
cm)
We were unable to test it's accuracy altitude wise - however - comparing
baro and GPS altitude recorded by some GPS data loggers - it's usually
accurate to a dozen feet - BUT - this accuracy is usually consistant.
So -this is with <�1k gear fitted to most competition gliders.
With the �10k gear that has a proper large GPS antenna ( usually looks like
a round tupaware thing on a large stick) - the accuracy can be even higher.
It's worth mentioning that in the past 18 months, the US military have
de-coded GPS which increased it's accuracy by a factor of between 5 and 10.
Futhermore - accuracy increases the more sats. you have connected. the more
accurate you are. 4 Sats and you'll be +/- 50m, 8 sat and you'll be
laughing!
Doug
"Peter Holt" <[email hidden]> wrote in message
news:[email hidden]...
> I've seen plenty of posts regarding the inaccuracy of circuits in some
> motor racing sims, Hungary in F1 2001 springs to mind, and I've noted
> software publishers proudly boasting that they use GPS data to create
> their versions of the tracks.
The only circuits modelled with GPS are those in F1RC as far as I'm aware.
And compared to F12001 and F12002 - IT SHOWS
DOug
Of course, that's assuming that the marketing wizs aren't stretching the
truth a bit with their "GPS used for the tracks" hype.
Nah, they wouldn't do that :)
-John
"Peter Holt" <[email hidden]> wrote in message
news:[email hidden]...
Quoted text (28 lines)
> I've seen plenty of posts regarding the inaccuracy of circuits in some > motor racing sims, Hungary in F1 2001 springs to mind, and I've noted > software publishers proudly boasting that they use GPS data to create > their versions of the tracks. > > My knowledge of the GPS system led me to doubt that the system was > accurate enough to create totally authentic rendition of a track so I > did some research and found the following quote > > "Typical normal accuracy (one-sigma) for welldesigned civil equipment > under nominal operating conditions with SA off should be about 10 m > horizontal and 13 m vertical." > > For those not yet cursed with Napoleon's measuring system (metric), > that's about 32 feet horizontal and 42 feet vertical. > > Now, for navigation, to find a harbour or an airport or to get you > home when hiking, that's more than accurate enough but I feel an error > of that size would make modelling a race circuit accurately very > unlikely. > > For further info on GPS accuracy, check out the links below > > http://www.ngs.noaa.gov/FGCS/info/sans_SA/compare/ERLA.htm > http://www.edu-observatory.org/gps/check_accuracy.html > > Peter >Peter Holt <[email hidden]> wrote:
: I'll repeat part of my quote, "welldesigned civil equipment". As to
: differential GPS and other methods of improving the accuracy, I have
: no evidence that any of that was used and given the errors in tracks,
: every reason to believe that they weren't.
Well, only those who made the tracks know what was used, but
I agree, the errors in the tracks are bad and obvious.
: Further, I would argue that 1 meter accuracy is still not sufficent,
: if you have a 1 meter error in where the apex of a corner is located,
: you would alter the nature of the bend completely.
That's also true.
The proper way to do the tracks would be to go to the people
who built the tracks, and get accurate drawings from them.
Then they wouldn't have to guess :) I'm sure there has to be
accurate blue-prints of all the tracks somewhere.
TTT
On Sun, 9 Jun 2002 15:47:22 +0100, "Douglas Ellison"
<[email hidden]> wrote:
If you check the links in my original post, you'll see that most of
the factors that degrade the accuracy of GPS are not related to the
receiver. The most significant factor is ionosphereic distortion of
the signal
>
>It's worth mentioning that in the past 18 months, the US military have
>de-coded GPS which increased it's accuracy by a factor of between 5 and 10.
That's the removal of SA (selective availability) in 2000. The
accuracy figures I quoted are with SA off.
>Futhermore - accuracy increases the more sats. you have connected. the more
>accurate you are. 4 Sats and you'll be +/- 50m, 8 sat and you'll be
>laughing!
>
>Doug
>
Quoted text (8 lines)
> >> "Typical normal accuracy (one-sigma) for welldesigned civil equipment >> under nominal operating conditions with SA off should be about 10 m >> horizontal and 13 m vertical." > > >With the �10k gear that has a proper large GPS antenna ( usually looks like >a round tupaware thing on a large stick) - the accuracy can be even higher.>"Peter Holt" <[email hidden]> wrote
> For further info on GPS accuracy, check out the links below
"GPS" track measuring is mostly a PR gimmick for game publishers. The
developpers mostly use it for basic things like the position of objects
around the track (from the grandstands to trees). The track itself is
modeled from the architectural layouts of the track and very inside
information. That is what I was told.
That is, if the developper has the time to model it that way. Most of them
(in the most blattant case, ISI and somehow Geoff Crammond's team) decide to
go the quick way. Realism in track layout is very difficult. The basics
are easy, but with each layer of detail come in another layer of complexity.
Therefore, time and therefore, money for the mentality of game producers.
Shame on EA, I say. Shame on ISI for not pushing the accuracy of tracks as
important to EA.
--
-- Fran�ois M�nard <ymenard>
-- http://www.ymenard.8m.com/
-- This announcement is brought to you by the Shimago-Dominguez
Corporation - helping America into the New World...
> The proper way to do the tracks would be to go to the people
> who built the tracks, and get accurate drawings from them.
> Then they wouldn't have to guess :) I'm sure there has to be
> accurate blue-prints of all the tracks somewhere.
The track wont necessarily match the blue prints.
Arial photograph, high scale maps, and plenty of photographic and video
research should be enough - GPS is a nice time saver in some respects.
Doug
F1RC uses GPS. I've seen F1RC screenshots and real pics merged together, and
F1RC has the tracks almost perfect.
David G Fisher
"Peter Holt" <[email hidden]> wrote in message
news:[email hidden]...
Quoted text (28 lines)
> I've seen plenty of posts regarding the inaccuracy of circuits in some > motor racing sims, Hungary in F1 2001 springs to mind, and I've noted > software publishers proudly boasting that they use GPS data to create > their versions of the tracks. > > My knowledge of the GPS system led me to doubt that the system was > accurate enough to create totally authentic rendition of a track so I > did some research and found the following quote > > "Typical normal accuracy (one-sigma) for welldesigned civil equipment > under nominal operating conditions with SA off should be about 10 m > horizontal and 13 m vertical." > > For those not yet cursed with Napoleon's measuring system (metric), > that's about 32 feet horizontal and 42 feet vertical. > > Now, for navigation, to find a harbour or an airport or to get you > home when hiking, that's more than accurate enough but I feel an error > of that size would make modelling a race circuit accurately very > unlikely. > > For further info on GPS accuracy, check out the links below > > http://www.ngs.noaa.gov/FGCS/info/sans_SA/compare/ERLA.htm > http://www.edu-observatory.org/gps/check_accuracy.html > > Peter >WITHOUT GPS
would it be possible to get a good and accurate
mapping ??
let's read some of the responses to this:=
:=))
"Peter Holt" <[email hidden]> wrote in message
news:[email hidden]...
> On 9 Jun 2002 14:24:29 GMT, Tauno Taipaleenmaki
> <[email hidden]> wrote:
>
> >
> > Yes, that's the standard normal accuracy for a GPSreceiver you
> > buy from your local store cheaply. But I doubt theyused those
> > cheap ones for the tracks, because you can rent/buybetter ones
> > if you have the funds.
> >
> > Differential GPS is the way to go for betteraccuracy. With good
> > equipment you can get <1 meter accuracy.
> >
>
> I'll repeat part of my quote, "welldesigned civilequipment". As to
> differential GPS and other methods of improving theaccuracy, I have
> no evidence that any of that was used and given theerrors in tracks,
> every reason to believe that they weren't.
>
> Further, I would argue that 1 meter accuracy is stillnot sufficent,
> if you have a 1 meter error in where the apex of acorner is located,
> you would alter the nature of the bend completely.
>
> > For a track layout, you don't need the absoluteposition, so it's
> > possible to use more than one GPS receiver toaccurately get the
> > relative position between the receivers.
>
> How does this work?
>
> Peter
Once upon a time a transit and a couple of people who knew how to use it
was considered good enough.
"John Metcow" <[email hidden]> wrote in message
news:[email hidden]...
Quoted text (10 lines)
> WITHOUT GPS > would it be possible to get a good and accurate > mapping ?? > > > let's read some of the responses to this:= > > > :=)) >>"John Metcow" <[email hidden]> wrote
> WITHOUT GPS
> would it be possible to get a good and accurate
> mapping ??
Of course, like I wrote it's mostly a PR gimmick. It just takes time, and
that is not always something developers have.
--
-- Fran�ois M�nard <ymenard>
-- http://www.ymenard.8m.com/
-- This announcement is brought to you by the Shimago-Dominguez
Corporation - helping America into the New World...
"John Metcow" <[email hidden]> wrote in message
news:[email hidden]...
> WITHOUT GPS
> would it be possible to get a good and accurate
> mapping ??
Yes there are mapping techniques far better suited to this type of modelling.
GPS provides point-data. And loosely accurate point data at that. At best you
end up with a XYZ co-ordinate at each side of the track, say every 10-15m along
the route. That is a loose model to build up the subtle rises and falls in the
surface. And requires about 1000 samples per track, which equates to a lot of
man hours. For a sloppy model.
What you really want is stream data providing continuous contours of the
surface. "Digital Stereo Photogrammetry" is tailor made. This uses overlapping
vertical aerial photographs to pick up the contours in the overlap. It would be
possible to pick up about a 10cm change in elevation and cover the *whole* track
with accurate co-ordinates in one evening. It is very difficult to explain
thoroughly, check it out here if you are curious:
http://www2.erdas.com/supportsite/focus3/explanation.html#DTM_extract
Photogrammetry is ifininately more thorough and accurate and more importantly,
doesnt require a team junket to every racetrack. Just hire a local surveyor at
each track to organise the aerial photo's and voila
I laugh every time I read about GPS accurate tracks.
Stu
ps. Attn. ISI / GP 2.3.2000.4 team - Im available to fix your models - for ever.
And i'm cheap - you can pay me in Ozzie dollars ;-)
Quoted text (50 lines)
> let's read some of the responses to this:= > > > :=)) > > > > "Peter Holt" <[email hidden]> wrote in message > news:[email hidden]... > > On 9 Jun 2002 14:24:29 GMT, Tauno Taipaleenmaki > > <[email hidden]> wrote: > > > > > > > > Yes, that's the standard normal accuracy for a GPS > receiver you > > > buy from your local store cheaply. But I doubt they > used those > > > cheap ones for the tracks, because you can rent/buy > better ones > > > if you have the funds. > > > > > > Differential GPS is the way to go for better > accuracy. With good > > > equipment you can get <1 meter accuracy. > > > > > > > I'll repeat part of my quote, "welldesigned civil > equipment". As to > > differential GPS and other methods of improving the > accuracy, I have > > no evidence that any of that was used and given the > errors in tracks, > > every reason to believe that they weren't. > > > > Further, I would argue that 1 meter accuracy is still > not sufficent, > > if you have a 1 meter error in where the apex of a > corner is located, > > you would alter the nature of the bend completely. > > > > > For a track layout, you don't need the absolute > position, so it's > > > possible to use more than one GPS receiver to > accurately get the > > > relative position between the receivers. > > > > How does this work? > > > > Peter >I agree with you on that. People seem to confuse the global GPS system
with the GPS technology itself. The tracks in F1RC are not done using
GPS satellites for data. Instead they had these small devices scanning
the surface of the track and they indeed made a worldtrip to get the
data. Correct me if Im wrong.
LM
