There's a bit of an argument on GPSTC between a couple of Dutch speed sailors regarding the credibility of max 2 sec speed recorded recently. Through some super modern technology (KA-72 website http://www.ka72.com/biglist.aspx) I had a look at the track which seems okay (from a layman's point of view) although the deceleration is extremely sharp. The track's 1 sec speeds either side of the peak are as follows;

29.892 knots
30.786
33.93
37.649
41.808
43.615
41.003
37.348
28.712
22.631

The argument against the validity of this was that you can't accelerate or decelerate this quickly [Edit - Apart from driving into a brick wall Choco]. I'm wondering what the experts think. Anyone care to comment.

Select to expand quote

sausage said...


The argument against the validity of this was that you can't accelerate or decelerate this quickly. I'm wondering what the experts think. Anyone care to comment.

Easy drive into a brick wall!

Select to expand quote

choco said...

sausage said...


The argument against the validity of this was that you can't accelerate or decelerate this quickly. I'm wondering what the experts think. Anyone care to comment.

Easy drive into a brick wall!

Choco,
I said experts....not smart arses That's exactly something I'd say.

Chris,
It didn't appear that the deceleration was caused by a crash either. The track stays around 20knots for a while after.

Yep - you can definitely decelerate that fast - but it aint pretty!

Seriously though, I haven't a clue.

Yoyo - this is your kinda question.

The deceleration looks especially suss.

It is about twice the deceleration of Tim's 41+ kn Melville track which was a track smoothed spike.

37kn to 28kn in one second is only seen in crashes.

it is also seen in spin-out...

1 knot = 0.514 m/sec. 30.8 to 43.6 knots in 4 seconds is an acceleration of (12.8 * 0.514)/4 =

1.65 m/(sec.sec) or 0.17 g.

Or applying F=ma, the net force acting in the direction of travel on a 100 kg rig and rider would be 100*1.65 = 165 newtons. That's a 16.8 kg component in the forward direction. The sail has to generate that plus the drag forces. All going thru the boom and the mast track - You'd have to be going broad to get that much sail pressure in the forward direction. You'd be hanging on!

so is that trackpoint or doppler???

Just checked for myself, here's the doppler table.
Still a big jump going from 27 to 38
So the biggest acceleration is 10kts in one second, the biggest decceleration is 5kts in 1 sec

27.14
27.33
38.80
42.20
42.94
44.77
39.60
36.45
31.82
30.63
29.24
28.07
27.78
27.27

When i was up in Qld went to Dreamworld and was tempted to take my GT31 on the Tower of Terror a magnetic propelled car that goes from a standing start to 160km in 7 secs before it reaches 38 stories and free falls back down.
The only reason i didn't take it was because to get onto the car we had to wait inside a large shed and i thought the gps would loose sink,would have been interesting though.

Select to expand quote

decrepit said...

so is that trackpoint or doppler???

Just checked for myself, here's the doppler table.
Still a big jump going from 27 to 38
So the biggest acceleration is 10kts in one second, the biggest decceleration is 5kts in 1 sec

27.14
27.33
38.80
42.20
42.94
44.77
39.60
36.45
31.82
30.63
29.24
28.07
27.78
27.27

Decrepit,
is the track you noted the same track I am refering to as the numbers seem completely different. It takes a little finding on the KA-72 big list (track date 4th April in Gruissan, Languedoc http://www.ka72.com/TrackRecalc.aspx?fid=4901)

PS - Choco, I went on that tower of terror about 3 years ago. The waiting at the top with the thing making ticking noises etc really does your head in. I gather as well that they sat you in the middle so the tower wouldn't fall over

Select to expand quote

PS - Choco, I went on that tower of terror about 3 years ago. The waiting at the top with the thing making ticking noises etc really does your head in. I gather as well that they sat you in the middle so the tower wouldn't fall over

Thankfully they did

Select to expand quote

sausage said...


PS - Choco, I went on that tower of terror about 3 years ago. The waiting at the top with the thing making ticking noises etc really does your head in. I gather as well that they sat you in the middle so the tower wouldn't fall over

When I went on it, they dropped us about 20m and paused just for long enough for you to think 'gee, they must be doing something', and then dropped it the rest of the way. You can't scream as there's no air in your lungs. Great stuff... when you are back on the ground

I think Choco was talking about the other half of the ride where you shoot up it in a roller coaster and then come down backwards. I think it takes about 20 to 30 minutes to get on a ride that lasts a few seconds. I wonder how many trackpoints you would get as it doesn't last very long at all.

Select to expand quote

sausage said...
>>>>>>>

Decrepit,
is the track you noted the same track I am refering to as the numbers seem completely different. It takes a little finding on the KA-72 big list (track date 4th April in Gruissan, Languedoc http://www.ka72.com/TrackRecalc.aspx?fid=4901)

Yep, the very same, but mine are the doppler figures not the trackpoints that you posted.
The doppler stuff is supposed to be more accurate, and there quite often is some sort of discrepancy between them.

The fact that a similar pattern is there, plus the fact there is no gap in the time sequence and 7 satellites are in view, gives me the feeling it's probably real.

Looking at the track, he starts bearing off several seconds before the acceleration, I think he's seen a strong narrow bullet on the water in front of him, and gone for it. Must have been very dead on the other side of the bullet though, I don't think he spun out, there's no deviation in his direction when he starts decelerating.

when i was young i used to go on the corkscrew and pretend i was doing cheese rolls.

^ Apologies Decrep, I just realised I haven't used realspeed on my work computer for over a year and it's still probably set to trackpoints. Home computer won't let me open this file on realspeed for doppler though.

Select to expand quote

FormulaNova said...

sausage said...


PS - Choco, I went on that tower of terror about 3 years ago. The waiting at the top with the thing making ticking noises etc really does your head in. I gather as well that they sat you in the middle so the tower wouldn't fall over

When I went on it, they dropped us about 20m and paused just for long enough for you to think 'gee, they must be doing something', and then dropped it the rest of the way. You can't scream as there's no air in your lungs. Great stuff... when you are back on the ground

I think Choco was talking about the other half of the ride where you shoot up it in a roller coaster and then come down backwards. I think it takes about 20 to 30 minutes to get on a ride that lasts a few seconds. I wonder how many trackpoints you would get as it doesn't last very long at all.

Yer one side is the giant drop and the other is the tower of terror.
Choco wouldnt go on the giant drop the big girl great view from up there too.

OT (well sort of)- would a GT-31 be able to record speed accurately (vertically) up and down the tower? or would it be read as a dead stop in the horizontal plane.

Not sure of that, but somebody did say once that the altitude measurement wasn't good enough for guys jumping waves.
Your home version of real speed must be fairly old. the early versions didn't recognise .sbn files.

I wasn't using realspeed, that doesn't give the doppler speed table. I was using gpsarpro

The PB run of Frank in GPSARPro:

"Looks" ok?

The discussion on the dutch forum (google translated): translate.google.nl/translate?js=y&prev=_t&hl=nl&ie=UTF-8&layout=1&eotf=1&u=http://forum.windsurfing.nl/viewtopic.php%3Ff%3D62%26t%3D13863190&sl=nl&tl=en

Select to expand quote

Gestalt said...

when i was young i used to go on the corkscrew and pretend i was doing cheese rolls.

were you eating the cheese or just rolling in it?

Select to expand quote

mathew said...

it is also seen in spin-out...

Not usually without a crash.

Seriously, I have had spin-outs at 40+ knots (bounced over wake wave) and the board keeps going frighteningly fast for a loooong time with the fin ventilating (if you can save it from crashing).

Just spent an interesting half and hour analysing the run and it is hard to see the error in the data in RealSpeed, except that RealSpeed sees the Doppler speed as invalid (too much accelleration - with acceleration option set to 3m/sec2).

What is interesting is that the Doppler defies norms and peaks higher and earlier than the trackpoint speed which RealSpeed accepts (must be just under the limit).

There is one other sign that may indicate an error. That is the sudden loss of altitude that begins right on the peak speed..
In any case I don't think I would accept this just because the Doppler peak is clearly higher than the trackpoint peak and 1 second earlier at that. On top of that is the unnatural accelleration and deceleration and the apparent sudden loss of altitude.

[URL=/] [/URL]

Decrepit & Andrew,
Thanks for your feedback - so we're really at a stalemate on the validity of this specific track as it appears from the Dutch website posted by Mark that it's a 50/50 split as well.

BTW - When looking at realspeed screen image on Sailquick's thread above it's washed out and very hard to decipher, but when you are in the reply box and scroll down it is crystal clear. What's the go with that?

Select to expand quote

sausage said...

Decrepit & Andrew,
Thanks for your feedback - so we're really at a stalemate on the validity of this specific track as it appears from the Dutch website posted by Mark that it's a 50/50 split as well.

BTW - When looking at RealSpeed screen image on Sailquik's thread above it's washed out and very hard to decipher, but when you are in the reply box and scroll down it is crystal clear. What's the go with that?

For clarity, just click on the bar at the top of the photo and it comes up in a new window at full size and resolution. :-) edit: You may then have to click on the photo to enlarge it to full size. My Firefox browser shrinks it to fit the screen by default.


This is a very difficult one as it is so close to meeting the requirements and it does not look obviously wrong in some views.

The filter setting for GPS-SS was agreed as 3m sec sq. accelleration. One problem is that the trackpoint accelleration is just under that but the Doppler is over it. I have yet to confirm this but it appears that GPS-Results may do the accelleration figures only on the trackpoints so GPS-Results users may not see the 3m sec sq. Doppler acelleration. Not sure about GPSAR-Pro yet. Maybe Decrepit can answer that one? I have acess to it but struggle with the finer details of using the more advanced features of it.

Hmm. Just had a play with GPS-Results and in the Dopper mode the accelleration figure comes up as 2.7 as reported on the Dutch forums. The accel. figure for the trackpoints comes up as 2.1 so there must be a different way of arriving at the accel. figure in each program. I played with the 'filters' for different calc. methods thinking that might be the difference but nothing changed. Got me stumped!
One thing I would say is that the trackpoint figures look a bit more realistic but not even sure about that.

I was trying to remember which tracks I had that show the strange divergence of the trackpoint and Doppler speed graphs but while I was playing around with GPS-Results and looking at the other results from the same session I came across another perfect example. Look at the speed graph at around 9.13.33 (data point 5713) and you will see this strange thing. Here it is clearly an error, but again there are adequate satellites and no other indications of error in the data.

So far I have not seen any of this type of error from the GT-31's.

[URL=/] [/URL]

Select to expand quote

decrepit said...

so is that trackpoint or doppler???

Just checked for myself, here's the doppler table.
Still a big jump going from 27 to 38
So the biggest acceleration is 10kts in one second, the biggest decceleration is 5kts in 1 sec

27.14
27.33
38.80
42.20
42.94
44.77
39.60
36.45
31.82
30.63
29.24
28.07
27.78
27.27

Hi Decrepit,
how did you get this list of doppler speeds? In your list the decceleration is less than in other lists mentioned. It is different from the screenshot by Sailquick, also doppler speeds list on the right in realspeed: /

Thanks,
Mark

I think you will find this is an anomaly that can be traced back to the Sirf chipset. In the two instances the sat number changes from 7 to 8 and the extra sats used are 04 and 31. They also happen to be the lowest on the horizon and have the weakest signal. The increased sensitivity of the Sirf which allows it to use sats with a weaker signal has vacillated between using these 2 sats in the nav solution. The doppler averaging lag can be seen as it attempts a dead reckoning best solution for the situation. The Sirf should have rejected these 2 sats due to different masks used to screen out the weaker sat signals. Additionally the GPS satellite system has recently undergone a major system wide software upgrade that has just gone fully operational.

First time poster, long time reader :P

Roo, I'm afraid I don't quite understand your post. This may be due to English not being my native language, so bear with me: I think you mean to say that the anomaly is due to the Sirf3 chip vacillating between two sattelites which are low on the horizon and therefore low on signal strength. Did I understand this correctly?

If so, the GPS unit the guy used here is the GT-11 which uses the older (and deprecated) Sirf2 chipset. If I understood your post correctly, this contradicts your post. :S

PS: I love your sail brand choice ;)

I stand corrected dBA! Glad you like the sails.

Select to expand quote

MarkvanOsch said...

decrepit said...

so is that trackpoint or doppler???

Just checked for myself, here's the doppler table.
Still a big jump going from 27 to 38
So the biggest acceleration is 10kts in one second, the biggest decceleration is 5kts in 1 sec

27.14
27.33
38.80
42.20
42.94
44.77
39.60
36.45
31.82
30.63
29.24
28.07
27.78
27.27

Hi Decrepit,
how did you get this list of doppler speeds? In your list the decceleration is less than in other lists mentioned. It is different from the screenshot by Sailquick, also doppler speeds list on the right in realspeed: /

Thanks,
Mark

Mark,
I just figured out Decrepit has accidently listed the doppler figures in reverse.

Select to expand quote

sausage said...

choco said...

sausage said...


The argument against the validity of this was that you can't accelerate or decelerate this quickly. I'm wondering what the experts think. Anyone care to comment.

Easy drive into a brick wall!

Choco,
I said experts....not smart arses That's exactly something I'd say.

Chris,
It didn't appear that the deceleration was caused by a crash either. The track stays around 20knots for a while after.

Red thumb him Snags