Forums > Windsurfing   Gps and Speed talk

Coros GPS watches

Reply
Created by choco > 9 months ago, 27 Nov 2020
choco
SA, 3995 posts
27 Nov 2020 2:06PM
Thumbs Up

Has there been any discussion on the Coros GPS watches? GPS Speed Surfing look like they are in the approval phase, nice looking hardware.

morts
WA, 172 posts
27 Nov 2020 1:38PM
Thumbs Up

Links to some info on GPS Speedsurfing

www.gps-speedsurfing.com/default.aspx?mnu=item&item=coros

www.gps-speedsurfing.com/default.aspx?mnu=forum&forum=6&val=166382

Looks like it's just the Vertix & Apex Pro that will have the Speedsurfing mode. They are not cheap!

sailquik
VIC, 6068 posts
27 Nov 2020 7:10PM
Thumbs Up

They have not shown that they do anything different from the Garmin watches with the GPSTC data field from Tbwonder. Not even confirmed that they record actual Doppler speed, and of course, no Doppler error data. Could be good for on the water feedback though.

It appears that they are taking the Positional data for the speed, and then trying to fiddle it with smoothing algorithms to get god only knows what?? They do not explain how they think they can get verifiability and they do not explain how they have evaluated accuracy. Nothing I have heard or read so far has given me any confidence they will achieve what they are proposing - verifiable accuracy at the level of the Motion and Locosys devices.

It appears to me that GPS-SS has a dilemma. They have come near the end of the line with the reliability/quality control problems with the Locosys GW-60 and they dont see the (far superior) Motion devices as the next step due to production delays caused by the lack of a suitable LDC screen. This would seem to threaten their business model so another alternative is being desperately sought.

I dont think it will change much though. The top GPS-SS speeders will still use the top line devices to protect their credibility, and those who are already using non verifiable, lesser devices will continue to do so. As GPS-SS have emphasised in the past, it is just a fun ranking anyhow, not a competition. All will be well as long as they dont try to pass it off as something it is not. (Both their Choros and their rankings)

tbwonder
NSW, 639 posts
27 Nov 2020 10:45PM
Thumbs Up

I have had a bit of feedback from the guys at GPS Speedsurfing.
Firstly this "Windsurfing" functionality is only available on the top end models ( APEX Pro and VERTIX)
When running in "Speedsurfing" mode they will use doppler speed.
They do not record SDOP data as that source code is apparently the property of Locosys.
However Coros have provided additional specific information from their chipsets and will use their own algorithm to produce some form of error data. My understanding is that this error data will only be viewable by GPS Speedsurfing admin in order to verify the accuracy of the files. So the error data will be linked to GPSSS when you sync your watch with the Coros cloud platform.
Apparently they have been testing the system against a 25Hz uBlox based logger and getting good results.

This is all very well and I look forward to seeing a scientific document explaining the mathematics behind the error data that they are collecting and a some statistical demonstration to justify whatever accuracy that they are claiming.

JulienLe
402 posts
27 Nov 2020 8:32PM
Thumbs Up

The whole deal of velocity/position accuracy is that it's a byproduct of velocity/position computation.

The whole deal of Sony CXD5603 (used by most recent watches) is to provide a very-low-power low-update-rate receiver and compensate with an IMU and an "algorithm" supplied by Sony.

Sony advertises a 6mW consumption, an uBlox M8 at 10Hz consumes 75mW, an uBlox M9 at 25Hz consumes 120mW.
"For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled." -- Feynman

sailquik
VIC, 6068 posts
27 Nov 2020 11:47PM
Thumbs Up

Yes, they will have to provide some solid evidence that the data is directly Doppler derived. It is possible to 'fake' Doppler. The chip maker would have that documentation.

Error data that is not viewable by the user or in the users analysis software is useless, being completely non transparent, unable to be peer reviewed. Sounds highly suspect to me! If they can see it, then anyone should be able to see it. That is IF they actually have anything!

The 'source code' description is not correct. The chipset itself produces the error and the Sony Chipset does not,can not. Even if it were some sort of 'source code' reverse engineering or something, the Locosys system (not a Sony Chipset) would not work with the Sony Chipset anyhow. It would have to be engineered from scratch for that chipset.

"25Hz ublox based logger"? Hmmm, I dont know that one. Maybe this is a new Ulbox chipset I have not caught up with? Possible, but why do that when there is very well known M8 10Hz multi GNSS data that is already well known?

I agree Andrew. They will have to do a lot more than throw around fancy rhetoric to get credibility. As has been said before. A few random comparison tests just do not cut it. Proper scientific documentation and explanation would be required at a minimum. Mathematics and a paper might if it is available for peer review.

It will be interesting to see what they come up with to say the least.

boardsurfr
WA, 2202 posts
28 Nov 2020 3:32AM
Thumbs Up

Select to expand quote
tbwonder said..
This is all very well and I look forward to seeing a scientific document explaining the mathematics behind the error data that they are collecting and a some statistical demonstration to justify whatever accuracy that they are claiming.


Really funny!

tbwonder
NSW, 639 posts
28 Nov 2020 10:31AM
Thumbs Up

Select to expand quote
sailquik said..
"25Hz ublox based logger"? Hmmm, I dont know that one. Maybe this is a new Ulbox chipset I have not caught up with? Possible, but why do that when there is very well known M8 10Hz multi GNSS data that is already well known?



This is the new 25hz logger they are using as a reference
shop.sailmon.com/product/max/

sailquik
VIC, 6068 posts
28 Nov 2020 12:58PM
Thumbs Up

Select to expand quote
tbwonder said..
This is the new 25hz logger they are using as a reference
shop.sailmon.com/product/max/



Wow! Nice device! I can see how this would appeal to the Yachting crowd. And at a price they can afford.

This reminds me again of how quickly things can move. It feels like it was only a few months ago that I studied the specs of the first of the M9 modules (the high prescision version) and they were saying it would take a while for the introduction of the normal precision M9 models. Well that didnt take long at all, and that is a massive jump up in specs! 4 concurrent GNSS @ 25Hz V's M8 3 concurrent @10Hz. WOW!

That much data is likely to make the analysis programs grunt a bit!


JulienLe
402 posts
28 Nov 2020 4:40PM
Thumbs Up

If anyone wants that in a (Mini) Motion: hit me up, I have some.

bayside inc
QLD, 399 posts
29 Nov 2020 5:02AM
Thumbs Up

Select to expand quote
JulienLe said..
If anyone wants that in a (Mini) Motion: hit me up, I have some.


WOW... do you ever sleep Julien!!!
Any real user-benefits in using it over the regular Mini-motion we already have?

JulienLe
402 posts
29 Nov 2020 3:40PM
Thumbs Up

For record attempts sure, they do 2.5meters every tick at 10Hz. I have a handful of these M9 for a non-windsurf record attempt next year.

I'm waiting for another chip to work on a newer version of Motion, in a year or more. M9 is a weird release.



Subscribe
Reply

Forums > Windsurfing   Gps and Speed talk


"Coros GPS watches" started by choco