Select to expand quote
JulienLe said..
- Notice how it bottoms out at 18 satellites used. ;) ;)
Actually, it does not "bottom out". All accuracy measures keep improving after 18 satellites. This trend continues to into the high 20s.
Select to expand quote
JulienLe said..
- Filtering by C/N0 had a short lifespan in Motions. It was overall a bad idea. Not recommending it to anyone.
But the Motion still is limiting the number of satellites tracked, correct? Which means that the GPS either throws out extra satellites randomly (which would be bad), or ranked by a measure that's most likely CNO. The only drawback of a fixed CNO requirement is that the number of satellites can drop low if the reception is poor. The graphs above actually show an example of that.
Select to expand quote
JulienLe said..
- Cold stationary tests are always a bad idea. The dynamic filter needs movement to sort itself out.
This was not a "cold" test. And re-iterating that stationary tests are a "bad idea" does not make it true. As Tom Chalko pointed out many years ago, the idea of "stationary" is an nonsense, given that GPS satellites move at more than 7500 knots relative to earth. The only issue that needs to be taken into account is the effect of speed directionality.
As for dynamic filters, those were quite heavily used by Locosys, especially by the GT-31. They distorted the alpha numbers quite a bit, and the "zero speed" filter that is still present in the GW-60 still creates errors at the start of runs. Only two dynamic filters that are apparent in the u-blox data: 1. the sAcc filters, which slows down the increase of sAcc numbers to make sAcc almost useless (sAcc usually exceeds thresholds only when the min. satellite filter is also triggered); and 2. the "extrapolate when reception is lost" filter, which can keep speed at pre-crash levels for several seconds after a crash, instead of letting it drop to zero. Those filters make some sense for GPS use in cars, but they are only detrimental for speedsurfing.
Select to expand quote
JulienLe said..
- Anyway, it's HDOP they want. So it's HDOP they'll get if everyone is cool with it.
The guy who asked the question appears to be somehow associate with a GPS watch maker, not with the GPSTC or any speed competitions. On the software side of things, HDOP filters are relics from pre-GT-31 times. With the current default setting of 5.0 in GPSResults, they are virtually never triggered when a GPS has reception of 5 satellites (the minimum for GPSTC postings).
If you want to spend your time modifying the Motion firmware to store HDOP values instead of PDOP values, because GPSResults uses the PDOP values from Motion files in the column labeled "HDOP", nobody can stop you, but don't claim that GPSTC or any analysis software would see any benefit from this.