AusENC S63 large scale cells offset

[aboon]

Dave,
Same location as before for the buoy.

The location given by SeaPro and the N2M is
43 9.7400 S, 146 58.8200 E.

Andrew

[bdbcat]

Andrew...

Time for more experiments.

I built a new version of s63_pi which may help us understand what is going on here.

Here is a link:
https://www.dropbox.com/s/3c96xhymis...win32.exe?dl=0

You may install this version over your existing s63_pi without requiring reload.

This version does two things:
1. The mouse rollover of the chart bar will show the Edition/Update field correctly now, so that we can be sure what OpenCPN is using for cell update application.
2. Any SENC files required are remade on every instance of OCPN startup. This will make the program run a bit slower, but will eliminate any ambiguity that may be caused by file cacheing.

So, lets take a look at the 5174P3 and 5174P4 cells, and note the Edition/Update fields on rollover.

Standing by
Dave

[aboon]

Dave,
Here are the results of the rollover on the two cells.
I have looked at the actual boundaries of these cells and compared them with the values in in the file INFO/PRODUCTS.TXT (in both Base Map and Updates data of AusENC). I haven't found any patterns but I think I can say that the eastern and western boundaries of AU5174P3 (and possibly P4) are 'shifted' by different amounts. Not sure if that is relevant or not.

Andrew

[bcn]

Andrew,

if you forward your PERMIT.txt file to Dave or to us, we will be able to install and run the files. Will make the analysis perhaps bit faster.

Thanks

Hubert

[aboon]

Dave, Hubert,
Here 'tis (a zip file). There has been no new permit file since Week 14.

Andrew

[bdbcat]

Andrew...

Please disregard my PM. I found another way to download the cells.

Thanks, more later
Dave

[bdbcat]

Andrew...

OK, sorted this out.
Here is a new s63 Plugin correcting the troubles, I hope.

https://www.dropbox.com/s/ddfzu05yjy...win32.exe?dl=0

Root cause of the problem was an inadvertent loss of numerical precision while processing a cell update. This was compounded by the simple distance from your longitude to the Greenwich longitude 0, so only readily apparent in Oz. Not a (very) noticeable problem elsewhere.

Let me know how this goes, so we can prepare the "official" update of s63_pi for all platforms.

Thanks for your patience.
Dave

[aboon]

Dave,
I loaded the new plugin and the location of the nav. mark is now OK in Object Query. The SW corner of AU5174P4 is in the correct location.
But (sorry!), the smaller scale chart AU444146 does not meet the P4 cell. It is about 15 m away, about the same as the previous error in the western bound of P4. Attached is a screen shot. The pink strip normally shows as blue, with no map data displayed.
The eastern boundary of P4 does not appear to have moved, giving a discontinuous shoreline, as before.
Definite progress though - well done.
If it would help you, I can give you an InstallPermit for my UserPermit. I have used 4 so have one left. (When Win 10 Pro updated recently, I had to get a new permit). Send me a fingerprint file if you'd like to do this.
You probably know by now, but AusENC base and updates can be downloaded from
www.hydro.gov.au/webapps/jsp/downloads/products.jsp

Andrew

[aboon]

Dave,
Sorry, wrong image. Try this one.

Andrew

[aboon]

Dave,
On the AU5174 P3 cell, the 'gap' (blue) is on the eastern boundary and is about 20 m (0.015') wide.

Andrew

[bdbcat]

Andrew...

Looks like you need to rebuild the OCPN chart database to capture the corrected cell coverage regions. If that does not work, then please try to re-import the S63 cell permits.

Thanks
Dave

[aboon]

Dave,
I guess that should have been obvious - not thinking.
Anyway, good news! Everything seems to be in the right place, no offsets, no discontinuous boundaries or Object Queries with wrong locations.
All of the light characteristics that I couldn't find are now are shown in Object Query results.
Thanks for your perseverance.
The issues with boundary marks remain, ie tick marks facing the wrong way and boundaries not appearing when 'Symbolised' Boundaries are selected at startup. Should I start afresh with them in a new thread?

Thanks again,

Andrew

[bdbcat]

Andrew...

That is good news. We will prepare an official Release of s63_pi for other platforms. Meanwhile, your Windows plugins from Dropbox should be fine for daily use.


re:
The issues with boundary marks remain, ie tick marks facing the wrong way and boundaries not appearing when 'Symbolised' Boundaries are selected at startup. Should I start afresh with them in a new thread?

Please do.

Thanks
Dave