Replacing files in s57data screws up "official" charts

[SandMan874]

Hello,

I'm a long time user and big fan of OpenCPN. Being used to Norwegian paper charts, I've always used to replace chartsymbols.xml and rastersymbols-day.png in s57data with the "swedish look" from OpenCPN's website.

For as long as I was using the "semi-legal" CM93 chart databases one could get a hold of on the Internet, all was fine. The two top attachments (with the fewest depth measurements) shows the old 2012 CM93 charts with original (left) and modified s57data (right)

However, when using the same technique with the official charts purchased from o-charts, the chart is absolutely filled with useless symbols, slowing down OpenCPN to the point that it's not possible to use it anymore.
The bottom two attachments shows the chart with original s57data files (left) and modified (right). And yes, that is about 150 000 pink question marks blotting out the entirety of the Shetland islands.

Anybody else encounter the same problem using the "swedish-look" palette with purchased charts from o-charts?

Thanks for any tips.

SandMan874

[bdbcat]

SandMan...
How about a zoom in on the problem area with o-charts? Hard to see much at this scale.


Thanks
Dave

[Hakan]

Sandman..
I tested on my Swedish o-charts files wo problem. Win10, OCPN v 5.6.2, o-charts_pi v 1.0.20.0

I used the alternative palette linked to here: https://opencpn.org/wiki/dokuwiki/do...vector_palette

[Hakan]

Dave.. bnc..Sandman..
Well I have now also tested the UK charts for the Shetlands and see the same for some cells in different scales.
There are also some diagonal lines wo a clear relevance?
If I use the original chartsymbols it's clear these chart files contains unusual symbols and O is also then corresponding irresponsible.
See a single chart file example in one of my shots. The last shot is the same chart file but with normal palette.

[SandMan874]

Yeah, sorry guys, I should have clarified that the problem is worst around the Shetland islands. I used a zoomed out view just to show how big the problem with the 150 000 pink question marks was. I tried zooming in on the Shetlands, but it actually crashed my OpenCPN.

I see that Hakan managed to replicate the problem anyway. Still don't understand why OpenCPN reacts this way to the changed palette files.

I do also encounter similar issues along the Norwegian coast, with pink question marks or pink X's showing up where there was none when using the modified palette.

SandMan

[bdbcat]

Just for reference, what are the "?" objects? What does object query say?
Thanks
Dave

[Hakan]

Dave..
There's no object query to catch for any question mark. I've tried to find a "base" point for the ?-symbol but wherever I try on the symbol the info is the same as surrounding areas wo a ?-mark. See shot.
I get a unverified impression the ?-marks are connected to lights in some way. Compare shot 2 and 3 about the same area with and without the "SE"-palette.

Today I didn't see the diagonal lines mentioned yesterday? O is reinstalled.
Normally I don't use this palette for vector charts. So never seen before.

[sailorF54]

My guess is that the Swedish chartsymbol.xml file is not up-to-date, that is not tweaked from the latest chartsymbol file...


PS I keep my own French SHOM file updated (and posted on a French forum, but not on https://opencpn.org/OpenCPN/info/vectorpalette.html

[Hakan]

@sailorF54
Yes, could be so. Now I'm not aware of the details but I can see there are differences in listed RCID in the lookup tables between yours "SHOM" and the available SE-file.
@SandMan874
It may bee you should not use the "Swedish" palette until it's adapted to the chart symbols used by O v 5.6.2?
I'm sorry but I don't have knowledge enough to maintain it.

[requiem]

If you have the ENC version of chart 1 installed, you should be able to zoom into the different symbol areas (located somewhere in the middle of Mali) and spot the ones that aren't rendering as they're supposed to.

I'll see if I can locate a copy later this evening; example screenshot attached...

[Hakan]

No I don't have "Chart 1". Is that a US chart? What I can found it's not beyond the noaa ENCs by Chartdownloader?
Thanks
Håkan

[bcn]

As far as I know there is no ENC Chart1 - its a booklet which explains very well the presentation of objects one will find in ENCs:



For example:
Chart 1 is a publication containing explanations of the symbols, abbreviations and terms needed to interpret nautical charts published by the Canadian Hydrographic Service. Member nations of the International Hydrographic Organization produce Chart 1 using a common format.


A very good reference:
https://nauticalcharts.noaa.gov/publ...s-chart-1.html

[requiem]

ECDIS (sorry!) Chart 1 was created to ensure that a system can correctly display all symbols in the current edition (4.0) of the IHO S-52 presentation library. As it's typically installed on ECDIS systems finding a copy in the wild is less common, but it appears you should be able to grab one from here:

https://support.navtor.com/support/s...iho-test-cells

[bdbcat]

OP...
The alternative chartsymbols.xml file available in the Wiki is missing a recently defined symbolization directive not often seen, but apparently common on UK ENCs.
An updated version is here:
https://www.dropbox.com/s/iwtbi6n1vl...mbols.xml?dl=0


Please take a look, and let us know how it goes.
Thanks
Dave

[sailorF54]

The updated file seems to do the job perfectly....