Generic qgs template for ENC/S-57

[free-x]

Hi @quantenschaum ,

could you provide generic template for gpkg which contains s-57 layers?

So far the templates are a mix of s-57 and OSM or other freely available data ( json, gpx etc) that are not s-57 compliant

Regards
free-x

[quantenschaum]

This is a good idea in principle, but practically I adjusted the QGIS files to the data that is actually available.

OSM tagging is not used in my QGIS files, only other schemes that I found in the data.

The data by BSH uses a S57-like tagging, but not exactly and it does not come in S57 layers. There is a script to filter and split it into some layers. It is called in the bsh target and the resulting layers are placed into data/bsh/layers.gpkg.

The data by RWS (NL) comes as actual S57 ENC, but the buoys and beacons are a little outdated, so I pulled those in from a different, more up to date dataset. The navaids from the ENC are available in rws.gqs as well in the aton group, but it is disabled. Just enable the aton group and disable vwm. Then you get a template that ist already pretty close to a generic S57 template. Not all layers are visualized, the yet untagged layers are in the import group, but the most important ones are visualized in the enc group.

Another problem is, that despite there is an official S57 specification, most real life ENCs out there do not stick exactly to these specs, so some manual fine tuning is always required.

I would be nice to have scripts for different data sources that convert the data to a standardized format (gpkg with S57 layers and tagging) and then one single QGIS template could be used for the actual rendering. The thing is, that adjusting the appearance in QGIS interactively is much faster and more comfortable, especially because there is quite a number of layers to handle.

[quantenschaum]

example problem: topmarks are in a separate layer and detached from their corresponding buoys/beacons. The spec mentions the LNAM_REFS field that links objects from different layers. But using it in QGIS is somewhat uncomfortable, and more important: it does not contain any values in the actual data!

[free-x]

rws.qgs was my first thought, I have already tried it. swap “vwm” with “aton”
Result is somehow not quite usable

[quantenschaum]

Hmm. Why? - But it is actually not bad, it needs some tuning. As mentioned, the actual data does often not follow the S57 spec (your data and/or the data by RWS which I created the styling for). What datasource are you using? o-charts?

[quantenschaum]

And you get a lot of duplicate objects. The RWS data only contain one data layer for all zoom levels, whereas the BSH provides them in 6 different usage bands. These usage band havt be assigned to zoom levels, which is done in the BSH QGIS file.

[free-x]

I am aware of the difference between ENC and iENC. ENCs in the name have a corresponding note in the third position.

1 - Overview - For route planning and oceanic crossing.
2 - General - For navigating oceans, approaching coasts and route planning.
3 - Coastal - For navigating along the coastline, either inshore or offshore.
4 - Approach - Navigating the approaches to ports or mayor channels or through intricate or congested waters.
5 - Harbour - Navigating within ports, harbours, bays, rivers and canals, for anchorages.
6 - Berthing - Detailed data to aid berthing.

iENC is always 7 ( fixed scale 2000 ?)

What datasource are you using? o-charts?

s-63

[quantenschaum]

OK. How where did you the S63 data from and how did you convert them into a GPKG? I'd like to have a look into this data, to see how it differs from what I have seen so far.

[free-x]

You can decrypt s-63 using PERMIT and then convert with ogr2ogr

sample

[quantenschaum]

Did you remove/filter any objects like contour lines?

[free-x]

In sample are following DEPCNT included
DE4OS10F-DEPCNT.geojson / Fehmarnsund
DE5OSBSK-DEPCNT.geojson / Burgstaake
DE6OSHLH-DEPCNT.geojson / Heihafen Stadthafen

[quantenschaum]

Well, this does not tell me, which contours are actually in the dataset. It only contains 2m, 5m, 10m and some are missing where there are contour area boundaries. No big problem, I am just wondering why?

[free-x]

I see also 15m

[quantenschaum]

Yes 15m, too. But a lot of contours are missing.

[quantenschaum]

Only the red contours are in the dataset.

[quantenschaum]

After some tuning, it is actually quite OK. The dataset itself is missing information (contours, soundings, etc.).

https://github.com/quantenschaum/mapping/blob/master/qgis/s57.qgs