Charts II: NGA - 2700 Charts

[HappySeagull]

yes,800 mb is a red herring ...but it's still going to be real work to apply them correctly,however it's done.

[cagney]

Originally Posted by LeaseOnLife

Could you describe how you do that? The ruby script just fails and does not create an image... Maybe use the wiki for now.

https://github.com/ngacharts/ngacharts/wiki

1 Download all available tiles. I use wget.
2 Run Allan Pedas perlscript. this will give you an idea about missing tiles. Replace all missing tiles with a copy of an all white tile, with correct names.
3 Run the perl script, and a chart pic will be produced.

I'm sure the ruby script can be changed to do the same thing, but why reinvent the wheel.....?
25525 had a lot missing, 61601 and 52047 just had one tile missing in less important areas.
I'm now working on 54334 and will then continue down the list, while checking for updates and uploads.

Thomas

[Netsurfer]

Originally Posted by cagney

2 Run Allan Pedas perlscript. this will give you an idea about missing tiles. Replace all missing tiles with a copy of an all white tile, with correct names.

Replacing the missing tiles with white tiles isn't ideal to me as it makes hard to recognize that there are missing tiles. I left them black!

3 Run the perl script, and a chart pic will be produced.

I'm sure the ruby script can be changed to do the same thing, but why reinvent the wheel.....?

I am using a quick'n dirty PHP script for this - no need to add the missing tiles.
So if anyone is interested in, drop me a note.

Gunther

[LeaseOnLife]

Missing Tiles:

I am checking on 61601, thanks Thomas for adding the missing tile!

Black vs white: As soon as we miss some significant info in the tile, I am all for black. White is fine if there is no info missing, like in 61601.

But, when looking at 61601 using the NGA online chart viewer, the tile is not missing there!

http://www.charts.noaa.gov/NGAViewer/61601.shtml

Not sure about the other charts, but the zoomify seems to be able to gather all tiles. Weird, can anyone explain? Is this the same for all other charts with missing tiles?

Here is the portion of 61601 with the missing tile:

[nohal]

Dirk,
isn't zoomify leaving a blurred zoomed in version from the previous zoom level there? Looks a bit like it is.
It seems this could be a way to reconstruct the missing parts of the charts - the quality will be a bit worse in those areas, but still...

Pavel

[LeaseOnLife]

Originally Posted by nohal

Dirk,
isn't zoomify leaving a blurred zoomed in version from the previous zoom level there? Looks a bit like it is.
It seems this could be a way to reconstruct the missing parts of the charts - the quality will be a bit worse in those areas, but still...

Pavel

Don't think so:
First picture is a screenshot of zoomify, no missing tile, and crisp.
The second picture is from our download-and-stitch, the oval is hand-drawn to show the missing tile.

Yes we could reconstruct, but I am wondering how zoomify gets the tile and we don't!


Edit: Pavel, maybe I misunderstood. Are you saying, there is another set of jpg-tiles, with a different zoom-level?

[nohal]

Dirk,
yep, there are different tilesets for the different zoom levels, we are stitching level 7 - the best quality and the tiles are really missing there.
On the attached screenshot from the NOAA website, it's visible, that the missing part is there, but the quality is different, from that my conclusion is that zoomify uses smaller scale tile and resizes it itself.
So if we download level 6, we can do the same. Will try.

Pavel

[LeaseOnLife]

Cool, your eyes are better than mine. To me it looked the same at first, but your screenshot shows the difference nicely! I guess we should then be able to create/stitch the last few charts after all... Who could modify one of the scripts to do so?

Dirk

[nohal]

The python script can do that for us already. Actually it gives a hint to do exactly that in the help

-z ZOOMLEVEL  zoomlevel to grab image at (can be useful if some of a higher
                zoomlevel is corrupted or missing)

There are really just a few charts, so I will manually do:

• Download level 6 with the python script
• Resize to the size of level 7
• Open both in Gimp and put them in two layers (level 7 on top)
• Make a hole for the missing tiles in the top layer.
• Flatten image
• Save as JPG
• Replace the one we have now

Should work, already downloading 61601.

Pavel

[nohal]

There are missing tiles for almost all of the problematic charts in ZL 6 and 5, so I'm finally going to learn a bit of ruby instead of wasting more time in front of stalled Gimp eating 2 gigs of memory

Pavel

[r.fairman]

Notice to Mariners

I am automating the application of NTM's both as a gpx file and hopefully as a graphic overlay to a graphic for reconverting to kap. I need the actual stored names of the icons used by OpenCPN The displayed names of the icons in Route Manager does not match the stored names that is used for GPX import and export.

regards

Richard

[nohal]

Richard...
Hope it's understandable:
https://github.com/nohal/OpenCPN/blo...eman.cpp#L1161
If you want to also use some new icons, the user icons are named according to the file name without .xpm extension

Pavel

[wdkester]

My graphics/picture editor (ACDSee) doesn't seem to be able to open large jpeg files. Can someone recommend a free program that can create a tinn and a lng file from jpeg?

[LeaseOnLife]

Originally Posted by wdkester

My graphics/picture editor (ACDSee) doesn't seem to be able to open large jpeg files. Can someone recommend a free program that can create a tinn and a lng file from jpeg?

I am guessing here, do you mean "create a tiff and a png file from jpeg"?

On XP, I use the command line tool "nconvert" to reduce size and convert to png:

nconvert -out png -resize 50% 50% filename.jpg

makes a quartersize png file, then I can view it in xnview, my favorite graphics viewer.

Dirk

[r.fairman]

Install ImageMagick and use the convert command at the command prompt

convert NAME.jpg NAME.png

regards

Richard