Charts II: NGA - 2700 Charts

[cagney]

Originally Posted by recri

.....
So what do you do when you find your chart plotter locating you on this chart? Will the plotter let you enter a locally determined correction? Or do you have to apply the correction in your head?

28201 is a well known problem chart. It should really be labeled "Don't use with GPS". GPS is the problem here. I can't see much problem using the chart otherwise. Anyone trusting GPS for navigation among coral reefs, such as the entrance to Puerto Morelos, is out of their minds. Only trust your eyes, do eyeball navigation, and use the chart as a guide only.
I think that phiggins tried to overlay GE, and correct the chart that way, but gave up.

Thomas

[cagney]

28201 was not correctly entered in the database.
The note on the chart explicitly says that the chart cannot be corrected to agree with WGS. Please don't enter private, guesstimated, corrections.
It is essential that the data in the database produces a chart the way the cartographer intended, not the way we personally think it should be.
But we should put a note in the comment field about the very poor result in relation to WGS.
Once the database is complete, it can be used in many ways, for different projects.
While waiting for an edition based on a modern survey, someone could for example produce a new chart based on the shoreline and reef-outline from GE and the depth figures from 28201.

Once again - Please enter the values found on the charts into the database -nothing else.

Thomas

Thomas

[HappySeagull]

Originally Posted by wdkester

I'm looking at 26030 which has a datum note that reads (approximately) "Positions obtained by satellite nav systems (WGS72) must be moved by .06 min S and .01 min W"

Does mean we enter .08 SOUTHWARD .01 WESTWARD ?

his char has a corners problem so isn't of immediate concern. I just want to know if Irun across another like it.

try this
http://www.cruisersforum.com/forums/...tml#post768652

[bcn]

Originally Posted by wdkester

I'm looking at 26030 which has a datum note that reads (approximately) "Positions obtained by satellite nav systems (WGS72) must be moved by .06 min S and .01 min W"

Does mean we enter .08 SOUTHWARD .01 WESTWARD ?

his char has a corners problem so isn't of immediate concern. I just want to know if Irun across another like it.

Yes, this is the right way. The program will take care to treat that information correctly. There are charts as well with correction in seconds. Divide by 60 and enter the information as minutes then.

Hubert

[HappySeagull]

Originally Posted by recri

Thanks Happy --

My project to reconcile Puerto Morelos and Yucatan Channel ended when I discovered that the two nav markers which appear on both charts are not the same distance apart. So how to resolve? Pick one? Split the difference? Rescale the chart? It was interesting to work through the puzzle and find no sound solution.

I see the NTM for this chart has 5 new lights, but no mention of any correction to the chart position. Plotting the lights on the chart, as the chart is labeled, puts them well offshore in deep water.

So what do you do when you find your chart plotter locating you on this chart? Will the plotter let you enter a locally determined correction? Or do you have to apply the correction in your head?

Welcome to the fun!why not download the image,and make your own personal Corrected KAP chart!

Try using just the NTM"CORRECT" positions of the lights as calibration points(forget the corners,or shift em to agree) in MapCal_2 and imgkap and see how it goes...you'll have to sift through the earlier posts and the imgkap thread etc etc and figure out whatll be your favourite method if you haven't made these before..
it'd be way too dangerous to be "remembering"about various charts unless you crack em into headers and images,scribble on the image and rebuild.Imgkap can do this easily.That's how I'd do it in any case(scribbling on the image with a picture editor)so the original stuff didn't confuse.

[recri]

Agreed, I'll confine any future chart surgery experiments to my own machine.

[paulr]

There's a numbering problem in the database or interface starting at chart 92170.

This chart is missing in the calibration interface and the subsequent files are all off by one, i.e. what the interface thinks is chart 92170 is actually 92190, and so on. It comes right again at chart 92220.

The problem doesn't appear to be in the uploads, since 92170 is correct on the ftp server, but the thumbnail http://opencpn.info/nga/chartimages/.../92170_zl0.jpg is wrong, as are the image that opens under "General chart and header data", and the corner cutouts.

Paul

[wdkester]

Originally Posted by recri

I processed the 489 charts in the Rectangular, Doable category about an hour ago.

I discarded all the charts with less than 5 digits in their chart numbers, all the charts which weren't MERCATOR, all the charts with missing corner coordinates, and computed the nautical miles/pixel ratio as described above.

There were 433 charts which were "square" to better than 9 parts per thousand, the ratio was greater than 0.991 and less than 1.009. The mean for those charts was 1.00436 with standard deviation of 0.00208.

I inspected all the charts outside the cutoff.

These 7 charts, all high latitude large scale, appear to be correct:

29002 29025 35000 37000 42046 43015 44000

These charts all have errors in corner coordinates:

24400 25400 44069 51002 51400 51460 52054 52262 53244 53281 71005 74002 81151

There are several places where 0, 6, and 9 got confused, there's one with the same longitude for east and west, and a couple where east ended up west of west or south north of north.

recri,

I'm hopeing that you run this error detection program at least weekly if not daily

Please

[nohal]

Originally Posted by paulr

There's a numbering problem in the database or interface starting at chart 92170.

This chart is missing in the calibration interface and the subsequent files are all off by one, i.e. what the interface thinks is chart 92170 is actually 92190, and so on. It comes right again at chart 92220.

The problem doesn't appear to be in the uploads, since 92170 is correct on the ftp server, but the thumbnail http://opencpn.info/nga/chartimages/.../92170_zl0.jpg is wrong, as are the image that opens under "General chart and header data", and the corner cutouts.

Paul

Thanks for the report...
All of them were uploaded with wrong names. Fixed.

Pavel

[recri]

Thanks for the vote of confidence.

I've finally figured out what I'm doing. The right way works much better than my last formula and it's much easier to explain. The rule now computes the aspect ratio (height/width) of the chart pixel area, the aspect ratio of the Mercator projection of the chart corners, takes the ratio of these two ratios, and expects to find a value close to 1.

The mean over 525 "Rectangular, doable in phase 1" charts with (0.99 < ratio < 1.01) is 1.005 +/- 0.002, the only charts which fail at this threshold are:

707 26050 43365 53244 26068 26140

and I just fixed them while verifying their problems, all wrong corner coordinates.

-----

I added in the Rectangular with insets classes and tightened the limits and fixed the following charts:

37326 44081 22052 25566 26066 28143 37228 37230 43374 44053 44075 52142 61270

In addition to incorrect coordinate entry, there were:
truncated fractions of a degree,
inset corner used for coordinates or REF points,
border frame used for REF points,
SW and NE corners were different charts, one looked like an inset
and one case where the corner coordinate degrees were obscured

This last run examined 705 charts, 634 passed, 11 failed, the mean and sd haven't changed.

The following charts were not checked because they aren't marked as Mercator in the database. I looked at one, and it was Mercator so I fixed it. Please check these, I'm finished fixing for the night:

14221 17421 18419 21141 21161 21621 24055 24444 24504 24511 26081 29123 29180 29200 29281 29282 29323 43064 43344 43363 43371 43377 43378 43386 44037 44068 54404 61551 62437 62442 62452 62463 62595 81127 81303 81327 81329 81511 81563 81565 81576 81616 81737 81791 83207 83495

The following charts were not checked because they're marked as Doable but they're missing REF points or corner coordinates. They should be finished if they're doable or marked undoable:

37166 51261 51461 51500 51637 52060 52062 52082 52085 52120 52164 52170 52281 53223

What I've claimed to "fix" should still be checked.

[recri]

Originally Posted by recri

Thanks for the vote of confidence.

That was directed to wdkester.

The following charts were not checked because they aren't marked as Mercator in the database. I looked at one, and it was Mercator so I fixed it. Please check these, I'm finished fixing for the night:

14221 17421 18419 21141 21161 21621 24055 24444 24504 24511 26081 29123 29180 29200 29281 29282 29323 43064 43344 43363 43371 43377 43378 43386 44037 44068 54404 61551 62437 62442 62452 62463 62595 81127 81303 81327 81329 81511 81563 81565 81576 81616 81737 81791 83207 83495

These charts have a blank projection, they're not Mercator because they're not marked as anything:

24444 24504 26081 29123 29180 29200 29281 29282 54404 62437 62442 62452 62463 62595

The rest claim to be something other than Mercator.

[paulr]

Database correction required:
the scale on chart 44069 doesn't correspond to the scale printed on the chart -
137000 versus 75000


Would you prefer such problems to be reported here, or as comments to the chart, or both?

Paul

[nohal]

Paul...
much better as comments to the charts, no need to post them here. The thread gets long and it's starting to be hard to catch everything important.

Thanks

Pavel

[recri]

Four new errors caught since last night:

chart ratio - problem
----------------------------
26145 0.00034 - 54 degree error in SW corner lng
26181 -0.36057 - SW corner lat and lng swapped
26219 1.86287 - 2 degree error in SW corner lat
97412 1.10891 - 1' error in SW corner lat

I have left these for the authors to fix.

But don't beat yourself up about it, I have found it hard to spot the errors even when I have reason to suspect they're there. And in one case I went back to get more details, and it was still hard to spot the error even though I knew exactly where it was supposed to be.

These charts have no projection specified:

no projection 29123 29180 29200 29281 29282 54404 62437 62442 62452 62595

These charts have missing corner coordinates or REF points:

missingcorner 37166 51261 51461 51500 51637 52060 52062 52082 52085 52120 52164 52170 52281 53223

[recri]

Keep up the good work!

A new summary format so I won't need to edit the output so much.
------
processed: 726 charts
accepted: 673 charts, mean ratio 1.004827280832095 stddev 0.0018468313130046647

problem: chart 26125 ratio 0.92272 - 2' error in SW corner lng
problem: chart 97412 ratio 1.10891 - 1' error in SW corner lat

no projection: 29281 29282 54404 62437 62442 62452 62595

missingcorner: 37166 51261 51461 51500 51637 52060 52062 52082 52085 52120 52164 52170
------
The script takes about 10 seconds to run. It fetches the current status from opencpn.xtr.cz for the doable and partially doable status charts, parses the values out of the html, computes, and prints the result. Then I check the problems by hand and add a description, or copy the description from the previous report.