Cruisers Forum
 


Reply
  This discussion is proudly sponsored by:
Please support our sponsors and let them know you heard about their products on Cruisers Forums. Advertise Here
 
Thread Tools Search this Thread Rate Thread Display Modes
Old 12-08-2009, 05:37   #1
Registered User

Join Date: Aug 2009
Posts: 39
CM93 Issues

Quote:
Originally Posted by Zoef View Post
Quote:
Originally Posted by GDD View Post
Hi all,

The ideas from the more technical members here regarding updating cm93 are all very interesting. But what about the simpler task of recording known issues with the different versions of CM93?
To avoid muddling this openCPN forum with (admittedly important) issues related to just one of the chart formats openCPN can handle, wouldn't it be better to start a specific CM93 thread and keep this forum concentrated on the development (features, performance, documentation, testing, release planning and you name it) of openCPN?

Klaas
Good idea - done.

"But what about the simpler task of recording known issues with the different versions of CM93?"

"But first, there would be a need to check that each version being compared isn't itself corrupted. For this I guess there would be a need to establish exact filesizes and CRC's for each commonly distributed version."


To illustrate my point, I've now downloaded 2 different 2009 folios - one from a well-known source in Norway, and the other which comes in 10 parts (source omitted, so as not to breach the Forums' rules).

The dates of the charts of both folios are mixed (which is not unusual), but large numbers of files are different in both size and date, and a comprehensive comparison is indicated (imho).

A few significant differences in filesizes I instantly spotted are:

____________________ Norway _ 'Ten Part' _File Date

02100420\E\02270432__76,338___0____(11/05/2009) 03300000\D\03750039__missing__2,724_(01/09/2008)
03300000\E\03770044__missing__6,010_(01/09/2008)
03300000\E\03780044__missing__2,171_(01/09/2008)
03900000\E\03990048__missing__6,282_(31/03/2009)

Sorry about the crude attempt at a table - couldn't see how to switch off the HTML.
Colin
sae140 is offline   Reply With Quote
Old 12-08-2009, 18:25   #2
Obsfucator, Second Class
 
dacust's Avatar

Join Date: Feb 2008
Location: Southeast USA.
Boat: 1982 Sea Ray SRV360
Posts: 1,745
How about doing a dir to text files for each set. You can compare them from there. If you don't have a way to do it easily, let me have the text files and I'll do an SQL to pick out the ones that don't match up. That'd be a start anyway.

That's what I did with the Dutch ENCs, but that had 15-20 different directories with sometimes 45 duplicates. This will be easier. But, once we have the results, seeing what the individual file differences are will be harder.

For the ENCs I used a simple ENC viewer, fired up multiple windows and looked at the duplicates side-by-side. Is there a way to do this with the CM93s?

-dan
dacust is offline   Reply With Quote
Old 13-08-2009, 00:38   #3
Registered User

Join Date: Aug 2009
Posts: 39
Hi Dan
I usually use Total Commander's 'compare' facility, which is fine for comparing just one directory at a time , but something a bit more comprehensive is needed here. Plenty of programs around which will identify duplicates, but what we want is the opposite ! (I suppose duplicates could be removed, to see what's left ?)

Ok - I've just done a "Dir /S". Before you commit yourself - here are some figures for one of the 2009 world folios:

Folio size = 1,407,179,055 bytes
with 21,400 files in 2,768 dirs/sub-dirs

Now there may possibly be less than a hundred or so file differences, but trawling through that lot to spot 'em presents something of a challenge !

The Dir listings are 1.4Mb each, even in plain ascii - if you want 'em, best to PM me with an email address which will accept attachments ?

Colin
sae140 is offline   Reply With Quote
Old 13-08-2009, 18:55   #4
Obsfucator, Second Class
 
dacust's Avatar

Join Date: Feb 2008
Location: Southeast USA.
Boat: 1982 Sea Ray SRV360
Posts: 1,745
Quote:
Originally Posted by sae140 View Post
The Dir listings are 1.4Mb each, even in plain ascii - if you want 'em, best to PM me with an email address which will accept attachments ?

Colin
Actually my webmaster@dacust.com address is public. Send it there and put cm93 somewhere in the subject so I don't delete it along with all the other spam.

-dan
dacust is offline   Reply With Quote
Old 14-08-2009, 04:08   #5
Registered User

Join Date: Aug 2009
Posts: 39
Hi Dan

You're welcome to those listings of course, but I can save you some grief, for this morning I've been using "MoleskinSoft Clone Remover" - 'stepwise' - firstly to remove true file clones, then those with the same name and size (albeit with byte-by-byte differences), and finally to remove those files with just the same names.
The following files were left:

Those files completely missing in the 'Norway' 05/2009 folio are:

Code:
03750039.D   2,724  01/09/08 
03770044.E   6,010  01/09/08 
03780044.E   2,171  01/09/08 
03600096.E   2,976  31/03/09 
03610097.E  20,667  31/03/09 
03620096.E  11,961  31/03/09 
03620097.E   6,713  31/03/09 
03600097.F  11,687  31/03/09 
03610096.F  21,301  31/03/09 
03600960.B  24,678  31/03/09 
03600990.B  95,991  31/03/09 
03870993.D   8,243  31/03/09 
03870996.D   8,408  31/03/09 
03870999.D     956  31/03/09 
03870995.F   7,753  31/03/09 
03870996.F  26,712  31/03/09 
03850994.G  29,817  31/03/09 
04350012.D  28,430  31/03/09 
04350015.D  15,668  31/03/09 
03990048.E   6,282  31/03/09 
03900960.B   1,642  31/03/09 
03900990.B   4,152  31/03/09
And those completely missing in the '10-part' 03/2009 folio are:

Code:
02270432.E  76,338  11/05/09
03780972.C   6,064  11/05/09
03860993.E   4,267  11/05/09
03860994.E   3,279  11/05/09
03811005.F  23,900  11/05/09
03870995.G  19,982  11/05/09
03870996.G  58,974  11/05/09
03900972.C   2,021  11/05/09
03900984.C   7,696  11/05/09
04520012.E     786  11/05/09
In addition to these, there are 896 files which have the same name and size, but which are 'byte-different' - which seems to me to be quite a lot for 2 folios just 6 weeks apart.

Hope the above is of use to somebody.

Colin
sae140 is offline   Reply With Quote
Old 14-08-2009, 15:11   #6
GDD
Registered User

Join Date: Jul 2009
Posts: 65
cmwfw

Hi Everyone

Great discussion!

I have used a directory compare tool (called 'Beyond Compare') to compare my 2006 edition to the 2008 world edition. I immediately noticed that the two were very different in that, besides each having files not present in the other, the sizes of the files in each edition were ussaly different.

Since I really have no way to read the files and compare their content, I switched to a more qualitative analysis using the old CMFwFW.exe utiltiy program. This program came with some of the earliest versions of CMAMPECS and allows for easy assesment of where areas are covered and to what level of detail. I"m happyo send to anyone who wants it.

Since a picture is worth a thousand words, I've attached a picture that shows the D level coverage right around Vancouver Island. The green cross-hatched area shows where coverage at E level is available. Notice the hatching where there is no E coverage? No need to zoom in and out a bunch in OPenCPn to see whats happening. Very useful!

I have used this program to compare the coverage in BC and I have found only one error in the 2009 edition, visible in the first image but also shown zoomed in in the second image.

Also, I have done some checking at places where I know the shoreline has changed due to construction, and sure enough the 2009 edition really does contain updates.

I have also 'swapped in' files from the 2006 edition to fix the missing coverage in the 2009 edition. Though this is a bit tedious and i've only done it at this one spot, it seems like the most promising way to create hybrid 'best' copy of cm93. I now have a 2009 edition where one small area is from 2006 ,making the coverage of BC complete.
Attached Thumbnails
Click image for larger version

Name:	cm932009VancouverIsland.jpg
Views:	385
Size:	469.0 KB
ID:	9351   Click image for larger version

Name:	cm932009Desoltation sound.jpg
Views:	354
Size:	473.8 KB
ID:	9352  

GDD is offline   Reply With Quote
Old 14-08-2009, 16:29   #7
Registered User
 
Viking Sailor's Avatar

Join Date: Nov 2006
Location: San Francisco Bay
Boat: Fantasia 35
Posts: 1,251
GDD,

How about trying to copy the 2009 charts on top of the 2006 charts. I'm thinking that maybe the 2009 charts are an update set. Just make a new folder, copy the 2006 charts into this folder, then copy the 2009 charts into the same folder with the option to "replace all" selected. I don't have the charts to do this myself. I beleve that there are chart sets for 2003, 2006, 2008, March 2009, and May 2009 out their on the web. It might be interesting to see what the composit chart set would look like.

Paul
Viking Sailor is offline   Reply With Quote
Old 14-08-2009, 17:41   #8
GDD
Registered User

Join Date: Jul 2009
Posts: 65
Copying/Overwriting

Paul,

I'm afraid its not so simple. Lets see if I can explain my understanding of how this works..

All cm93 chart files have a name where the first part is a number describing in some weird number system the charts position. The second part of the file name (the extension) is a letter that describes the charts scale/detail level.

The folder names use the same number system as the file names. Folders group files together by area. The first half of the number is the latitude which I believe counts up north from the south pole to the north pole. The second have of the number is the longitude which counts upwards going east around the world from Greenwich/prime merdian.

Also, sometimes a file is split into parts with a letter at the start of the name describing which part it is. These show up as A420704.E , B420704.3 etc.

So, the file 420704.E is an 'E' level, which means a 40 000:1 (or so...they vary) chart that is about 42/50 of the way from the south pole to the north pole and 704/1080 of the way east around the world from Greenwich. Or something close to that. I use the opencpn log file to figure out exactly which file i'm looking at in OpenCpn.

The catch with pasting/overwritng files is that the chart outlines are diffferent in different editions, hence the file names for the same point on earth can also be different ( I think).

What makes pasting/overwriting not work is that while on the one hand you end up with directory that contains all the files from both editions, on the other hand a viewer program doesn't always no what to do with this directory. This is because the different editions break the source maps up into parts differently. That is, the chart outlines are different. So you can actually create MORE errors by mixing charts from different editions.

I believe this is the case because when I swapped 2006 files into the 2009 edition to fix that error in described my last post, I actually had to swap 3-4 E level files to make it work. If I only swapped the one file right at the missing spot I would just 'move the error around' because I was combining jigsaw shaped pieces that didn't want to go together.

This is just a theory though, and a badly explained on at that.

Hope it helps.

On another note, the other major problem with the 2009 edition for western north America is that coverage of western Mexico seeme to be totally gone.

Intrestingly, this is an area renowned for having charts with funny/flawed datums. The charts that have 'disappeared' may have been removed because they were inaccurate. The coverage for this area in cm932006 is a couple of hundred yards off google earth.

Does any one else know of areas where chart datums were bad, and now the maps are missing (or fixed, as that would be even better)?

gdd
GDD is offline   Reply With Quote
Old 14-08-2009, 19:17   #9
Registered User

Join Date: Aug 2009
Location: between the devil and the deep blue sea
Boat: a sailing boat
Posts: 20,420
Huh, very interesting - since CM93 stands for C-map 93 and there was only one update - in 96! So what all those 2009 or 2006 'editions' are, god only knows. And why not ask at source? BTW why should CM be more accurate that Google ??? After all - if you look at the files they clearly state what datum they are and VERY many are not pre-WGS. A good software will adjust, but if your plotting software sucks then you may make a stupid mistake. A point to remember though - when within soundings a good sailor will go by the pilot rather than stick to whatever the nav gadgets onboard show, unless there is a good reason to to do otherwise.
barnakiel is offline   Reply With Quote
Old 15-08-2009, 04:29   #10
Registered User

Join Date: Aug 2009
Posts: 39
Quote:
Originally Posted by barnakiel View Post
Huh, very interesting - since CM93 stands for C-map 93 and there was only one update - in 96! So what all those 2009 or 2006 'editions' are, god only knows.
My earliest CM93/2 folio is dated 1998, which was allegedly bundled with a 1999 Latvian demo CD of the Russian (Morintech.ru) DKart Navigator 5.10 - or so the story goes ....

Some commercial operators and navies stayed with CM93/2 long after the introduction of CM93-3, so I guess there must have been a contractual obligation to support the earlier format for some years. I've heard talk that the Norweigian coastguard (at least) still uses CM93/2.

I became a tad obsessed with DKart Navigator software and collected numerous demo versions (including DKart Nav 810 - for submarines !!). Certainly the DKart 2000 series (CM93/2 & S57) were being actively promoted at least up until 2004.

Both the 2009 folios I have display the fully operational Wind Farms off Skegness (my nearest offshore development), albeit marked as 'under construction'. But when I try reading the information about them (using CMAPECS 4), the system locks-up - so there's a naughty error somewhere.

Another 'funny' can be seen at 0.57'113"S, 89.35'933"W - SW of the Isle of San Cristobal in the Galapagos. If you look around, you'll see some tidal arrows - unrelated to date or time - which look like an amateur annotation to me.
sae140 is offline   Reply With Quote
Old 15-08-2009, 05:14   #11
Registered User

Join Date: Dec 2005
Location: Helsingborg
Boat: Dufour 35
Posts: 3,891
I have found the following explanation on the internet:

Quote:
CMAP File naming explained:
The System for c-map is very simple. All folders contein segments of 20 x 20 degrees.
00300000 means 80S 0E, 00300060 means 80S 20E, 00900120 means 60S 40E. The first 4 digits
is south to north 0150=40S +20 north, 0210=20S, 0270=0S (equator and 20 degrees north),
0330=20N, 0390=40N. The last 4 digits is East from 0 meridian.. 0060=20E, 0120=40E,
0180=60E, 0240=80E, and so on 0480=160E, 0540=180W, 0600=160W and so on..

The folders in the next level is a - z, where z is scale 40x40 degrees,
a is20x20 degrees, b is 10x10 degrees, down to f and g which is scale of
20x20 arcminutes, which is harbour level.
There are so many issues with CM93 that I would say: Don't trust it unless you have compared the coverage of a given area with a reliable source. One big issue is the lack of worldwide B level coverage (at least in the versions I have access to). Just some examples. Conway Reef or Theva-i-ra on 21 44 S 174 37 E, a small reef SSW of Fiji, 2 m high, doesn't show up at all. Same thing for a few shallow areas in the Coral Sea.
cagney is offline   Reply With Quote
Old 16-08-2009, 04:58   #12
Registered User

Join Date: Aug 2009
Posts: 39
Quote:
Originally Posted by GDD View Post
I switched to a more qualitative analysis using the old CMFwFW.exe utiltiy program. This program came with some of the earliest versions of CMAMPECS and allows for easy assesment of where areas are covered and to what level of detail.
Hi - did you receive my PM and email about CMWFW ?

'best, Colin
sae140 is offline   Reply With Quote
Old 16-08-2009, 05:22   #13
Registered User

Join Date: Aug 2009
Posts: 39
Quote:
Originally Posted by cagney View Post
There are so many issues with CM93 that I would say: Don't trust it unless you have compared the coverage of a given area with a reliable source.
That of course begs the question of what constitutes a reliable source ?
Many would say 'British Admiralty charts' without a moment's hesitation - but even these were originally drawn up by the use of lead line soundings back in the days of sail. Only those areas used by commercial traffic have since been surveyed by modern methods - and even then some obstacles have been missed.
The UK coastline is one of the most intensively surveyed in the world, yet there was an interesting case on British TV recently in one episode of a series called 'Coast', in which a heavily-laden vessel had briefly grounded in a channel which should have been clear for heavy shipping. A survey vessel was duly dispatched and an underwater ridge was identified, with one previously unknown peak rising very close to MLWS. There was nothing marked on the paper charts, so therefore nothing on any of the electronic charts based on them.

I would say CM93, along with other electronic and paper charts can pretty much be trusted by shallow-draft leisure vessels in those areas frequented by commercial traffic.
Charts (of any kind) for off-the-beat places, like south sea islands and indeed anywhere where coral heads grow, should be treated with a huge margin of distrust. You may be using modern charts and equipment, but the surveys on which the chart data has been based may be a century or more old, and all sorts of hazards might have been missed.
sae140 is offline   Reply With Quote
Old 16-08-2009, 07:44   #14
Registered User
 
blubaju's Avatar

Join Date: Mar 2008
Location: where my little boat is ;-) now Philippines
Boat: Catamaran Schionning Wilderness 1320, built myself
Posts: 475
I have a new Navionics Far East Asia (CF31/XG; Indian Ocean & China Sea) and CM93 of 2004 and 2009. The elder CM93 looks and shows quite different details than the 2009 version. On the other hand, the 2009 version is almost identical with the Navionics data. Could it be that the latest update of CM93 used the same database from the Hydrographic offices as Navionis?

Regarding displaying old CM93 with newer Software, ever tied to read new Excell or Works data with an old office?

The ConwY REEF IS SURPRISINGLY "BLUED-OUT" on Google, but the Name Ceva-I-Ra pops up. This does not mean it does not exist as the Scarborough reef does exist and in the more detailes levels it is "blued out" too, but it does exist in CM93 and Navionincs. Anyone knows wether it exists or not?
blubaju is offline   Reply With Quote
Old 16-08-2009, 09:24   #15
Obsfucator, Second Class
 
dacust's Avatar

Join Date: Feb 2008
Location: Southeast USA.
Boat: 1982 Sea Ray SRV360
Posts: 1,745
Quote:
Originally Posted by blubaju View Post
Regarding displaying old CM93 with newer Software, ever tied to read new Excell or Works data with an old office?
I think you got your comparison backwards. Ever tried to open an old spreadsheet with newer software? It almost always works.

I could see that an old version of nav software might not be able to display newer charts, but would not find it surprising at all that it could read older charts just fine.

-dan
dacust is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
CM93 International Charts Jeff Millar Navigation 39 24-11-2012 13:17
CM93 C-Map cooper Navigation 18 17-07-2011 17:43
CM93 chart program riginos Navigation 6 20-04-2008 07:42
Where do I get CM93? ude123 Navigation 1 12-06-2007 02:53

Advertise Here


All times are GMT -7. The time now is 01:09.


Google+
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Social Knowledge Networks
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2024, vBulletin Solutions, Inc.

ShowCase vBulletin Plugins by Drive Thru Online, Inc.