Beta Test / Technical

[bdbcat]

Fabbian....

Well, maybe....

One of the advantages of S57/52 is that many of the "hard" objects have their least depth encoded as attributes. This value can be compared to the Mariner's specified "Safety Depth", dynamically. If there is a potential conflict, the system should symbolize the object as "dangerous".

This is the real payoff of vector charts. With the proper interpretation in an ECDIS or ECS, they are not dumb, like raster charts.

So, the S52 rules tell us how to symbolize these "hard" things like obstructions and wrecks. In theory we could see a foul ground, whose least depth is known, and is always greater than the Safety Depth. In that case, the SY(FOULGND1) may be appropriate. I have never seen such a case. Generally, obstructions encoded as "foul" have an unknown least depth. Being "always submerged" is not enough info to navigate safely over the obstruction. Thus, it is dangerous for ANY safety depth.

This is all very conditionally complicated. The S52 CSPs are regularly updated for anecdotally discovered cases. Often, the onus goes back to the cartographer, to better (i.e. more usefully) encode the S57 attributes.

Anyway, that's what I know about this.

Thanks
Dave

[fgd3]

Compare the raster (18424) and vector (US5WA45M) charts of the Bellingham area. On the raster chart I can tell immediately that there is a foul area at 48 45.7377 N 122 31.0300 W, in an area that's 3 to 4 feet deep. The K-31 symbol tells me it's not dangerous to navigation but should be avoided for anchoring or trawling.

The vector chart shows something at 048 45.7703 N 122 31.0228 W. I have to do an object query to learn that it's a foul area, depth unknown. Since we don't know the least depth it should be categorized as dangerous and symbolized as K-32.

Clearly one of the charts is wrong with regard to the least depth of the foul area. The raster chart is probably correct but since we don't know the least depth of the foul area it makes sense to err on the side of caution and treat the foul area as dangerous. I'm sure that's what was done when the raster chart was converted to a vector chart.

Displaying either K-31 or K-32 on the vector chart seems better to me than displaying an X in a red circle. Is that SY(ISODGR01)? If so, I think it's being misused. It's supposed to represent "isolated danger of depth less than the safety contour" (page166, IHO PUBLICATION S-52 APPENDIX 2 ADDENDUM TO ANNEX A, PART I, USERS´ MANUAL EDITION 3.4 (2008)). No matter how I set my safety depth (even to a negative number) I can't make the symbol change.

It doesn't seem like a good idea for that symbol to replace other symbols just because the object's depth is less than the user's safety depth. Most of us are more concerned with hazards that are less than our safety depth than the ones we can safely sail over. Losing information about the objects that most concern us seems like a bad idea. Maybe that symbol is meant to be displayed along with the object's regular symbol. It would make more sense to provide an additional warning about depth without removing any information.

Wouldn't it make for better charts if, in the absence of a clear S-52 symbol specification you used the closest approximation from Chart No. 1, informed by the raster chart of the same area? Once the IHO publishes a clear symbol specification you could make the appropriate change in the next release.

Fabbian

[bdbcat]

Fabbian....

Sorry, what is K32? I cannot find it in symbol libraries....

Dave

[bdbcat]

Fabbian...

Looked at this again.

I'm afraid the interaction between S57 and S52 is not as intuitive as we might like it to be. Add to that the errors in S57 encoding, and we can get mush...

At the particular point you mention (NW of the marina in Bellingham), the object is encoded as a POINT object OBSTRN, with attribute CATOBS6. This seems wrong to me. CATOBS6 decodes to "Foul Area", should be associated with an AREA object, and symbolized with area pattern for "Foul", whatever number that is. We have an AREA LUP rule for that, but it doesn't get hit in this case because the OBSTRN in question is encoded as a POINT object.

We have a rule in the tables to symbolize point OBSTRN, CATOBS7 as symbol K31. So it would work correctly if the S57 cell was correct. But in this case, it is not, IMHO.

PLIB3.2 conditional symbol procedure for OBSTRN04 is pretty clear. The only symbols that can be emitted are various DANGER, LNDARE, and UWTROC symbols. If a POINT OBSTRN with CATOBS of "Foul Ground" is to be shown, no CSP is involved. Its a simple rule, resulting in K31. Never seen this, but surely exists.

Over the course of development, I have often been tempted to editorialize/improve S52. I have resisted so far. I suspect that many of these types of questions have been raised in professional circles, and addressed in more modern PLIB editions. A future activity for us, to which I look forward

Further discussion on this subject is welcome anytime.

Thanks again
Dave

[bdbcat]

Fabbian....

Another look, and I found a bug in the OpenCPN's OBSTRN04 CSP.

Our favorite point should actually be symbolized as OBSTRN01, according to PLIB3.2

This resolves your concern about ISODGR01 as related to safety depth.

What we get now is a little dotted line circle which must be queried to discover that it refers to "foul area". Note that the chart area covered by the little circle in this case has no relation to the size of the physical area.

There is no way this S57 cell can produce SY(FOULGND1) at this point in 3.2.

Thanks for beating on this. We can improve one point at a time.

Dave

[fgd3]

Originally Posted by bdbcat

Fabbian....

Sorry, what is K32? I cannot find it in symbol libraries....

Dave

It's in Chart No. 1, page 45. K refers to the section on Rocks, Wrecks, and Obstructions. The number follows the Chart Specification of the IHO. K-32 is a blue oval with "Foul" (or "Wks" or "Wreckage") inside and a dotted line around it.

Maybe you've already found it by now.

Fabbian

[bdbcat]

Fabbian...

Thanks for the info.
Is there an S52 symbol defined for this? I don't see it in the 3.4 Addendum...

If there is, we might expect a more modern PLIB to use it for POINT OBSTRN, CATOBS6.

Wondering
Dave

[fgd3]

Originally Posted by bdbcat

Fabbian...

Looked at this again.

I'm afraid the interaction between S57 and S52 is not as intuitive as we might like it to be. Add to that the errors in S57 encoding, and we can get mush...

That's certainly true. The more I look at vector charts the harder it gets. Add to that the difficulty of finding any documentation of the S-52 specs and the greater difficulty of understanding what you find and it's a wonder anyone is able to make S-57 charts look anything like the nautical charts we're familiar with.

Originally Posted by bdbcat

At the particular point you mention (NW of the marina in Bellingham), the object is encoded as a POINT object OBSTRN, with attribute CATOBS6. This seems wrong to me. CATOBS6 decodes to "Foul Area", should be associated with an AREA object, and symbolized with area pattern for "Foul", whatever number that is. We have an AREA LUP rule for that, but it doesn't get hit in this case because the OBSTRN in question is encoded as a POINT object.

Could there be a problem with how you deal with POINT and AREA objects? Look at http://www.iho-ohi.net/iho_pubs/stan...I_Addendum.pdf, page 148, where SY(FOULGND1) is described as "foul AREA of seabed safe for navigation but not for anchoring" and page 574 where AP(FOULAR01) is described as "foul AREA, not safe for navigation". Both reference K-31 (although the latter should reference K-32, I think).

If the rules you're using are correct perhaps this should be reported to NOAA's Office of Coast Survey as a chart discrepancy.

Originally Posted by bdbcat

We have a rule in the tables to symbolize point OBSTRN, CATOBS7 as symbol K31. So it would work correctly if the S57 cell was correct. But in this case, it is not, IMHO.

PLIB3.2 conditional symbol procedure for OBSTRN04 is pretty clear. The only symbols that can be emitted are various DANGER, LNDARE, and UWTROC symbols. If a POINT OBSTRN with CATOBS of "Foul Ground" is to be shown, no CSP is involved. Its a simple rule, resulting in K31. Never seen this, but surely exists.

What is CSP?

Originally Posted by bdbcat

Over the course of development, I have often been tempted to editorialize/improve S52. I have resisted so far. I suspect that many of these types of questions have been raised in professional circles, and addressed in more modern PLIB editions. A future activity for us, to which I look forward

You're right to stick with the S-52 standard. If the standard produces incorrect results it would be better to write the IHO about the problem than to bypass it with a non-standard workaround in OpenCPN.

Originally Posted by bdbcat

Further discussion on this subject is welcome anytime.

Thanks again
Dave

[fgd3]

Originally Posted by bdbcat

Fabbian...

Thanks for the info.
Is there an S52 symbol defined for this? I don't see it in the 3.4 Addendum...

If there is, we might expect a more modern PLIB to use it for POINT OBSTRN, CATOBS6.

Wondering
Dave

I don't think so. The only symbols shown are in column 2, which is for NOS symbols. If there were an S-52 symbol I think it would be shown in column 5. Sorry I didn't pick up on that earlier.

Fabbian

[bdbcat]

Fabbian....

CSP is "Conditional Symbology Procedure" S-52 speak for process to determine the correct symbol(s) to render based upon the attributes of an object, its and relation to mariner settings and Safety Depth particularly.

Dave

[janp391]

I don't know if I am right in this tread, but if we discuss what we can find in the vector charts I see a difference between 2 different softwares.
I compare the same cm93 between Open CPN and another software, this show 2 shoals in one program but not in Open CPN. (Compare north and south of the light house in the attachment) This is my home water and I know where I could find them but looking in OpenCPN I only see a red cross in the chart. I have tried to alter the values of safety depth and the other values, but I can not come to the point that I can see that there is a shoal in open CPN. Is there a way that I can be sure of that I see all of them?
Utanfor Skarhamn.doc

[Netsurfer]

Originally Posted by janp391

... if we discuss what we can find in the vector charts I see a difference between 2 different softwares.
I compare the same cm93 between Open CPN and another software

And your screenshots show another missing in OCPN - the degree value off the bearing line (the 50.0).
Is there anything we can do to also have these information shown up in OpenCPN?

Gunther

[P_Dub]

Change your version for the one 2.4.412 and you will probably see a difference. For me it did work.
Regards.
Pierre

[bdbcat]

P_Dub....

Please remember that the effort for Version 2.4 is a Beta test program.

If you have specific, reproducible errors with 423, please bring them with screen shots as appropriate to the forum. Compare to Version 412 as useful.

If unable to assist, please desist in recommending 412. 412 is obsolete, and no Beta test information for 412 is today useful. That version has known bugs, and is closed.

If you need a stable, production quality system, then choose version 3.2.1.

Thanks for your help
Dave

[bdbcat]

Janp391....

I cannot read your attachment. What is the format, please? Does not look like a .doc file to me. Or can someone else please re-upload for me.

Thanks
Dave