Celestial Navigation Plugin Redux

[P_Dub]

Rick take a look at your private message.

[rgleason]

Stelian, It works great. Thank you.

PS: Have you tried this?
https://github.com/stelian42/celesti...on_pi/releases

I think Norton looks at that a little differently.

[stelian]

Originally Posted by rgleason

Stelian, It works great. Thank you.

PS: Have you tried this?
https://github.com/stelian42/celesti...on_pi/releases

I think Norton looks at that a little differently.

No, didn't tried, will do next time. Thanks.

[rgleason]

Stelian, Boat_Alexander, P_Dub
Can you explain this? It does not make any sense to me.
I've set the date and time to 10/10/2017 UTC 180000.00
Why have the sights moved since yesterday?
Perhaps my understanding and assumptions are wrong?

See this Post

Simplification: Parameters that are always the same:
Clock Offset =0
Time: Boston Time ( UTC-5): Oct 10, 2017, 13:00 so UTC 10/10/17 18:00
DR Shift: Distance=0 Bearing=0
Parameters: Eye height=2.0 m Temp 10 c, Pressure=1010, Index Error 0 min
Latitude: 42.35, Longitude to -71.1
-----
Using "Find" and Altitude set for the Star and enter the Lat/Long above:
Star - Altitude Entered
Alkaid 79.502
Kochab 58.133196
Arcturus 66.507224
Sun 36.888867
etc

I have managed to "Move boat to lat\long" to get the boat closer to the exact position of 42.35 N 71.1 W by using VDR_pi and running the CN-bos-boat-pos.txt file with
these nmea0183 sentences:
$GPRMC,180000.000,A,4221.0000,N,07106.0000,W,0.00, 180.00,101017,,,*3E<cr><lf>
$IIHDM,205,M*3B<cr><lf>
$GPGLL,4221.000,N,07106.000,W,180000.00,A,*18<cr>< lf>
This makes entry a little easier.

Also, quite honestly, I cannot understand why the sights are not crossing in the same spot!! They are really quite out of whack and make it impossible to find a good fix. So I am going to attach the sight.xml file again too. So you guys can check what I have done first hand.

I am actually disturbed by this big shift when the sights are set on a given data and time. Why?

Copy and rename Sights-rick3.xml.doc to Sight.xml after renaming yours.
Copy and rename CN-bos-boat-pos.txt.doc to CN-bos-boat-pos.txt and run with VDR_pi.

[rgleason]

Well, what I found out caused that shift was again I found -10000 in the "Clock Offset" and the reason was this. Entry of this value does not take effect until you close the entire plugin and then open it again! At the end of the last session a day ago, I had entered that value -10000 to see what happened. Nothing happened, so I went on and just closed it. Then this morning I opened Celestial_Nav and the whole thing had shifted east -10000 minutes about. So I guess it is working in that respect.
However I think there is some explanation needed about this value!
Certainly more needs to be put in the manual about how it is used.
--Question: Why doesn't the value change the sights immediately? It seems like it should be able to.

I don't understand why the sights are not crossing in one place better though!

Also I am getting something else funny. Playing the VDR_pi file CN-bos-boat-pos.txt, the boat goes Red and I can zoom all the way in as far as possible and locate my cursor right over the center of the boat and the coordinates are 42.350003 and -71.10001 which is good.

But then when I open the Celestial Navigation plugin and select one of the "Sights" > "Edit" > "Find" the Coordinates appear: 42.6032 -68.5565
Which is not 42.35 -71.1 !!! It is consistently 42.6032 -68.5565 no matter what sight. Why shouldn't it be 42.35 -71.1 !!!
Is there some formula that is not doing math right???

Maybe if there were not such a discrepancy the sights would be crossing better!

PS: Also I'd like to have a [checkbox] Global change of Sight Properties > Minutes of Uncertainty!!

[rgleason]

Here is a way to check the Sun..
Altitude and Azimuth of the Sun given Time and Location

[rgleason]

Now the sights seem to have shifted south, away from 42.35 -71.1
I've double checked all the clock offsets, and I have the VDR_pi CN-bos-boat-pos.txt running with RMC and GLL positions and a Red Boat:

$GPRMC,180000.000,A,4221.0000,N,07106.0000,W,0.00, 180.00,101017,,,*3E<cr><lf>
$IIHDM,205,M*3B<cr><lf>
$GPGLL,4221.000,N,07106.000,W,180000.00,A,*18<cr>< lf>

Now I am pretty sure this is not working correctly.

See sights shift south screenshot.

[stelian]

Originally Posted by rgleason

I am actually disturbed by this big shift when the sights are set on a given data and time. Why?

I see that your times are set to 06:00 instead of 18:00. Changing them back to 18:00 (and setting the time offset to 0) will put them at the right place.

Did you change the time of the sights manually on purpose or did it somehow change without user interaction ?

[stelian]

Originally Posted by rgleason

Well, what I found out caused that shift was again I found -10000 in the "Clock Offset" and the reason was this. Entry of this value does not take effect until you close the entire plugin and then open it again! At the end of the last session a day ago, I had entered that value -10000 to see what happened. Nothing happened, so I went on and just closed it. Then this morning I opened Celestial_Nav and the whole thing had shifted east -10000 minutes about. So I guess it is working in that respect.
However I think there is some explanation needed about this value!
Certainly more needs to be put in the manual about how it is used.
--Question: Why doesn't the value change the sights immediately? It seems like it should be able to.

I don't understand why the sights are not crossing in one place better though!

Yes, it looks like the time offset is taken into account only at the beginning or when each sight is opened and validated. It's a bug, I'll look into it tomorrow.

Originally Posted by rgleason

Also I am getting something else funny. Playing the VDR_pi file CN-bos-boat-pos.txt, the boat goes Red and I can zoom all the way in as far as possible and locate my cursor right over the center of the boat and the coordinates are 42.350003 and -71.10001 which is good.

But then when I open the Celestial Navigation plugin and select one of the "Sights" > "Edit" > "Find" the Coordinates appear: 42.6032 -68.5565
Which is not 42.35 -71.1 !!! It is consistently 42.6032 -68.5565 no matter what sight. Why shouldn't it be 42.35 -71.1 !!!
Is there some formula that is not doing math right???

Maybe if there were not such a discrepancy the sights would be crossing better!

As I said in a previous post, I believe this is because "Find" is giving true altitudes while the altitude you enter will be considered as an instrumental measure, which will be corrected for eye height, refraction, limb etc. So at the end you will find a true altitude which is different than the one "Find" found.

edit: after re-reading I understood you were talking about the coordinates of the Find dialog not being the exact same as the ones of the boat. I'll look into this.

Originally Posted by rgleason

PS: Also I'd like to have a [checkbox] Global change of Sight Properties > Minutes of Uncertainty!!

Yes, this would be nice.

[rgleason]

Stelian,

I think you read this from the Sight.xml from this post?
http://www.cruisersforum.com/forums/...ml#post2497658
--Correct?

If so, I believe it says in the Sights.XML "6:00pm" which is 1800.
This is part of the top and one record. ClockError appears to be Global!
I changed it at the end of the day yesterday and nothing happened!
Then this morning loaded it and it took awhile to figure it out that the -10000 had decided to work on reboot. - I am glad you agree this is a bug!

Could we have the (value) of ClockError actually be written somewhere at the menu level? Maybe in the table with values for DRdist DRbrg, EyeHt, T, P, IndErr, CkErr
(Hope this isn't too difficult. It seems really easy to check things using the table. You can pick things that are out of whack.)

PS: Later: I guess there are some other values from Sights.xml that aren't in there. Maybe be able to pick simple version and full version and you can right click -Copy to clipboard either (however that won't get the calculations included.)

... <ClockError Seconds="-10000" />
<Sight Visible="1" Type="0" Body="Arcturus" BodyLimb="0" Date="2017-10-10" Time="6:00:00 PM" TimeCertainty="0" Measurement="66.5072" MeasurementCertainty="5" EyeHeight="2" Temperature="10" Pressure="1010" IndexError="0" ShiftNm="0" ShiftBearing="0" MagneticShiftBearing="0" ColourName="blue" Colour="rgba(0, 0, 255, 0.588)" Transparency="150" />

Originally Posted by stelian

I see that your times are set to 06:00 instead of 18:00. Changing them back to 18:00 (and setting the time offset to 0) will put them at the right place.

Did you change the time of the sights manually on purpose or did it somehow change without user interaction ?

[rgleason]

Stelian wrote:

Yes, it looks like the time offset is taken into account only at the beginning or when each sight is opened and validated. It's a bug, I'll look into it tomorrow.

edit: after re-reading I understood you were talking about the coordinates of the Find dialog not being the exact same as the ones of the boat. I'll look into this.

Great Thanks!

I added this to the List in Github Improvements and a few small bugs #21
which I'm trying to keep current. Going to simplify it in a little bit.

This is turning out to be a little more involved than I realized.
I need to have some instructions how you actually take sights and use all the parameters, so we can do one of those.
I am going to assume a DR direction and speed from 42.35 N 71.1 W for two sights (Lets say 180 magnetic at 5knots)
and try entering two sights that way to see what we run against.

Also I'd like to have a [checkbox] Global change of Sight Properties > Minutes of Uncertainty!!

[rgleason]

Found in HELP html in the plugin.

Date and Time for the sight must be the same as the exact moment the measurement is taken. If the sight was taken a while ago, and we have moved since then, do not modify the date and time values. Instead, enter course dead reckoning for distance traveled and bearing to shift it into position to use along with current sights. Once a sight is shifted, it is no longer be used for a position fix (it is very complicated to implement this math) but a fix can still be visually determined.

Parameters These values slightly affect the accuracy, but are normally the same for all sights.

Computing a Fix

...The regression is least squares to find a position on earth at sea level.... For sights which are shifted (from dead reckoning movement over time), or for magnetic north azimuth sights, it is extremely complicated and difficult to implement the math required to perform a fix operation.

In this case, the computer would find the fix the same way a human can visually (which is not implemented). It is also currently possible to compute a fix only from altitude shots (true north Azimuth could be implemented in the future)

I am not sure those two boxes in the FIX MENU for Lat and Long are working properly. They are a little funky. What format should be used??
I've been assuming that it is for the DR position to help the calcs, and have been using decimal degrees 42.35 -71.1 but sometimes it just resets to -90 and -180 !!!

Stelian - I see you did "Time(UTC)" which is fine as far as I am concerned. I had been thinking under the time entries in the form, but maybe this is better.

LATER: Found it in the wiki plugin manual.
=== 3. Sight Properties - Date and Time Tab ===
Enter Date (based on GMT/UT) and Time in GMT/UT, Certainty and Shift. I think you could use UT or GMT. I vote for UT because it is smaller and more familiar.

From Help html Should name it Time(GHA) instead of Time(UTC)? I think UTC is more familiar, but Sean may prefer the other. Hope he'll speak up.

GHA : GHA (Greenwich Hour Angle) indicates the position past the plane of the Greenwich meridian measured in degrees. Equivalent to longitude on earth. It can be followed by East or West.

"GHA(star) = SHA(star) + GHA(aries). Ideally the GHA of aries ..."

"Note that astronomers and land surveyors use RA (Right Ascension) instead of SHA, usually measured in time (hours and minutes) and increasing in an Easterly direction, to convert just remember that 1hour=15 degrees, 1 minute of time = 15 minutes of arc and don't forget to change the sign. Instead of GHA Aries they use GST (Greenwich Sidereal Time).

[stelian]

Originally Posted by rgleason

Stelian,

I think you read this from the Sight.xml from this post?
http://www.cruisersforum.com/forums/...ml#post2497658
--Correct?

If so, I believe it says in the Sights.XML "6:00pm" which is 1800.

You are correct, but the table in SightsDialog shows 06:00 UTC.

I guess there is some other bug here, maybe similar to the date (mm/dd/yyyy versus dd/mm/yyyy) which I already fixed.

This explains probably why the sights have shifted west (you created them with time set to 18:00, then closed opencpn or just the plugin; when reopened, they were loaded with time set to 06:00 (UTC) -> so shift west).

I'll look into this.

[rgleason]

Sean wrote that once a sight is shifted with course and bearing it is no longer to be used for a calculated fix ( but it can be done visually.
I don't believe this calc affects clock offset but perhaps so, needs to be checked.

Additionally, and another point, sights are entered with clock offset of 0, then if clock offset is entered, the sights shifted, then clock offset 0 again, sights should shift back to the exact same position! (I believe).

In my example, it was shifted back say 7 to 8 hours later and I am not sure if it went back to the same position, in fact I think the screenshots may show it is slightly different. Then through the day, as I checked, the sights got worse and worse, moving south. Also shown in the screenshots I believe. This is wierd behavior for a dead simple set of sights that are supposed to mark a certain spot IMHO.

[rgleason]

Yes, Late one night after working with Cel Nav I believe I entered -10000 into Clock Offset. Nothing seemed to change. I don't believe I added any sights that night. I closed down cel nav, opencpn and the computer. Next morning 7-8 hrs later I opened up cel nav and it had shifted east across the atlantic. I found the clock offset entry of -10000 and zeroed it nothing happened. Then I closed cel nav and opened it and everything had shifted back, but I and not at all sure it shifted back to the same place. Then throughout the day the sights seemed to drift south enough for me to noticed it.

As it happens I am going to see Sean today and will ask him about this and how clock offset should be used and what it should do! How it is used.

Also why the drift of sights.

And why dont they cross properly? Large inaccuracy here.

To. Label time with UT rathet than GMT? Actually I realize UTC is universal time correction, in this use it is not a correction, we are to use UT!

Anything else?

Thanks again for helping to get this fixed and confirned.

Rick