Real Stars in Charted Space

[Garnfellow]

The source of much cartographic confusion in Traveller.

I'll admit it has bitten me.

 

[whulorigan]

Early Map of Charted Space (MT-CDROM)

It's interesting to note that the MT CD Rom from FFE has a hand-drawn annotated map of charted space that shows Deneb beyond Zhodani space in what would be Viajlefliez Sector. It also shows Antares in Ley Sector. It shows the location of a pretty fair number of stars in the region of the Solomani Sphere. I'm not sure how close it is to being accurate, but it's worth looking at as part of this discussion.

I have one of the original MT-CDROMs (from about ~2005); I am wondering whether or not the current edition on FFE is an updated 2nd Edition of the Disc (and whether or not I should repurchase the disc)?

Hand-drawn, annotated map of charted space is in the Images directory.

At least on *my* version of the CD ROM. It looks like it is from 2012.

Looks like I need to get an updated CDROM . . .

OK I went and got an updated CDROM, and I can now see the documents you mentioned.

That is a very interesting early map of Charted Space. In fact, it is clearly and directly related to the Map shown in an old (and early) issue of Different Worlds magazine with an article entitled: The Imperium - A Traveller Campaign (© JTAS 1980). That map has very few of the stars labeled, but does have all of the rifts shaded as in the hand-drawn map, as well as the alternate position of Terra at the same location as on the hand-drawn map.

Deneb is at roughly what its correct position should be (relative to the positions of other stars) on the other side of the Consulate. Interestingly, the Pelican Nebula is noted as being in the lower claw of the Great Rift, whereas on the canonical maps it is located in the midst of the Zhodani Consulate (which is a much more correct position, BTW).

Kuzu (Kusyu) on both maps is roughly located in Subsector J of Dark Nebula (its current canonical position) as opposed to its previously debated position of it having been located in Subsector G.

Spica appears to be located in the Hinterworlds Sector in the middle of the Outcast of the Whispering Sky. "Kaitain" (also an alternate name of Alpha Piscium) is located in Subsector P of Dark Nebula. And "Kursa" (Beta Eridani) is located in Subsector P of Magyar Sector (is this the homeworld of the Kursae?).

Enif (Epsilon Pegasi) is marked as being located in Tsadra Sector, and Baten Kaitos (Zeta Ceti) is marked in a small "island" of the Great Rift at the Spinward extreme of the Aslan Hierate.

It is also interesting to note the world of Troy located on the border of Subsectors D & H in the Trojan Reach (though the world name did not survive on Traveller Map, it still went on to give the Sector its name).

EDIT: Also interestingly, Spinward is specifically listed as 0.00h Right Ascension, Rimward is specifically listed as 6.00h Right Ascension, Trailing is specifically listed as 12.00h Right Ascension, and Coreward is specifically listed as 18.00h Right Ascension. I wonder if this is where the 45^o Galactic-Coordinate axis-discrepancy (i.e. "Traveller"-Coreward versus "True"-Coreward) arose?


Re-buying the MT-CDROM was worth it just to get that map. (And I got a Noble Patent Card for "Baroness of Northammon" - don't know what I am going to do with that yet ).

 

[Garnfellow]

And "Kursa" (Beta Eridani) is located in Subsector P of Magyar Sector (is this the homeworld of the Kursae?).

Weird to place Beta Eridani there; I'd have to find my protractor to see if that's actually 5h 7m RA.

For what it's worth, I put the star in Kursa subsector (H) of Aldebaran. The T5 rulebook notes the Kursae are "From an unidentified homeworld somewhere in now-Charted Space," so Beta Eridani -- wherever it is -- probably wasn't the true homeworld, but the first Kursae-settled world discovered by Terrans very well might have been in or around the Kursa system.

I also placed an unidentified minor race on a nearby world in subsector H, just in case.

 

[G. Kashkanun Anderson]

Weird to place Beta Eridani there; I'd have to find my protractor to see if that's actually 5h 7m RA.

For what it's worth, I put the star in Kursa subsector (H) of Aldebaran. The T5 rulebook notes the Kursae are "From an unidentified homeworld somewhere in now-Charted Space," so Beta Eridani -- wherever it is -- probably wasn't the true homeworld, but the first Kursae-settled world discovered by Terrans very well might have been in or around the Kursa system.

I also placed an unidentified minor race on nearby world in subsector H, just in case.

Beta Eridani would be a terrible idea for any species' homeworld anyway -- or any life-bearing planet at all, really -- since it can't be any older than 1.7 billion years, and it's already well into its post main sequence phase (in other words, about where the Sun will be in about 5-6 billion years).

 

[sudnadja]

This list gives distance from Terra and position in galactic longitude. Assuming 0 degrees galactic longitude corresponds to coreward, and 90 degrees corresponds to spinward, using a little trigonometry we can then calculate the approximate position within Charted Space of these stars. I was hoping to get some eyes on my Google Sheets to make sure my methodology works:

Do the canonical traveller maps use the galactic plane as a basis for the mapping?

It's more than a year since you posted this, but this is something I've taken a shot at a few times (never with any success). Here, for example, I took the 70 closest stars to the sun, took their heliocentric equatorial cartesian coordinates and rotated them into the galactic basis (and then rotated counterclockwise by pi/2), and then simply chopped off the +/- Z data. This approach is never effective because space close to the sun is very three dimensional, so you'll see Barnard, even though it's the second closest star system to the sun seeming to be just one of a whole fleet of nearby stars that are even closer to the sun.

When you get to very long distances in traveller charted space, the relationship to real positions becomes more possible to do, as the stars tend to have much less of their distance in the Z component.

I'll take a look at your data too - thanks for posting it

 

[sudnadja]

The Google Sheets workbook contains a Notes sheet that goes into more detail, but basically I translated the galactic longitude and distance into polar coordinates with Terra (Solomani Rim 1827) at position (0,0). I then converted those coordinates into hex offsets, and then determined where the resulting position fell within Charted Space to get a sector and hex number.

There are several places where I could have screwed up, so I would welcome any questions, thoughts, or corrections.

I plotted your close in stars vs close in stars from catalog data. I had to rotate counterclockwise by pi/2 to get the match (my data with no rotation matches galactic Cartesian coordinates). Alpha Centauri, Procyon, Eta Cassiopeiae and Altair are good matches. For some reason your position for Tau Ceti is significantly different than mine. I plotted my star points with a greek lowercase alpha and yours with a red dot.

At a longer range for the brighter stars, our catalogues don't contain many of the same stars but there are a few near exact matches:

We definitely do not have the same location for Tau Ceti, which is odd.

From scratch, for me, Tau Ceti is at RA=26.0214 degrees, DEC=-15.9396, proper motion RA,DEC in mas/year={-1721.82,854.02}, parallax in mas=274, and radial velocity=-10.6 km/sec

Only RA,DEC and parallax are needed for a three dimensional solution, so doing the spherical to cartesian mapping you get {3.153575,1.53956,-1.002277} parsec - this in the equatorial plane. Rotate that into the ecliptic plane and you get {3.15358,1.01384,-1.53197} and then rotate that into the galactic plane and you get {-0.317353,-0.0371296,-3.63562} - 3.64 parsec away, which is catalogue distance, just that most of the distance is in the z component. In this case, Tau Ceti is almost directly under us relative to the galactic plane. It looks to me like you might have taken the z distance component for y?

 

[Garnfellow]

Only RA,DEC and parallax are needed for a three dimensional solution, so doing the spherical to cartesian mapping you get {3.153575,1.53956,-1.002277} parsec - this in the equatorial plane. Rotate that into the ecliptic plane and you get {3.15358,1.01384,-1.53197} and then rotate that into the galactic plane and you get {-0.317353,-0.0371296,-3.63562} - 3.64 parsec away, which is catalogue distance, just that most of the distance is in the z component. In this case, Tau Ceti is almost directly under us relative to the galactic plane. It looks to me like you might have taken the z distance component for y?

First, thanks for taking a look at this -- better late than never!

I am not certain what is going on with Tau Ceti, but I do think it has something to do with the Z distance. It looks like it is in the same line as your position, just way farther out.

 

[whulorigan]

According to SIMBAD, Tau Ceti (i.e. "Iilike") is located at:

Radial Distance: 3.65pc
Galactic Longitude: 173.1007^o
Galactic Latitude: -73.4397^o

This would put it at a highly Galactic South latitude, and at a bearing slightly spinward of rimward.

However, when looking at the Traveller Map coordinates, keep the following in mind:

NOTE CONCERNING STAR POSITIONS:

When comparing real-astronomical star positions to the Traveller Map of Charted Space, it needs to be observed that the axes of the Charted Space Map appear to be tilted almost exactly 45^o counter-clockwise from the correct position (i.e. "true" Coreward would seem to lie parallel to the Lesser Rift). If one makes that adjustment, about 70% of the named stars will fall roughly along their correct bearing. However, there are some exceptions in which the star in question lies in roughly the correct position relative to the Charted Space Map without the need to rotate the coordinate axes.

My general philosophy is that if one is assigning the position of a Real-Universe star to a Traveller Map hex, it should be acceptable if:

1) The star's longitude angle (θ) is within a 45^o bearing-arc between the "true" coordinate axes and the ''Traveller'' Charted Space Map coordinate axes, and​

2) The star's assigned distance from Terra on the hexmap falls somewhere between its true distance R in parsecs, and its 2D-projection onto the flat map as seen from above, found by D_proj = {R * cos(ɸ)}, where (ɸ) is the latitude angle. If desired, the distance above/below the plane can be determined by Z = {R * sin(ɸ)}.
​

EDIT:

Do the canonical traveller maps use the galactic plane as a basis for the mapping?

Not sure, but I believe the original Traveller Map was based on Right Ascension and Declination data, based on the axis labels of the hand-drawn map on the MT CD.

One thing to keep in mind is that while the Earth is tilted to the ecliptic by about 23.5^o, the entire ecliptic itself is tilted to the mean Galactic plane by about ~ 58^o-62^o, if I am remembering my numbers correctly. Tilt Images.

 

[sudnadja]

According to SIMBAD, Tau Ceti (i.e. "Iilike") is located at:

Radial Distance: 3.65pc
Galactic Longitude: 173.1007^o
Galactic Latitude: -73.4397^o

This would put it at a highly Galactic South latitude, and at a bearing slightly spinward of rimward.

However, when looking at the Traveller Map coordinates, keep the following in mind:



My general philosophy is that if one is assigning the position of a Real-Universe star to a Traveller Map hex, it should be acceptable if:

1) The star's longitude angle (θ) is within a 45^o bearing-arc between the "true" coordinate axes and the ''Traveller'' Charted Space Map coordinate axes, and​

2) The star's assigned distance from Terra on the hexmap falls somewhere between its true distance R in parsecs, and its 2D-projection onto the flat map as seen from above, found by D_proj = {R * cos(ɸ)}, where (ɸ) is the latitude angle. If desired, the distance above/below the plane can be determined by Z = {R * sin(ɸ)}.​

I tried an alternate approach

. In this case, I extract the 2d components and then normalize and scale by the distance from the sun. Thus all of the distances from the sun are correct (but any other star to star distance is not).

I'll add more commentary later, but one thing to notice - if you see the line extending from Barnard's Star, that's the proper motion of the star (output of an nbody integrator) in the 3622 years between now and 1116 imperial. This is enough to displace it by a hex or two vs 2017-earth mapping. I don't know if the original traveller mappings were trying to place stars relative to their positions in the 20th century or their future positions.

 

[sudnadja]

According to SIMBAD, Tau Ceti (i.e. "Iilike") is located at:

Radial Distance: 3.65pc
Galactic Longitude: 173.1007^o
Galactic Latitude: -73.4397^o

This would put it at a highly Galactic South latitude, and at a bearing slightly spinward of rimward.

However, when looking at the Traveller Map coordinates, keep the following in mind:

My general philosophy is that if one is assigning the position of a Real-Universe star to a Traveller Map hex, it should be acceptable if:

1) The star's longitude angle (θ) is within a 45^o bearing-arc between the "true" coordinate axes and the ''Traveller'' Charted Space Map coordinate axes, and​

2) The star's assigned distance from Terra on the hexmap falls somewhere between its true distance R in parsecs, and its 2D-projection onto the flat map as seen from above, found by D_proj = {R * cos(ɸ)}, where (ɸ) is the latitude angle. If desired, the distance above/below the plane can be determined by Z = {R * sin(ɸ)}.
​

EDIT:


Not sure, but I believe the original Traveller Map was based on Right Ascension and Declination data, based on the axis labels of the hand-drawn map on the MT CD.

One thing to keep in mind is that while the Earth is tilted to the ecliptic by about 23.5^o, the entire ecliptic itself is tilted to the mean Galactic plane by about ~ 58^o-62^o, if I am remembering my numbers correctly. Tilt Images.

I clipped out the Sol subsector from TravellerMap, aligned it with the galactic plane, and then applied a rotation of -pi/4 so that sol and alpha centauri would align appropriately:

The red dots are the closest stars (out to Sirius A/B), appropriately transformed into the galactic basis. The black arrow is a pointer to the true direction toward Sagittarius A*, which I will take to be the center of the Milky Way. I've also plotted the equatorial and ecliptic planes, which is interesting in just making note of how rotated they are compared to the galactic plane. There is a bit of a parallax problem viewing the map from above - though some of the stars kind of line up, let's look at the map from the side:

There's a huge z component to these and though not violating the +/- 45 degrees rule based on longitude, the flat map gives a very misleading picture of local space.

I've tried several dimensions reduction methods (principle component analysis, self organizing maps) but not had really good results yet.

I do think that for the players that are more interested in physics and astronomy, and play traveller more for those aspects than others, it's not bad to have a sound methodology for placing real astronomical objects on a 2d map, or just do away with the 2d map and try to transform traveller into a 3d game (which it really, really does not want to be).

 

[whulorigan]

...I tried an alternate approach . In this case, I extract the 2d components and then normalize and scale by the distance from the sun. Thus all of the distances from the sun are correct (but any other star to star distance is not).

I'll add more commentary later, but one thing to notice - if you see the line extending from Barnard's Star, that's the proper motion of the star (output of an nbody integrator) in the 3622 years between now and 1116 imperial. This is enough to displace it by a hex or two vs 2017-earth mapping. I don't know if the original traveller mappings were trying to place stars relative to their positions in the 20th century or their future positions.

One of the most mis-matched stars is Epsilon Indi ("Meshan"), which is on entirely the wrong side of the coreward-rimward axis. Epsilon Indi is also a star that is at a significant Galactic South Latitude, so I am curious if perhaps it was placed where it was due to an attempt to preserve various stellar distances when transforming from a 3D to a 2D environment (i.e. preserving an approximate correct distance from Terra, while also attempting to keep it far enough away from other stars that would otherwise be far to close by a simple projection to 2D).

The only oddity about that theory is that in GDW's 2300AD universe, they used a very accurate catalog (for the time - 1969 Gliese catalog) as the basis for an actual 3D starmap out to 50 lightyears. But Epsilon Indi is in the wrong position on that map as well.

 

[sudnadja]

One of the most mis-matched stars is Epsilon Indi ("Meshan"), which is on entirely the wrong side of the coreward-rimward axis. Epsilon Indi is also a star that is at a significant Galactic South Latitude, so I am curious if perhaps it was placed where it was due to an attempt to preserve various stellar distances when transforming from a 3D to a 2D environment (i.e. preserving an approximate correct distance from Terra, while also attempting to keep it far enough away from other stars that would otherwise be far to close by a simple projection to 2D).

The only oddity about that theory is that in GDW's 2300AD universe, they used a very accurate catalog (for the time - 1969 Gliese catalog) as the basis for an actual 3D starmap out to 50 lightyears. But Epsilon Indi is in the wrong position on that map as well.

Wow, Epsilon Indi is way off. Here, with more stars, the z component flattened out to zero and labeled for context:

 

[aramis]

The only oddity about that theory is that in GDW's 2300AD universe, they used a very accurate catalog (for the time - 1969 Gliese catalog) as the basis for an actual 3D starmap out to 50 lightyears. But Epsilon Indi is in the wrong position on that map as well.

There is an interesting thing about real world data in tables...
If the data is all real data, the table's contents cannot be protected by US copyright; it's there for academics to be able to update each others' data with new data, and for the press. Cite.

If some entries are intentionally fictive, the table's content (not just the layout) is protected as a work of fiction...

There are 3 or 4 apparently intentional errors in the 2300 data... which makes the 2300 NSL copyrightable as a work of fiction. So, GDW, (now Marc) can enforce copyright over versions which include those intentional discrepancies.

 

[sudnadja]

There is an interesting thing about real world data in tables...
If the data is all real data, the table's contents cannot be protected by US copyright; it's there for academics to be able to update each others' data with new data, and for the press. Cite.

Though the Traveller names are fictional, regardless of the accuracy of the positioning. It isn't all that easy to find the names of real stars that the fictional ones are analogues of, except in a few cases. Would the entire table still be protected by copyright if one star were mispositioned among hundreds?

I suppose that could indeed be the reasoning, but it's a little unfortunate when the players happen to be astronomy buffs.

 

[aramis]

I suppose that could indeed be the reasoning, but it's a little unfortunate when the players happen to be astronomy buffs.

The tables for drives in CT have 2 (1977) or 3 ( 1981) non-formulaic includes, too...

 

[sudnadja]

According to SIMBAD, Tau Ceti (i.e. "Iilike") is located at:

Radial Distance: 3.65pc
Galactic Longitude: 173.1007^o
Galactic Latitude: -73.4397^o

This would put it at a highly Galactic South latitude, and at a bearing slightly spinward of rimward.

However, when looking at the Traveller Map coordinates, keep the following in mind:

My general philosophy is that if one is assigning the position of a Real-Universe star to a Traveller Map hex, it should be acceptable if:

1) The star's longitude angle (θ) is within a 45^o bearing-arc between the "true" coordinate axes and the ''Traveller'' Charted Space Map coordinate axes, and​

2) The star's assigned distance from Terra on the hexmap falls somewhere between its true distance R in parsecs, and its 2D-projection onto the flat map as seen from above, found by D_proj = {R * cos(ɸ)}, where (ɸ) is the latitude angle. If desired, the distance above/below the plane can be determined by Z = {R * sin(ɸ)}.
​

EDIT:


Not sure, but I believe the original Traveller Map was based on Right Ascension and Declination data, based on the axis labels of the hand-drawn map on the MT CD.

One thing to keep in mind is that while the Earth is tilted to the ecliptic by about 23.5^o, the entire ecliptic itself is tilted to the mean Galactic plane by about ~ 58^o-62^o, if I am remembering my numbers correctly. Tilt Images.

If the original traveller map was based on right ascension and declination, then the fundamental plane is equatorial, which throws a lot of this into question.

I take what I think is the same approach to finding distances above and below the plane of the galaxy, but from a linear algebra approach, I have a single rotation which will transform from the equatorial plane to the galactic plane (in J2000) ( {{-0.0548755,-0.873437,-0.483835},{0.494109,-0.44483,0.746982},{-0.867666,-0.198076,0.455985}} ) and I take catalog RA,DEC and parallax data (and proper motion data if I am computing motion as well), do a spherical to cartesian mapping, then transform them with the above matrix into the galactic basis. The z component then is the distance above or below the earth, in the galactic frame.

 

[sudnadja]

I associated the known stars list traveller names with their presumptive real world stars and plotted them in 3d over the aligned 2d map. Blue dots are known traveller stars, smaller red dots are other real world stars that are not known to be associated with a traveller star. There are some oddities that fall out of that, the 2d Ys is placed where the 3d Fenris should be, Sirius is where Fenris is placed, and so on.

From the 2d map you get the feeling you can go from Terra to Prometheus to Ember as a set of jump-2, all in the same direction, yet Prometheus to Ember in 3d is 4.14 parsec, requiring jump-5.

What is the most distant traveller star that has a real analogue? Deneb? If so, what's after that, Ishdar (epsilon Cygni)?

 

[whulorigan]

What is the most distant traveller star that has a real analogue? Deneb? If so, what's after that, Ishdar (epsilon Cygni)?

Deneb is the farthest, I believe. But it is actually way to close. In reality it lies to spinward of Terra, but allowing for the 45^o rotational anomaly on the TravellerMap stars, it falls along a consistent bearing realtive to other TravellerMap stars. However, it is at about half its true distance from Terra. It ought to be somewhere along the coreward border of the Zhodani Consulate.

Interestingly, the early hand-drawn map of Charted Space by GDW on the MT CD does place Deneb in the Delta Quadrant of Viajlefliez Sector corespinward of the Zhodani Consulate border. But for some reason it got moved on the published maps.

 

[sudnadja]

Deneb is the farthest, I believe. But it is actually way to close. In reality it lies to spinward of Terra, but allowing for the 45^o rotational anomaly on the TravellerMap stars, it falls along a consistent bearing realtive to other TravellerMap stars. However, it is at about half its true distance from Terra. It ought to be somewhere along the coreward border of the Zhodani Consulate.

Interestingly, the early hand-drawn map of Charted Space by GDW on the MT CD does place Deneb in the Delta Quadrant of Viajlefliez Sector corespinward of the Zhodani Consulate border. But for some reason it got moved on the published maps.

Ok, using catalogue data for Deneb, I get the "real" location of Deneb to be in the sector to the Core-Spinward of Stinj Tainz, -10 sectors x and +9 sectors y from the Sol sector. This presuming that Deneb is 477 parsec from Earth. Fortunately, the z component in the galactic plane for Deneb is very small relative to earth, it would have been a good reference had it been placed with care on the traveller maps.

Our Galactic core is at +192 sectors x, +134 sectors y in this reference, and the Large Magellanic Cloud would be at +1235 sectors x, -596 sectors y. I imagine neither of these have been canonically placed?

Thanks!

 

[sudnadja]

In reality it lies to spinward of Terra, but allowing for the 45^o rotational anomaly on the TravellerMap stars

When I zoom out on TravellerMap far enough, I do see the overlay of the galactic core straight up from charted space, but of course close in stars in so far as they are known are rotated. At what point does the rotation anomaly go away? Was it done this way intentionally or was there a mistake somewhere in someone's math?