Learning Curve------

[landnrailroader]

Colleagues,

I read whatever instructions we get with a simulator, and also look at the video tutorials. All nice, but an explanation of what the various keys do while in the editor function would be nice.

Having said that, however, learning that you don't automatically get the white/yellow switch to indicate a tanget was a big help. I am attaching an image. The track from the left is the James River sub. which is ascending at just under a 1% gradient (1 in 100) and that from the right is the Mountain Sub., now the Buckingham Branch RR, over which the Amtrak Cardinal operates 3 days a week, each way. Near the center is the turnout at "JD Cabin". Along the former C&O control towers were called "cabins" and so the name of this junction tells us that at some time in the past there was a tower at the jct. called "JD". The tangent doubletrack is level and will be so for about 1.5 miles to Smith Creek at which point it will start to ascend at about 1 in 250. The dead end track at the right is the lead to a small yard at the east end of the bigger Clifton Forge Yard, the bulk of which will start after the curve in the distance and be to the right (south) of the mainline.

If KujuAdam sees this, I have a suggestion for some future release. When producing a curve, it would be very helpful if there was a arrow or other symbol, about 200M in lenght, extending off the far end of the curve as if it was a tangent of 200m. This would make it much easier to get the required swing of the curve correctly. I was able to eventually get my tangents close, though not exact, by trimming off the end of the curve until the swing was correct. This is tricky with double track because as soon as you bail out of the track laying session, the double track becomes 2 seperate tracks. I hope there is a way to bond the tracks together so that a clip workds at the same point on both tracks.

landnrailroader

[landnrailroader]

2nd try to get the image attached

[theokus]

2nd try to get the image attached

How did you get that curve along the flags?
Is there a procedure to lay tracks according the DEM?

thx.

[landnrailroader]

The procedure that I used is called T&E, that is, trial and error. As I have stated in
other threads, I much prefer to use named markers. I have found that there is an apparent limit to the number of named markers that one can have, since when I pass 500 in number, there seems to be a random drop out of some of them.

Here is what I have learned. If you lay multiple tracks at the same time by enumerating the number required in the appropriate box, it is easier to get X number of tracks, spaced the same distance apart and parallel. problem is that once you double right click and get out of track laying, you now have X number of seperate tracks.

I use UsaPhotomaps because I am familiar with it and have used in it several MSTS routes. I also have DEMEX which has a routine in it for converting UsaPHotomaps waypoint files to MSTS marker files. However there are some freebies in the library to accomplish this also. Then I use Mike Simpsons tools to convert the marker file to a .csv file and to make a blue print etc. Once the flags are displayed in the route, it is a matter of careful trail and error fitting to get the curves to follow, I would put it, "the average of the flag positions". There is probably a 2-5 meter error +- in positions taken off of UsaPhotomaps which would probably be no better if you were using Google Earth or whatever, except in areas where there are very high resolution images. In some areas (Seattle, Mobile, ATlanta, DC to name a few) Terraserver has high res. urban images and UsaPhotomaps will work with those, in which case the error need not be worse than 1/2 the track gage or about 2 feet.

So the short of it is --- T&E, Trial & ERROR. I refitted those curves 5 or 6 times until I was satisfied with the merge. The two center tracks converge at what is supposed to be a #15 Righthand turnout, and I have this such that it looks reasonable, but what it actually is is anybody's guess. KUJU is apparently not interested in what they call "fixed junctions" and I don't have the skill with 3DS to make some. This limits this simulation as far as American practise is concerned, but I am going to play with it a bit, and then check out MSTS-X if and when it is released.

Turnouts are another matter and take some getting used to. If you select the curved portion of the turnout and get a message that "turnout rails do not intersect mainline", I found it necessary to relay the straight portion of the turnouit back to just short of the previous frog (it was a #10 ladder, approx.). Usually this meant that I was trying to put my points either at, or beyond, the previous frog, which is a no-no I guess. The solution is to continue the branch track straight toward the ladder until a bit closer, and then curve it into the ladder to form the switch, making sure that when the "mainline, actually the straight side" flashes pink, the curved portion does not snap beyond the previous frog.

landnrailroader

[livercup]

Looking Good! I am having a whale of the time laying a straight section 10.5 miles from point to point. Not much play room there. How do you slow down the object rotation tool to be more precise. We need an aim straight utility.
Tom

[zbobg]

Hold the shift key to slow down the gizmo movement options. Doesn't slow it down by much but it helps.

I've got a speed adjustable mouse so I turn the speed down to minimum when moving things around. Just the opposite of increasing the speed/response for HL2 for example. For a speed adjustable mouse lowering the speed means the same thing as increasing the mouse resolution.

The control key sometimes speeds 'gizmo' movement up if you press ctrl and release while grabbed onto a gizmo and you are moving your object - haven't figured out if that's intended and it doesn't work reliably.

[Paddington bear]

Hi,
yip, The "T & E" Procedure is wonderful.
Gets better with Time & Experiance, after a couple of months use it is like second nature really. Track laying is a bit like a famous TV add............... "It wont happen over night, but it Will happen". There is NO magical short cut. The results improve all the time.

Ross

[landnrailroader]

I have no idea how you slow down the rotation,and in my case it seems to jump about
2 degrees at a time. Would be nice if you could slow it down, and what would be even nicer is if you could SPECIFY the Azimuth angle of a tangent.

If you had that (plus see my comment about a pointer on the end of a curve) you could start your route at any long straight section that you knew the elevation of, and work to and fro from there.

landnrailroader

[Paddington bear]

hi,
With any object selected, the Gizzmo/object can be "controlled" using the keyboard.....

F, G rotate one way/other way
V, B up/down
C,N move object closer to camera or away from the camera along the cameras line of sight axis. you can move the camera anywhere in the sphere and C, N moves the object towards or away from the camera. Can be fun fun to get the hang of but very useful after that.

None of which solves the "Azimuth Angle of tangent". Sniper has all the maths sorted on that one to give you just how long a piece of track you need at what radius to give you exactly the right number of degrees of curvature required to change heading from one bearing to another.
http://forums.flightsim.com/vbts/showthread.php?t=267932&highlight=radius

Ross