[_o_OOOO_oo-Kanawha]

Thanks for fixing the superelevated turnouts in x1539. At least in the player train path, the diverging track of a turnout is not superelevated anymore using my WP 3rd Sub ScaleRail .stf file.

I think it is generally a good idea to only have the path of the player train superelevated. That leaves all of the other tracks in place.
Perhaps in a later phase, the paths of AI trains within visual range, or in approach of, the player train could be superelevated as well.
Still later, when proper track profiles have been sorted out, an automatic routine could select the proper .ace files from original track the train is travelling over.

Don't know wether that will save on resources and CPU cycles, set off against the extra calculations and logic determining the paths.

Now that California Zephyr glides real smooth through the serpentine Feather River canyon, gently leaning into each and every curve. The ScaleRail track looks real good as well. Only some small gaps remain.
Are these caused by rounding errors, or is my texture alignment slightly off?
Only quibble are those dancing shadows when the headlight is on. Are the 'normals' in my file properly set up? Or is it a deficiency of the still experimental dynamictrack.cs and superelevation.cs routines when multiple light sources have to be taken into account?

So far, Open Rails has shown impressive progess from my point of view, and being able to browse the source code, I think I even understand how it is done. :dance3: :D ;) :cool3: :good2:

[markus_GE]

As superelevation seems to work from the "outside" now, I got one more question:

Will it once be possible, that similar to the cab sway option one could turn on that the cab is kinda "turned" when running through superelevated curves?

I know this already has been dicussed, maybe even in this thread, but I think it has to be talked about once more: from the inside, that cool option looks a bit odd - carbody tilting, cab not, eh?

Yours Markus


BTW: Can´t be mentioned once too often: great feature!!!!! ;)

[_o_OOOO_oo-Kanawha]

No doubt it will be applied in the future, Markus.

The engineer's viewpoint must stay level and solid, only the cab needs to lean a little into the curve.
Could pose a few problems with the alignment of the instruments and controls in a live cab, from where you actually handle your engine by reaching for those levers and buttons.
Having a static cab lean a little is no problem IMO, but a cab full of controls?

So far we have no live '3D' cabs in OR and that feature will probably have to be implemented first.
There are still more important things to implement or improve, more eyecandy can wait from my part.

[markus_GE]

Yeah, can imagine from my own programming experiences (actually i´m envy on the people writing OR ;) )

I just thought of that again today when running through a curve and changing to the head out view: cab = horizontal, body = tilted. But I can live with it :D

[_o_OOOO_oo-Kanawha]

markus1996, on 05 April 2013 - 10:35 AM, said:

Yeah, can imagine from my own programming experiences (actually i´m envy on the people writing OR ;) )

I just thought of that again today when running through a curve and changing to the head out view: cab = horizontal, body = tilted. But I can live with it :D

If you are a programmer and into trains, why don't you join them?

Superelevation and cab sway/tilt is implemented in Railworks 4/Train Simulator 2013, but there are few routes with proper superelevated track.
I like the Open Rails approach and think it is well done so far for the existing routes. No doubt it will be refined over time.

We do need more track profiles, esp. UK FineScale and DB Tracks.
I got the rail shape from DB Tracks correct, but am still struggling with the ties, ballast and rail fasteners. Norbert Rieger has promised to supply Open Rails with proper DB Tracks profiles. Currently, it seems converting all of those X-Tracks pieces to DB Tracks keeps him fully occupied. So one should better wait for his official release.

I have no experience with UK FineScale. Can somebody recommend a route that uses it? Does it use bullhead rail with those wooden kegs? Getting those textures properly spaced is a PITA.

[Csantucci]

 _o_OOOO_oo-Kanawha, on 05 April 2013 - 09:54 AM, said:

I think it is generally a good idea to only have the path of the player train superelevated. That leaves all of the other tracks in place.

Sorry, but I don't agree with this. On double track it's very nice to see also the incoming train superelevated, see e.g. my Youtube clip at minute 2:50.

[JTang]

 _o_OOOO_oo-Kanawha, on 05 April 2013 - 10:14 AM, said:

No doubt it will be applied in the future, Markus.

The engineer's viewpoint must stay level and solid, only the cab needs to lean a little into the curve.
Could pose a few problems with the alignment of the instruments and controls in a live cab, from where you actually handle your engine by reaching for those levers and buttons.
Having a static cab lean a little is no problem IMO, but a cab full of controls?

So far we have no live '3D' cabs in OR and that feature will probably have to be implemented first.
There are still more important things to implement or improve, more eyecandy can wait from my part.

Check this video:

[_o_OOOO_oo-Kanawha]

I understand, Carlo. Of course it looks strange when one train leans into the curve while the other doesn't.

That's why I wrote that perhaps the paths of AI trains within visual range should be superelevated as well. I don't care too much for 'static' track being superelevated or not as I would rather not have all that rusty ScaleRail yard track suddenly turn shiny.

What actually happens with 2t pieces? There isn't that much of it on the WP 3rd Sub, are both tracks superelevated in one go? That perhaps settles your wish?

[_o_OOOO_oo-Kanawha]

 JTang, on 05 April 2013 - 11:14 AM, said:

Check this video:

Sorry, JT. I didn't know of that.
There are a lot of cabs around, and I do prefer those that are photo-realistic.
It would probably require permission and cooperation from their authors if the 3D cab cannot be applied generic.

The 3D cabs are better saved for another thread, to keep this thread focused on superelevation.

[markus_GE]

 _o_OOOO_oo-Kanawha, on 05 April 2013 - 11:07 AM, said:

If you are a programmer and into trains, why don't you join them?

Well, there´s just one problem: I´m 16 years of age, and all i can do i learned all on my own - tht´s just some basic skills in BASIC...

But maybe i´ll join them later on.. I´m planning to - after high school - possibly study software development (but it´s WAAAYYY to go ´till i´ll be able to do anything like the OR-code) ;)

Just wrote that line, ´cause i think I know the problems occurring while of coding (i can´t even write working buttons...)

[_o_OOOO_oo-Kanawha]

Talent has no age, Markus. I should have known from the 1996 in your nickname you are still a student.

Anyway, it is not the actual programming, but rather the idea and how to implement it that makes original and creative minds stand out. Programming is a skill that can be learned by anybody who is determined and committed.

Young people like you are not yet fixed by their roots and think 'out of the box' more freely.

You do read http://www.tssf.eu/ since it is in German and more focused on Germany and its neighbours?
I am reasonably fluent in German also, but have not posted there yet. It is where Norbert Rieger has his own thread on DB Tracks, MS Tiefbau in German.
Hopefully somebody will pick up on a .stf track profile for superelevated DB Tracks.

(for non-German speakers, Google Translate will provide you with a coarse translation.)

[markus_GE]

 _o_OOOO_oo-Kanawha, on 05 April 2013 - 11:47 AM, said:

Talent has no age, Markus. I should have known from the 1996 in your nickname you are still a student.

Anyway, it is not the actual programming, but rather the idea and how to implement it that makes original and creative minds stand out. Programming is a skill that can be learned by anybody who is determined and committed.

Young people like you are not yet fixed by their roots and think 'out of the box' more freely.

You do read http://www.tssf.eu/ since it is in German and more focused on Germany and its neighbours?
I am reasonably fluent in German also, but have not posted there yet. It is where Norbert Rieger has his own thread on DB Tracks, MS Tiefbau in German.
Hopefully somebody will pick up on a .stf track profile for superelevated DB Tracks.

(for non-German speakers, Google Translate will provide you with a coarse translation.)

We´ll see, what time will tell - may skills have way to go, and OR too, sowhos knows? ;)
Nope, I actually don´t read much in german/austrian forums, ´cause my real "love" are the aerican RR things (might also be a bit about bigger, stroger, heavier :D ) - That´s also why I hang out at Elvas: You get to talk to people who can help you, and also know about the real american things, which i just know from my PC :/

[That Genset Foamer]

Apologies for the bump and I'm unsure as to whether this has been stated, but what are the Openrails algorithms for Superelevation transition, and would support of certain algorithms (i.e. the "2/3ds Rule") be possible?

Algorithm: One of the formulas my engineering class was given for a superelevation transition from tangent to curve is L = 150*D*e, where D is the distance from the centerline to the lane edge. In ORTS, D would be the distance from the center line of the track to 1/2 the gauge, and E would be the percent-grade of superelevation - (Level/Grade).

Assuming these variables aren't in use, this could be how transition length is computed.


SPElevation = [Super Elevation Level]/[Gauge]
SPDistance = [Gauge]/ 2
SPTransition = 150 * SPD * SPE


2/3ds Rule: This specifies that 2/3rds of the superelevation must occur on the tangent preceding the curve, and the remaining 1/3 will occur on the subsequent curve. Of course, a whole wealth of issues are introduced in this. It would mean the Superelevation script would need to have a 6m tangent preceding the curve at its lowest setting and 9.5m at the highest setting, or to parse the preceding x meters of track and and apply the transformation.

Placement could be done by detecting the difference between "profiled" track and tangent track, and profiling the preceding / following SPT * (2 / 3) meters. If that distance is on a curve it could sum the transition in degrees into the other curve, but I'm not sure what kind of issues that'd introduce.

[vince]

I use multiple decreasing radii 1 degree track sections to simulate a spiral entry and exit on a curve.
i.e. ===straight===,9,8,7,6,5,5,5,5,5,6,7,8,9,===straight===.

Would such an algorithm handle the multiple sections?

regards,
vince

[That Genset Foamer]

 vince, on 30 October 2014 - 01:50 PM, said:

I use multiple decreasing radii 1 degree track sections to simulate a spiral entry and exit on a curve.
i.e. ===straight===,9,8,7,6,5,5,5,5,5,6,7,8,9,===straight===.

Would such an algorithm handle the multiple sections?

regards,
vince

I use something similar, the issue is such a system doesn't exist outside ScaleRail unless someone generates...say...USTracks analogues to all the ScaleRail shapes. According to Tim Booth's rules that would require a Dynatrax license though. Then of course most routes using the Kuju profile do not have this either.