Distant Mountain Scenery And Such

[conrail1]

This post is dedicated to those, if any existing, that think that distant mountains aren't soo important.
You have comparisons of MSTS screenshots and ORTS screenshots taken from same locations in the Bernina freeware route.

Here’s how distant terrain is handled (without performance issues) in a game engine that was written to properly utilize modern hardware -

http://3.bp.blogspot.com/-wm1v7tG_fP8/ULYYN8z4q0I/AAAAAAAAAlw/yCi8NiomfXA/s1600/screen_1354110814.jpg

http://www.dsogaming.com/news/outerr...-environments/

http://www.moddb.com/engines/outerra-engine

http://outerra.blogspot.com/

[seagoon]

With respect to mnjrr and the obelisk, and also to conrail1 and the Outerra engine rendering individual blades of grass at 20 kms, I don't think that level of "distance" is necessary to achieve a Distant Mountains effect.

The most pertinent post here is that of Mr. Santucci and his series of screenshots, that of the Cadera comparison being the perfect example of all that is really needed insofar as DMs for Open Rails.

IMO I don't think that we really need to strive for anything other than to fill in the vacant space.
OR has much more to offer on the operational side and that is where the focus should remain.

Bruce

[conrail1]

With respect to mnjrr and the obelisk, and also to conrail1 and the Outerra engine rendering individual blades of grass at 20 kms, I don't think that level of "distance" is necessary to achieve a Distant Mountains effect.

That’s up to the game developer to decide, if that kind of detail that far out isn’t needed then it’s not used. The point is if the game engine is written properly (again to utilize today’s hardware) distant terrain/scenery isn’t problem and it certainly isn’t in Outerra.

[seagoon]

That’s up to the game developer to decide, .....

I don't believe that is quite correct for Open Rails.
I'm of the opinion that much of what the future holds for the program is based upon the input from we users and supporters of OR.

Undoubtedly the team will have their own agenda but from what I've seen so far they are certainly sympathetic to the suggestions and comments of the users.

I for one would consider it a complete waste of programming time to go to the depth's of detail with the vistas and vegetation produced by the likes of Outerra; we are, after all, talking about a train simulation, not a landscape and terrain simulation.
This is Open Rails, not Railworks.

Bruce

[baldwin]

Trains and landscape are like bread and butter. One is not so good without the other. If time is not spent on developing a convincing landscape generating part of OR then we might as well settle for a tabletop railway set. Agreed that some of the capabilities of Outerra are probably OTT for a train sim. However, having used it, and keeping up with updates, the landscape generation and representation of gravity are probably the best I have seen so far.

[seagoon]

Trains and landscape are like bread and butter. One is not so good without the other.

I don't agree, but then we are all entitled to our opinions and we all have our preferences.
Having spent considerable time over the years with train-sims such as TM4, BVE (all iterations) and Zusi 2, I have learned that the realistic simulation of train operation/handling can be attained with the most basic of graphical enhancements outside the window.
There are times when I consider that the worst thing that ever happened in the development of train sims was the ability to soar like a bloody eagle at 1000 feet.
Very pretty but damn-all to do with train operations.

If time is not spent on developing a convincing landscape generating part of OR then we might as well settle for a tabletop railway set.

That is somewhat of an exaggeration, to put it mildly.
Railworks (Train Simulator 2013) has glitzy graphics yet all the dynamics of a table-top railway set.
There are some glorious routes for Trainz yet take the gloss out and the physics are weak.

Outerra.......However, having used it, and keeping up with updates, the landscape generation and representation of gravity are probably the best I have seen so far.

Exactly. It produces some beautiful landscapes.

Bruce

[ChrisD]

As long as I can see the trees grow out of the ground in front of my train, the draw distance can be improved.

For the users that do not bother, make it adjustable.

ChrisD

[Bananarama]

How a sim looks is important part of what I do, and I'm fairly certain there are other artists who feel the same. A good looking back-end helps the creative juices flow. Granted, Railworks doesn't have the best physics or game play, but the graphics keep me interested in developing for the platform.

I'm also considering developing for Trainz, as there are some great features about it and some of the latest route/model screens look fantastic! I just need to get my head around some of the creation techniques, and also that I don't care much for the Jet-1 engine.

MST-ala-OR is looking interesting, but it lacks certain features that I feel are important from a graphics standpoint (ambient occlusion, normals, etc.), not to mention the horrid editors and other MSTS ball and chains. Future OR developments will only improve, however, which is likely to lead to better models and improved user experience.

[conrail1]

I don't believe that is quite correct for Open Rails.
I'm of the opinion that much of what the future holds for the program is based upon the input from we users and supporters of OR.

Undoubtedly the team will have their own agenda but from what I've seen so far they are certainly sympathetic to the suggestions and comments of the users.

I for one would consider it a complete waste of programming time to go to the depth's of detail with the vistas and vegetation produced by the likes of Outerra; we are, after all, talking about a train simulation, not a landscape and terrain simulation.
This is Open Rails, not Railworks.

Bruce

It has nothing do with being a waste of time as far as programming is concerned and all to do with developing a game engine that is capable of that kind of detail without bringing the house down (as far as performance is concerned).

Outerra has been specifically designed for simulation games so of course it has to be capable to simulate the world in great detail if that’s what a particular game developer is after.

Does a train simulator need detailed grass as far as the eye can see?, no, but the fact that the engine can handle that level of detail while still maintaining excellent performance is an example of its capability and how useful it would be for a rail operations simulation.

[conrail1]

Trains and landscape are like bread and butter. One is not so good without the other. If time is not spent on developing a convincing landscape generating part of OR then we might as well settle for a tabletop railway set.

Or you could just settle for the junk train games we’ve already seen from the last ten + years.

In the real world part of getting T&E qualified on a particular division/sub division is learning the visual characteristics of the territory, so yes getting the terrain/scenery right would be important to a rail operations simulation.

Agreed that some of the capabilities of Outerra are probably OTT for a train sim. However, having used it, and keeping up with updates, the landscape generation and representation of gravity are probably the best I have seen so far.

I’ve been following its development since 2009 and then was invited into Outerra’s closed beta, it’s an impressive game engine to say the least.

[lindsayts]

A couple of points on viewing distance.....

With a camera angle of 45 degrees (somewhat wider than a standard normal fix lens) the width of 1 pixel on a 2560x1600 monitor at 20 kilometres is around 13 metres as a consequence there is little point in trying to render anyhthing smaller.

Another point is OR is meant to be a realistic sim, now I have a pilots licence its __VERY__ rare to have atmospheic conditions that is clear most days even viewing down sun you will not pick up an object shorter than a silo 30 ft diameter and 80 ft high at 20 kilometers. Its not difficult to miss a town even at 10 kilometers when you know its there.

The only thing one an see clearly is the shapes of large objects such as mountains, lakes can be visible depending on the direction of the sun at great distances. When flying back to my home field you can pick up the Winton lake at least 20nm (say 40 kilometres) before one sees the town.

I have flown on plenty of days when one could not see ANY detail beyond around 10 to 15 kilometres below 6000 ft altitude. Note atmospheric muck is usually confined to below 6 or 7000 ft, when one is higher than that one can pick up isolated mountains many kilometres away, sticking above all the brown haze and polution.

From this I personally think its VERY non-realistic to have high detail distant terrian and OR is meant to be a realistic sim.

There's a limit to the total number of triangles (and therefore objects) a single home PC an render you simply CANNOT have high detail on every thing, loco's, rolling stock, track, signals, buildings, trees, shrubs, distant terrain and expect to have anything like a decent frame rate. You have to chose what is important and render the rest in low detail.

Lindsay

[conrail1]

From this I personally think its VERY non-realistic to have high detail distant terrian and OR is meant to be a realistic sim.

As I already said part of getting qualified in the real world is learning the visual characteristics of the territory so yes distant terrain/scenery is important.

Also if high detail terrain isn’t important to a simulation then why would the military be eyeballing Outerra for use in an aviation training simulator?

There's a limit to the total number of triangles (and therefore objects) a single home PC an render you simply CANNOT have high detail on every thing, loco's, rolling stock, track, signals, buildings, trees, shrubs, distant terrain and expect to have anything like a decent frame rate. You have to chose what is important and render the rest in low detail.

You can when the game engine is written to utilize modern hardware. I can fly around in a 172c in Outerra now and see detailed terrain as far as the eye can see at triple digit frame rates.

Again it’s up to the game developer to decide what detail is important and where that detail level is needed.

[seagoon]

Well, if the OR development team can continue giving us excellent performance (high fps and no lag) and also cater to the 20 kilometre obelisks and insects in the weeds, that's fine by me.

But the moment that performance and operational realism is traded for glitz then that's the day that OR loses the plot.

Bruce

[seagoon]

Or you could just settle for the junk train games we’ve already seen from the last ten + years.

Have you not used Run 8?
A far cry from "junk train games" yet the graphics are perfectly acceptable because they quickly become secondary to the operational aspects.

In the real world part of getting T&E qualified on a particular division/sub division is learning the visual characteristics of the territory, so yes getting the terrain/scenery right would be important to a rail operations simulation.

True, but then it is not part of the familiarisation process to know what type of trees are located on a mountainside 8 kilometres away.
The shape and profile of the terrain is what is recognised.
Ninety percent of the familiarisation of the route comes from the course and grade of line itself and what is on either side of it in close proximity.

Bruce

[conrail1]

Have you not used Run 8?

I’ve had it since a week after its release; it’s far from anything I would associate with quality.

The graphics are far from acceptable (another XNA mess) in this day in age (with lack of image quality control in some ways it’s worse than MSTS, if that’s possible lol).

I have T&E experience as well as train dispatcher experience and I wouldn’t go crazy over the operational aspects of Run8 either.

True, but then it is not part of the familiarisation process to know what type of trees are located on a mountainside 8 kilometres away.

As I’ve said before that’s obviously up to the game developer to decide what detail is important and where. If you don’t need detailed grass or trees 20 miles away then you put that capability to good use in the simulation elsewhere.

The game engine is the platform for the game, if the platform can’t handle the weight the games future is limited.

Ninety percent of the familiarisation of the route comes from the course and grade of line itself and what is on either side of it in close proximity.

I’m well aware of what goes into training T&E; I’ve worked in the railroad industry as an instructor.