Elvas Tower: Forest Object Density - Elvas Tower

Jump to content

  • 2 Pages +
  • 1
  • 2
  • You cannot start a new topic
  • You cannot reply to this topic

Forest Object Density Rate Topic: -----

#1 User is offline   CGW121 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 533
  • Joined: 29-December 06
  • Gender:Male
  • Location:Genoa, Illinois
  • Country:

Posted 12 August 2011 - 07:19 AM

I just got around to trying Open Rails, and I am impressed. I do not know if this has been addressed on other Forums but I was curious as to if Open Rails will allow for greater Forest Object density?
Thanks

#2 User is offline   James Ross 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 5,488
  • Joined: 30-June 10
  • Gender:Not Telling
  • Simulator:Open Rails
  • Country:

Posted 12 August 2011 - 08:37 AM

View PostCGW121, on 12 August 2011 - 07:19 AM, said:

I just got around to trying Open Rails, and I am impressed. I do not know if this has been addressed on other Forums but I was curious as to if Open Rails will allow for greater Forest Object density?


There doesn't seem to be any limits set in the OR code (I looked at this recently). What limits have been found in MSTS?

#3 User is offline   caldrail 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 588
  • Joined: 14-April 08
  • Gender:Male
  • Country:

Posted 13 August 2011 - 01:35 AM

This was mentioned in a tutorial long ago but I can't locate it anymore. If I remember right, anything above 5000 is pointless as it won't be rendered any denser than that, though I do use larger density values in route building without harm.

#4 User is offline   CGW121 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 533
  • Joined: 29-December 06
  • Gender:Male
  • Location:Genoa, Illinois
  • Country:

Posted 13 August 2011 - 04:46 AM

View Postcaldrail, on 13 August 2011 - 01:35 AM, said:

This was mentioned in a tutorial long ago but I can't locate it anymore. If I remember right, anything above 5000 is pointless as it won't be rendered any denser than that, though I do use larger density values in route building without harm.


Exactly what I mean about msts. Will this change in ORTS? Was thinking about messing with the numbers in the world tile and seeing if it makes a difference in ORTS.

#5 User is offline   James Ross 

  • Open Rails Developer
  • Group: Status: Elite Member
  • Posts: 5,488
  • Joined: 30-June 10
  • Gender:Not Telling
  • Simulator:Open Rails
  • Country:

Posted 13 August 2011 - 01:46 PM

View PostCGW121, on 13 August 2011 - 04:46 AM, said:

Exactly what I mean about msts. Will this change in ORTS? Was thinking about messing with the numbers in the world tile and seeing if it makes a difference in ORTS.


As I said before, there doesn't appear to be any fixed limit set in OR so it should work with higher values... there may be data structure limits though, like 32,768.

#6 User is offline   CGW121 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 533
  • Joined: 29-December 06
  • Gender:Male
  • Location:Genoa, Illinois
  • Country:

Posted 16 August 2011 - 04:51 AM

View PostJames Ross, on 13 August 2011 - 01:46 PM, said:

As I said before, there doesn't appear to be any fixed limit set in OR so it should work with higher values... there may be data structure limits though, like 32,768.


I went into the world tile and changed a value

Forest (
UiD ( 5800 )
TreeTexture ( oak25_2.ace )
ScaleRange ( 0.6 1.4 )
Area ( 207 145.125 )
Population ( 500 )
TreeSize ( 16 18 )
StaticFlags ( 00100000 )
Position ( -5.18713 193.401 -462.263 )
QDirection ( 0 -0.0553678 0 0.998466 )
VDbId ( 5 )
StaticDetailLevel ( 0 )
)

The route editor had the population value set at 80. I had tried changing the value in the world tile before but msts ignored it. I ran past the area in Open Rails and the forest looked like a forest!. ORTS rendered the object at the written levels with no change at all in frame rates. I do imagine that if I edited the tile in msts route editor it would default back to a value of 80.

#7 User is offline   charlie 

  • Engineer
  • Group: Status: First Class
  • Posts: 722
  • Joined: 17-February 04
  • Gender:Male
  • Location:Maryland, USA
  • Simulator:MSTS
  • Country:

Posted 17 August 2011 - 09:42 AM

View PostJames Ross, on 12 August 2011 - 08:37 AM, said:

There doesn't seem to be any limits set in the OR code (I looked at this recently). What limits have been found in MSTS?


For forest items translating into OR, I think there are two present difficulties, but first a little about how they are handled in MSTS. It's probably more complicated than most folks would have thought, and there isn’t a magic cast-in-concrete rule.

Firstly, forest object density is always recalculated by MSTS even if you just leave it alone at the default 50.0 initial setting. For a long time as I observed this happening, I thought that the density was being adjusted based on tile object count, world file size, or some such limiting attribute.

I was wrong, the final calculated density is a function of the size of the forest object "footprint". Want more density? Either overlap forest objects or plant smaller ones. I've done both...

Here's Demo 1. This is on a modest route, the one-tile Dual Gauge Track Demonstration. It has a world file size of 389K and 755 objects. The forest object on the left is stretched to the maximum possible and covers an 8x8 array of terrain patches. The middle object covers a 4x4 array with the small one at 2x2.

I left the density at 50.0, saved and exited. When I popped back into the RE, left to right the new densities were 50.94905, 50.58142, and 50.05504. I edited the densities to 99999.0, saved and exited. When I popped back in, the final densities are 736.26373, 1506.08215, and 2114.77222. Density change is apparent in the JPG.
Attached Image: Lite_Tile.jpg

This is Demo 2, and it has a world file size of 646K and 1,264 objects. Through the same sequential experiment, final densities are 746.00018, 1479.99902, and 2120.62085. Pretty similar to Demo 1...
Attached Image: Hvy_Tile.jpg

So here's a another exp't for today. To Demo 1, I added a forest object covering just one terrain patch, resulting in an indicated 2325.00806 density. Here's a repeat of the approximate RE densities, apparently indicating some diminishing returns for the last two as I reduced size.
From left to right - 736 1506 2114 2325
Here are the “Population” entries from the world file.
From left to right - 737 401 138 36
Attached Image: Population.jpg

Remembering that the forest on the left is stretched to the max, the 2nd has 1/4th the area of the first, 3rd has 1/16th, and last has 1/64th.
I can then rewrite the population entries like so:
(737 x 1) (401 x 4) (138 x 16) (36 x 64) = 737 1604 2208 2304

A pretty fair match to the RE indicated densities, and Population entries are clearly dependent on forest object size. I could probably put the “errors” down to my inability to stretch the forest objects to exactly the size I wanted, and I had not thought to hand-edit the Area entries in the world file for exact size.


So, back to some difficulties. The first difficulty I have is that the RE (or anything else) cannot give a preview of where individual forest object elements are going to be arranged when seen in OR. In the Demo 2 tile above, in the interests of frame rates separately-placed vegetation was banished and replaced with very carefully arranged forest objects. In OR I have trees and bushes all over track and roads and in the river with no way to tinker with the arrangement.

The second difficulty is only my own. :rolleyes: The way my mind works, I would have liked to have some sort of easy algorithm to intelligently set the population entry in the world file. Looking at the Area entries, I think I could write one, but it isn’t necessary. However, arbitrarily editing world files for OR leads me right back to my first difficulty.

One last experiment.
Note that arbitrarily editing all population entries to 1000 (greater than 737) in any of the above demos buys exactly nothing since they are already at MSTS max. In the RE in Demo 1, I edited the forest object densities back to the original 50.0, saved and exited. I edited all of the world file Population entries to 1000. Back in the RE, the resulting indicated densities are exactly to the last decimal place identical to those I reported in Demo 1 above when I used the RE to edit density to 99999.0.

And I think that’s as good as you can do in either MSTS or OR. Looking at the last exp’t in OR, subjectively there’s greater density.

However with the edited world file there still is an obvious density difference between forest areas in OR. As in MSTS, the larger the area covered, the less density you get. Just in case I’m imagining things, it’s somebody else’s turn to look at that.

As for me, I will never do this type of world file editing for OR. Mostly because of the lack of preview capability and also that I can get what I want for density anyway. They will look OK for density in OR.

And I have hundreds. The Demo 2 world file alone has 63, and it ain’t even up in the higher elevations where the real forests are. If I ever finish this project and somebody loads it in OR, it won’t be my fault if vegetation is all over track/roads/river. :wallbash:

regards,
charlie

#8 User is offline   CGW121 

  • Engineer
  • Group: Status: Contributing Member
  • Posts: 533
  • Joined: 29-December 06
  • Gender:Male
  • Location:Genoa, Illinois
  • Country:

Posted 17 August 2011 - 09:55 AM

I edited a couple of world tiles yesterday, changing all forest object population line to 3000. In msts it made no difference visually. On Open Rails it looked like a forest. Trees tightly packed, a carpet of trees. The frame rates dropped to about 30 in those tiles whereas I usually get 70 or 80. Not too sure if I will do more but in that route I will wait until I get the mileposts and other stuff done. Its nioce to see a feal forest tho

#9 User is offline   Shay 5 

  • Superintendant
  • Group: Status: Inactive
  • Posts: 1,476
  • Joined: 12-April 10
  • Gender:Male
  • Location:Back in the sticks of Virginia
  • Simulator:MSTS/ ORTS
  • Country:

Posted 17 August 2011 - 10:23 AM

In what file can you find this? I would like to do this to a route I have, but don't know where it would be.

#10 User is offline   charlie 

  • Engineer
  • Group: Status: First Class
  • Posts: 722
  • Joined: 17-February 04
  • Gender:Male
  • Location:Maryland, USA
  • Simulator:MSTS
  • Country:

Posted 17 August 2011 - 11:26 AM

View PostCGW121, on 17 August 2011 - 09:55 AM, said:

...changing all forest object population line to 3000. In msts it made no difference visually. On Open Rails it looked like a forest...


That's not a surprise if you were already maxed out in MSTS density. Neither is seeing more trees in OR. That situation matches my last experiment above.

I think that any edited Population entry larger than 737 (the apparent maximum) will produce exactly the same result. I just used 1000 to make sure for the tests I was conducting.

Obviously arbitrarily setting everything to the max for OR can affect computing resources. No way to test but cut and try.

For MSTS purposes, before you edit you can check if a forest is maxed out with the method I gave earlier. The forest area ratio normalized to a max sized forest is:
(X times Y)/(1,000,000), where X and Y are the two entries in the Area statement.

It's 1,000,000 square meters because a max sized forest is really (1,000 x 1,000) instead of the (1,024 x 1,024) I thought it was a while back.

Have fun...

regards,
charlie

  • 2 Pages +
  • 1
  • 2
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users