No Speed Trees!

amtrak2041864

High Speed Rail!!
Lets stop the oppressive rule of speedtrees killing our computers, and how about we post our non-speedtree routes/creations or what ever, here. I will start it off that everyone else can follow.
With your help we can save computers everywhere and not make Trainz routes the next Crisis(1,2 and 3).

amtrak2041864_20131225_0019_zps4d183a29.jpg
amtrak2041864_20131225_0015_zpscca170c7.jpg
 
The new Trainz Italia Rollins Pass NG36 route is a TS12 route.

The route has billboard trees and looks fine. Clam1952 billboards work great.

Why do we need resource draining Speedtrees for people who want to operate?

Harold
 
how does one know if speed trees are used? what is the author's name?

Most of the built-in trees are speed trees, Auran an Pofig are the two most popular speed tree makers, billboard trees are two 2D trees that have two surfaces with pictures of trees on them.
 
I have 'converted' from Speedtrees to billboards.. here's 2 from me...

Winter:
ddyk.jpg




7s79.jpg

Autumn scene... Both spline billboards and static billboards. Each only 2-4 poly's!!

In Ts12, my personal opinion is that billboarded trees look better than speedtrees. This has to do with the light (or lack of) effect so that they produce a very dark appearence. Billboarded trees do not have this wierd effect. What we see in the new Trainz game can be another story and we just have to wait and see. If you make billboarded trees for Ts12 you can get rid of the 'edge' effect by using the free GMax software and apply the correct light settings like it's done her in these 2 shots...


Jan
 
Last edited:
Most of the built-in trees are speed trees, Auran an Pofig are the two most popular speed tree makers, billboard trees are two 2D trees that have two surfaces with pictures of trees on them.

Not strictly correct as you can use more than two planes I use 4 twisted planes.

 
I can feel a rant coming on... :( When I got TRS2010 and opened my route in it, I could hardly believe the havoc which had been wrought. It wasn't simply that my hard work had been, without asking, ruined. Even if I needed to see trees swaying about like dervishes in a 70 mph wind. Why should a Birch be replaced by an American Boxwood in the heart of the UK? Anyway I have banished the evil weeds from my routes and here is the proof
speed trees gone par PinzaC55, on ipernity Here's some video of prototype Speed Trees http://www.youtube.com/watch?v=FqrLqg3w6AU
 
Aaaaaaghrh - SpeedTrees!!!

.
Here’s one from me – totally SpeedTree free!

9hei.jpg




This herbicide Industry has proved to be very helpful.

weedzfree.jpg




If I may be allowed the inclusion of just one (photoshopped) speedtree, here’s a shot showing the effectiveness of WeedzFree™. They instantly run for miles at the merest sniff of this unique product.

Cheers
Casper
:p



2045d87b.jpg
 
here we go again with the speedtree bashing.

to be fair I did a comparative test and here are the results.

1 baseboard with 3220 speedtrees on it.

total buffers = 3
total vertices = 4973
total indices = 23640
materials = 3
speed trees = 3220
speed tree types = 8
scene load = 0.28%

using a one for one replacement with clam1952's tree, I get the following

total buffers = 22
total vertices = 198173
total indices = 641880
materials = 4
scene load = 7.65%

As you can all guess, you'll know what I'll be using and it certainly won't be billboard trees.

Peter
 
Interesting...I was planning on using 2d trees for my project but after gawpo50's post I'm giving this a second thought.
 
here we go again with the speedtree bashing.

to be fair I did a comparative test and here are the results. ~ results snipped~ As you can all guess, you'll know what I'll be using and it certainly won't be billboard trees.

Peter
I'm not sure many of us understand the bottom line of those figures.

What do they look like? A picture paints a thousand words, and this is a screenshot thread after all.

How about a comparative videos of each example so that we can see the differences in stutter rate and artistic merit?
 
I'm not sure many of us understand the bottom line of those figures.

What do they look like? A picture paints a thousand words, and this is a screenshot thread after all.

How about a comparative videos of each example so that we can see the differences in stutter rate and artistic merit?

Same test, more trees from both types. First two images are speedtrees by myself, pay attention to the screenshot of the stats


e255fa1a83ff8eb4bdc32229db786f79.jpg


95be8a9a3e045407ba8076def964e141.jpg


Next two images from Clam1952 trees. Exact one for replacement for identical number of trees. Again pay attention to the first image with stats and then compare both stats.

89e48c30752d0356bce5ee69ca70c7a7.jpg


3bea17ad917ef596b50d7c13f938f033.jpg


The figures to pay particular attention too are, Total Buffers, Vertices, Indices and Scene Load.

The higher these figures go, the more load on the system. Splines are your biggest frame rate killers if not used properly. Certain splines you stay right away from, they will bring any system to it's knees.

Hope this info clarifies things.

Peter
 
Splines are your biggest frame rate killers if not used properly. Certain splines you stay right away from, they will bring any system to it's knees. Hope this info clarifies things.Peter
All splines are framerate killers ... so track should be avoided at all costs, and a rails to trails single texture should be the ideal framerate route:hehe:
 
Well, these comparison test (by Gawpo50) are very...suprising and interresting. So I Wonder why do we (most of the community) still think speed trees sucks more than bilboard ones if it is not the case ? I personnaly use speed trees but, prior to this comparison, I was planning to replace them with billboard ones, but I won't.

Rail4Pete
 
Interesting stats, and what I would expect, I'm under no illusion as to what the game stats are supposedly telling us. As a rough rule of thumb if I try to keep the scenery load under 80% there are not usually any problems, I have more problems with buildings boosting the load rate than with Billboard Trees.

Lets now confuse the issue by real time usage on a route.
Note this was done in 49922.
30 minute AI session using tracking Camera on my route Route as in mine, I have SpeedTree version and non Speedtree Versions

With Speed Trees: Averaged 52fps. With stutter.
With billboard Trees: Averaged 89fps virtually no stutter.

SP1 HF4 frame rates are about 10% lower than 49922 on both my systems here.

The Grumbles.....
One of the reasons I reverted to bill board trees is the SpeedTrees available on the DLS and of the correct type are mostly far too big for my route. The other is that pre HF4 I get better frame rates and no stutter using billboards. With SpeedTrees I got loads of errors in jetlog.txt to do with shaders and such, more than likely the cause of the stutter.
Pofigs trees would be a better size apart from the sparkle effect but as I'm producing a route where everything is on the DLS I opted for something that works and fits.
The peculiar lighting effects in SP1 on SpeedTrees was also very off putting, those that once looked OK now look unrealistic depending on the viewing angle / time of Day. Also the Billboards do not disappear in SP1 as seems to happen with SpeedTrees on occasions.

Interestingly though on a quick check it looks as if the lighting issue has been sorted in HF4 and jetlog is not reporting the errors any longer.

I actually have the Unreal SpeedTree editor, however it's not permitted to use it for anything other than Unreal otherwise I would have sorted out some Trees of appropriate size and varied appearance that don't need clicking on each one to shrink the flipping things, even then some are still from the land of the giants. Yes I'm aware one can tweak the config, however if others are going to use the route that's not a solution. When you are attempting to produce a realistic protypical route you need items that look the part and of approximately the right size and colouring and that don't leap around even with the wind on minimum setting or cause stuttering in addition to that we are stuck with due to the overstretched game engine.
 
Back
Top