How About a Poly Rating System for Routes?

I have to agree with Ed on this one. I don't see any way to do this with results that would apply to everyone. There are 2508 layouts on the download station. No telling how many are available on other web sites. I have never bought a store bought computer. I have built my own with what parts I could afford at the time of building. The best thing to do is download a layout and try it. If it won't run on your computer it just means your computer won't run it without some kind of upgrade. Besides, a high end machine today is tomorrows low end machine. My two cents worth.

Jack
 
I have to agree with Ed on this one. I don't see any way to do this with results that would apply to everyone. There are 2508 layouts on the download station. No telling how many are available on other web sites. I have never bought a store bought computer. I have built my own with what parts I could afford at the time of building. The best thing to do is download a layout and try it. If it won't run on your computer it just means your computer won't run it without some kind of upgrade. Besides, a high end machine today is tomorrows low end machine. My two cents worth.

Jack

Some layouts do not have a lot of detailed scenery objects and give very reasonable frame rates on many computers. Some layouts simply die on anything other than the latest high end video card and cpu combination.

No matter what the machine the first type of layout will always run faster. To identify these layouts would be doable no matter which machine you use for benchmarking them.

The problems at the moment are two technical ones, you need repeatability which means you need to exit at the end of a session. Is there a way to do this? Causing an exception which crashes Trainz would be acceptable as long as the average fps are recorded in the jet.log file. The other one is to start the session in a conistant manner. Ideally from a windows script or .bat file that could be fed the layout name and session name, however if you simply started a session manually and any delays in starting it had no impact on the average frames per second in the jet.log fire that wold be acceptable as well.

Cheerio John
 
Yeah, Thats a neat idea but. The FEC Unit kills my FPS.


So if we can start this! I'll put up a low end computer specs.




Dell DIMENSION DIM2400
Intel(R)
Celeron(R) CPU 2.40GHz
2.39 GHz, 512 MB of RAM.

And i get about 20 or 10 FPS in Trainz. :'(
 
Sound s like a great idea but one can only assume also with a low poly route and a old comp that more trains ran gonna eat up your frame rates anyways.....:'(

But like any game the bigger the comp the better results be nice TRAINZ witha XBOX game engine........
 
Does have a definition of a low end machine?

Does have a definition of a low end machine? I have my own ideas but would like input from the group.
 
We don't have to define high/med/lo machines. How about this....

Suppose we (whoever 'we' is) defined a specific session on a specific route viewed in cab mode as a 'standard'.

I run it on my machine and get say 80 fps from jetlog.

I now run the same consist (still in cab view) on my new route and get say 40fps.

My new route is defined as 80/40.

Anybody can run the 'standard' session and from the fps rate they get on that work out ROUGHLY how their system compares to mine. If they get less than about 40 fps on the 'standard' it is prolly not worth their while downloading my route, since they can expect only about 20fps with a single consist run the least demanding view (cab).

OTOH if they get 120 fps on the 'standard' they can expect a reasonable framerate on my new route even with lots of rollingstock in tracking view...

or is that just what Linda said???

Andy :)
 
or is that just what Linda said???

Andy :)

That was my interpretation.

I have some instructions on how to start Trainz in driver so if that works then all we need is some way to stop it.

Anyone know how to crash Trainz in such as way as the jet.log still has the frames per second?

Thanks

Cheerio John
 
Help wanted.

Auran came up with the following .bat file.

*****

cd \Program Files\Auran\TC

bin\Trainz.exe -driver=60349:100002 -intro=disable -1024 -Render=DirectX

*****

Unfortunately on my machine it doesn't work as 60349:100002 doesn't find it's route. However if I change the session kuid to one that is connected then it works.

I need a small layout with everything on the DLS that comes up no missing content with an associated session that automatically starts and runs for say five to ten minutes.

Thanks John
 
It seems to me the answer to this dilema is for everybody to own a high-end system. Now before you get all bent out of shape wondering what makes me think everybody can afford a top notch system, let me remind you there are myriad ways to obtain sufficient funds for such a system. There are always jobs available as greeters at the local Wal-Mart or dishwashers at one of the fast-foods. One could rob a convenience store (but please don't lay your weapon on the counter while you count the money as did some dufus in Houston). As a last ditch effort One could sell the spouse and children into slavery. With a little thought and effort it is possible to obtain funds so you can thoroughly enjoy your hobby by purchasing a system that prevents engines and cars from melting or skipping merrily down the tracks and through the train yards regardless of how much scenery candy the creator has added. ;) And I still need a loaded gondola.
 
It seems to me the answer to this dilema is for everybody to own a high-end system. Now before you get all bent out of shape wondering what makes me think everybody can afford a top notch system, let me remind you there are myriad ways to obtain sufficient funds for such a system. There are always jobs available as greeters at the local Wal-Mart or dishwashers at one of the fast-foods. One could rob a convenience store (but please don't lay your weapon on the counter while you count the money as did some dufus in Houston). As a last ditch effort One could sell the spouse and children into slavery. With a little thought and effort it is possible to obtain funds so you can thoroughly enjoy your hobby by purchasing a system that prevents engines and cars from melting or skipping merrily down the tracks and through the train yards regardless of how much scenery candy the creator has added. ;) And I still need a loaded gondola.

Yes but the problem is which is the best performing expensive bit of hardware. Do you want a low energy or quiet solution.

Be nice help me sort out the benchmarking thing.

http://forums.auran.com/trainz/showthread.php?t=14714

I have a basic one working for TC and it's consistent. Once we have the procedure worked out we can move onto TRS2006 etc and benchmarking routes.

Cheerio John
 
Thanks to Auran we now have a tool for benchmarking. See this thread.

http://forums.auran.com/trainz/showthread.php?t=14993

Basically if you run the standard benchmark to calibrate your machine then run a benchmark using the same technique on the route of interest you can divide the fps by the standard benchmark. This gives an indication if it is lighter load or a heavier load than the benchmark system.

I've only set it up for TC because that is the only one I'm running at the moment but if someone who has TRS2004 or TRS206 can come up with a standard session with .bat file on a standard route but remember all content must be on the DLS so using an Auran route that comes with the game makes the most sense then I think we can do the other platforms as well.

Cheerio John
 
On the face of it this can sound a good idea. My websites are hosted with a server that "stars" the sites. However Trainz is such a wide remit. For example there are train people who have no interest in trams and there are those of us who like trams and building routes. How does one compare a city based tramway system with a a far wider railway system? Not as easy as suggested I may muse?
 
On the face of it this can sound a good idea. My websites are hosted with a server that "stars" the sites. However Trainz is such a wide remit. For example there are train people who have no interest in trams and there are those of us who like trams and building routes. How does one compare a city based tramway system with a a far wider railway system? Not as easy as suggested I may muse?

We are only comparing the frame rate on a layout to the frame rate on a "standard" layout. If your machine runs the Harlem at 25 fps then another layout that some one has run and got the same frame rate perhaps 56 fps on the Harlem benchmark and the test route would suggest your machine would probably give about 25 fps on the test route.

It's simply a way to help people identify which routes are best suited to slower machines and for people with higher end machines to identify which routes have more detail or animated figures in.

Cheerio John
 
While this is a very good and needed idea, just using average fps has one pitfall and its a big one, an average of 30 fps could mean a min/max of 20/40 and all is good, but on some routes there is a lot of detail in the towns and practically none in between so the min/max is more like 2/58, not so good :(

Now the hard part, is there any way to record the min/max fps other than fraps or running in a window ?

Cheers David
 
While this is a very good and needed idea, just using average fps has one pitfall and its a big one, an average of 30 fps could mean a min/max of 20/40 and all is good, but on some routes there is a lot of detail in the towns and practically none in between so the min/max is more like 2/58, not so good :(

Now the hard part, is there any way to record the min/max fps other than fraps or running in a window ?

Cheers David

I agree but it's the best we have at the moment.

Cheerio John
 
While this is a very good and needed idea, just using average fps has one pitfall and its a big one, an average of 30 fps could mean a min/max of 20/40 and all is good, but on some routes there is a lot of detail in the towns and practically none in between so the min/max is more like 2/58, not so good :(

Now the hard part, is there any way to record the min/max fps other than fraps or running in a window ?

Cheers David

Thinking about this one you can use the same method to record the built up areas and the country areas. Just build two sessions, one starting near a town and one in the country. You don't need to run for five minutes that was just a scientific approach way of making sure the sample was large enough for benchmarking hardware. Modify the .bat file to select first one session then the other and you can get a reasonable indication of the fps in different parts of the layout.

Cheerio John
 
Back
Top