How About a Poly Rating System for Routes?

S

sethmcs

Active member
Wouldn't it be nice to have some sort of rating system to inform members the expected performance of a route on a low medium and high end computer before a member downloads a route? Maybe a color code system:

Green - expect decent frame rates on low end computers.:)
Yellow - Will work but expect some slow spots:confused:
Red - you need a cray computer to run this sucker.:eek:

Any ideas on how to develop this kind of standard?
 
Yes absolutely in favour of this.

One way would be to select a reasonably challenging built-in route (I'm talking TRS2006 for this example) and scenario say Toronto, run it through with a fixed camera view, probably external with no view or angle changes from default, and note the Trainz reported frame rate. This would be a standard which could be used as a base line e.g.

"my base line frame rate is 25, here is my new route 'Bembridge to Ventnor' and this route has an average frame rate of 34 when a train runs the length of the route".

This would give all potential down-loaders a chance to asses whether their own PC was likely to be able to run the route.

"my base line is 30, I will definitely be able to run Bembridge to Ventnor'"

or

"my base line is 15, there might be problems"

The snag with predefined categories/colours is how can any one player know what categories to put his route into as they are not likely to know what range of PC power is out there unless they are a keen student of PC construction and graphics capability.

An additional option is to establish a benchmark using products like 3dMark06 which seems to provide some kind of objective benchmark. Whilst it can always be argued that this does not necessarily measure the types of graphics processing that Trainz requires I think it would be pretty close. It is possible to map base line value to 3dMark and develop a reasonable assessment of PC power.

Segy
 
Hi,

An idea that probably is not very good to follow up on would be to sort of custom make a Trainz version with one route in it, and bake in (if that is a good word) a FRAPS like tool into it so one can download this version and use that as a way to test the computer to see how it might respond to this particular route.

It would be a non expandable or user changeable Trainz version, just the one route and working with the Trainz game engine of either TRS04/06/TC (preferable have all three as 3 programs from my point of view) and having one default session that starts as soon as one open the program.
All automatic, all you can do is sit and watch till the end.
Or, have a couple of choices, extern view, cab view, camera view - those is the views we have in Trainz today, and having them as a choice would let a person get an idea how their computer behaves on this route in what view.

This would give everyone that downloads this program/route a chance to have a equal experience when it comes to what happen on the screen - but with various result in the frame rate department due to computer specs.

Having a separate program with all preset would give all a equal start point - using a route included in a Trainz version, even with a session included still allows for differences as the amount of content one have is not alike, the way one opens the test route/session might not be totally the same, and so on.

The problem lays in the making of this default test program/route - and the cost associated with it.
But, I think it could be done - it should been done years ago in my opinion - it could even have be used as the Trainz demo program except one would have no surveyor in it then. :)

Either way, it was just one girl's opinion. :)

Best wishes all

Linda
 
Hello Linda, no, that is an excellent girl's suggestion and I totally agree that something like this should have been considered many months ago, and I appreciate what sethmcs was thinking when he wrote his thread. However, one little flaw in this topic is that many of us play our Trainz on laptops and as they were never really developed to withstand this type of simulator game specifics from the outset, frame rates are pretty poor on nearly all routes because of the underpower. Depending on any given route my frame rates can hover between 12 - 24 fps, but someone using a proper PC hard-drive may get a steady 30 fps on the very same layout. I will never complain personally though, as it was my decision to play this sim on a laptop, so whatever fps I obtain on a route I have to accept, and I can cherry-pick my favourite layouts that offer me the best fps, to enable me to enjoy my hobby.

Cheerz. ex-railwayman.
 
Shortline2 said:
Hi,

An idea that probably is not very good to follow up on would be to sort of custom make a Trainz version with one route in it, and bake in (if that is a good word) a FRAPS like tool into it so one can download this version and use that as a way to test the computer to see how it might respond to this particular route.

It would be a non expandable or user changeable Trainz version, just the one route and working with the Trainz game engine of either TRS04/06/TC (preferable have all three as 3 programs from my point of view) and having one default session that starts as soon as one open the program.
All automatic, all you can do is sit and watch till the end.
Or, have a couple of choices, extern view, cab view, camera view - those is the views we have in Trainz today, and having them as a choice would let a person get an idea how their computer behaves on this route in what view.

This would give everyone that downloads this program/route a chance to have a equal experience when it comes to what happen on the screen - but with various result in the frame rate department due to computer specs.

Having a separate program with all preset would give all a equal start point - using a route included in a Trainz version, even with a session included still allows for differences as the amount of content one have is not alike, the way one opens the test route/session might not be totally the same, and so on.

The problem lays in the making of this default test program/route - and the cost associated with it.
But, I think it could be done - it should been done years ago in my opinion - it could even have be used as the Trainz demo program except one would have no surveyor in it then. :)

Either way, it was just one girl's opinion. :)

Best wishes all

Linda
Click to expand...

The idea I like very but it would also need to have the performance settings set to a particular level.

So basically it would need a setup .bat which would set the performance levels to the default or prederetermined setting, then load up fraps, finally start up Trainz to run the session.

Actually no omit fraps and go for the average fps from the jet log. This gives a 'value' for the computer.

Sounds doable to me. Then layouts would be calibrated in a similar way either rough and ready or by using a session.

I like the idea. Some computers / layouts respond better to different performance settings than others but as a rough guide I like it.

Cheerio John
 
You'd definitely need a standard route with a standard length train in it...

You couldn't do it any other way due to the different amounts of stock that people may place on a given route...

The more trains, the less frames. see the problem with poly-rating routes now?

regards

Harry
 
Any Rating System However Imperfect is Better than None

Any rating system is better than none.

Here is an Idea:

Test a route with minimum specs. The ones on the box that comes with the game. There is your low end.

Test a route with mid range computer. There is your mid range.

Test with High end. There is your high range.

Form a test committee of volunteers who get perhaps a free first class ticket to participate provided that they provide the specs of the compter and the test results to Auran who could make the determination of the rating. OR if Auran is far too busy developing the new Trainz of the future than have the Route Builder themselves submit the route to Test Committee for rating. Then you would have this:

Color Coded Routes:
White - Non Rated
Green - Suitable for low end computers
Yellow - Suitable for Mid Range computers
Red - High end only

The rating committee would have its own thread that could be viewed on the forum by members who could add input from there experience with the route to validate the rating.

The added bonus would be a sort of beta testing of routes so that route builders could fix errors and submit higher quality routes.

I dunno what do you think?
 
Although ideas like this have been bandied about forever, and they aren't without merit, the implementation is fraught with disaster. The wide range of various computer builds regarding the hardware and software installed and the base settings of each system, not to mention the ongoing processes of each one would render such a ratings system undependable at best.

If you are intent on creating such a system, first you need to define the parameters of what constitutes a low, mid level and high end system. I think after scratching your heads over that for awhile, you'll realize how difficult such an undertaking can be.

Ed
 
Last edited:
If you are intend on creating such a system, first you need to define the parameters of what constitutes a low, mid level and high end system. I think after scratching your heads over that for awhile, you'll realize how difficult such an undertaking can be.

Ed[/quote]

The reality is that trainz members have low, mid and high end systems now.

Whats wrong with this: Open XYZ thread

I have reviewed XYZ route with a Athlon 3200+ 1.5 Ghz Ram 6800GT and found that I got 15 fps for most of the session. The lowest fps 9 going through the city and the highest 32 in farm area. Nice driving experience. Beautful Route.

I have reviewed XYZ route with Athon 1Ghz 512K Ram 6200 and found that I got 7 fps for most of the session. The lowest fps was 0:'( it crashed. definately not a green route.

etc...etc
 
So you are, in effect defining....what? A high end system and a low end system......or a mid level system and a low end system?

By the every nature of what you have typed; you have begun to define the parameters. Now you need to make certain your parameters are consistant with everyone ele's opinions.

Certainly a range of systems will have to BE DEFINED to be able to give the user of the ratings system some type of inkling of how it applies to them.

Ed
 
Euphod said:
So you are, in effect defining....what? A high end system and a low end system......or a mid level system and a low end system?

By the every nature of what you have typed; you have begun to define the parameters. Now you need to make certain your parameters are consistant with everyone ele's opinions.

Certainly a range of systems will have to BE DEFINED to be able to give the user of the ratings system some type of inkling of how it applies to them.

Ed
Click to expand...

I Agree. What do you suggest?
 
I'm sorry.....it's a great idea, but I suggest it would be of little use because of the wide range of computer builds, hardware, operating systems, software, and system settings.

A public "review" board could be more useful, but the value of the findings would be limited to the extent the reviewer is willing to document his or her system.

In other words; let's assume you set up a website, and every week you recommend a route to download and review. You state your own system specs to the most complete extent possible and post your review. Then you invite others to do the same. Over time, owners of many varied machines and specs may respond and write their own reviews, and supply their own system specs. Eventually, if enough people post their own reviews, a user curious about a route could read all the information, and perhaps find a review by a reviewer who has similar system specs to their own. That user MAY decide that based on the reviewer's system, the route may run in a similar fashion for them.

That approach is not withoout problems however, running a website is not easy, and especially not easy when others are allowed to post on them. It would require endless attention and moderation, and, in the end, may be of little use if not enough reviewers respond, or a wide enough range of systems is not represented.

The problem lies in the endless permutations of putting hardware and software together. It's one thing to report "I have a Core 2 Duo system with such and such MB and a GeForce 8800 Ultra GPU", and quite another to report the system wattage, temperature under load, background processes running, sata/non sata interface, and all graphics settings IN ADDITION to all Trainz settings.

I'm not saying your idea is a bad one....it's a great one! I just think it will be impossible to put it into practice in such a way that you can present bonafide reproducable findings and recommendations.

I suppose you can present a "estimated rule of thumb" concerning the performance of certain routes, but the results will end up being more anecdotal than scientific.

Ed
 
Last edited:
Hi all,

Hmm, I feel stupid here - I still will let the post stand - but I just realize I misread the threads subject!
It was about routes polygon rating - not how the computer one have is acting under Trainz, which I read into it.
I said English was hard, didn't I? LOL

Anyway, carry on people, the below is obsolete but I spent time typing it so I let it stand. ;)

-----
I get the feel from above that somehow I wasn't able to make my self understandable - so against better judgment I try again. ;)

* Make a freestanding Trainz program based on Trainz (game engine) and put in a route where you have no control over anything other then what views you want (Cab, Fly-over or track side camera).
* Put that out as a download to install as a free standing program.
* Let people install it, run it, and read out the frame rate from a log the program makes.
* This would give everyone a chance to have a equal starting point and reference frame point - like the above mention 3DMark program.
* This would allow everyone to have a certain idea what their computer can run Trainz at when compared to other people's computer.

Using the default TRS04/06/TC and a route from that might work too, but as mention above, there is so many things that can be done differently getting to that route in question that it might not be totally similar.
Hence, the idea of a stand alone program - which again could be used as sort of a Demo for Trainz too. ;)

Hope I haven't repeated my self to much - not that easy to always get my thoughts down in written (English).

Best wishes all

Linda
 
Last edited:
Hmm... I like this idea, since I have such a dinosaur of a computer this would be useful to me and many others with ancient machines. Problem is somehow integrating it, much less even writing out the programming bits for it.

WileeCoyote:D
 
I think it's much less complicated than people are saying. Linda's English thoughts still hold true.

We are comparing routes, nothing else.

We take a standard route with a session and get a number from the jet log about fps on your own computer. The reason you take the jet log number is it is an average over the session not a point in time fraps number. Also running fraps affects the frame rate on a low end machine.

Now to compare it against any other route you simply divide the fps from the standard into the fps you get on the new route. This gives you a ratio. You probably need a session to get the repeatability.

If the route runs twice as fast it is low end. If it runs at half the speed it is high end.

On a low end computer the second route will still run twice as fast as the standard route, the third route will still run at half the speed of the high end route.

Now for the magic on the high end computer the second route will still run twice as fast as the standard route, the third route will still run at half the speed of the high end route.

It doesn't matter what the computers are. For comparing routes you don't need to know. It will be approximate because some routes will respond to a specific hardware configuration than others but it's a lot better than nothing.

We can also use the system to compare hardware but that is more complex as we have seen. You need a lot more data but it could be done. For that you would need the absolute fps not the relative fps.

Cheerio John
 
I understand your POV John, but that is still apt to be fraught with errors. What section of each route do you choose for your comparison.....or are you suggesting that the reviewer run the whole of each route to aquire the ratio you seek? Each camera method will deliver different FPS averages, and I imagine using faulty items could skew the numbers as well, and, of course, if you must run the whole route, you may have to reverse through an area to take a different branch.

As you have said, it won't be anything other than an estimate in the end.

I say fairly useless information, given the fact that a low end machine may run pretty much every route badly, while a high end machine may run every route well. There will only be a difference to use to create your ratio up to a point. I contend that a Duo Core machine with an 8800 Ultra will max out to system capabilities, providing no ratio at all.

Of course, I was married, so I very well could be mistaken again.

I'm sure with enough discussion we could make this happen.

Ed
 
Good Idea John......

johnwhelan said:
I think it's much less complicated than people are saying. Linda's English thoughts still hold true.

We are comparing routes, nothing else.

We take a standard route with a session and get a number from the jet log about fps on your own computer. The reason you take the jet log number is it is an average over the session not a point in time fraps number. Also running fraps affects the frame rate on a low end machine.

Now to compare it against any other route you simply divide the fps from the standard into the fps you get on the new route. This gives you a ratio. You probably need a session to get the repeatability.

If the route runs twice as fast it is low end. If it runs at half the speed it is high end.

On a low end computer the second route will still run twice as fast as the standard route, the third route will still run at half the speed of the high end route.

Now for the magic on the high end computer the second route will still run twice as fast as the standard route, the third route will still run at half the speed of the high end route.

It doesn't matter what the computers are. For comparing routes you don't need to know. It will be approximate because some routes will respond to a specific hardware configuration than others but it's a lot better than nothing.

We can also use the system to compare hardware but that is more complex as we have seen. You need a lot more data but it could be done. For that you would need the absolute fps not the relative fps.

Cheerio John
Click to expand...

Good Idea using the jetlog for base comparison. Still the basic vision is discussion forum where each route has a thread and people could comment on their performance experience with the route or, if you like, their overall experience with the route.

Ed, I am talk about creating another place where you could disagree with everything. You would love it.:hehe:
 
Euphod said:
I understand your POV John, but that is still apt to be fraught with errors. What section of each route do you choose for your comparison.....or are you suggesting that the reviewer run the whole of each route to aquire the ratio you seek? Each camera method will deliver different FPS averages, and I imagine using faulty items could skew the numbers as well, and, of course, if you must run the whole route, you may have to reverse through an area to take a different branch.

As you have said, it won't be anything other than an estimate in the end.

I say fairly useless information, given the fact that a low end machine may run pretty much every route badly, while a high end machine may run every route well. There will only be a difference to use to create your ratio up to a point. I contend that a Duo Core machine with an 8800 Ultra will max out to system capabilities, providing no ratio at all.

Of course, I was married, so I very well could be mistaken again.

I'm sure with enough discussion we could make this happen.

Ed
Click to expand...

You need a number of routes with a known session to calibrate this thing. Then on a mid range system you find a ratio for three groups. You also need some sort of windows script to automate the process. The automation is important.

After that it's either a sorted list which is difficult on the forum unless you do some thing like UK layouts and have multiple messages that you can spread into or you simply start three threads, first one simple layouts suitable for low end machines, mid range, and high detailed ones.

In fact that might be the simplest idea.

Cheerio John
 
You must log in or register to reply here.

Similar threads

J
My new $2K gamer is a furnace!
2
Replies
22
Views
1K
Paulsw2
P
n3v_laurence
  • Locked
Trainz Plus Beta TLR Phase 2 - 127542
2
Replies
35
Views
3K
n3v_laurence
n3v_laurence
n3v_laurence
  • Locked
Trainz Plus Beta TLR Phase 2 - 127161
Replies
19
Views
2K
n3v_laurence
n3v_laurence
J
An Open Letter to N3V
9 10 11
Replies
212
Views
30K
boleyd
B
T
Trainz Plus 116912 ready for testing
4 5 6
Replies
111
Views
14K
Sanek17
S
Back
Top