FPS Question, need lots of answers please!

[Pappy44]

As far as I know, KORD is the busiest airport in the world.


.

IIRC, KORD has lost its busiest airport in the world to Atlanta's INTL airport the last couple of years.

I think it might be the biggest in terms of size though.

[ATI_7500]

Brett, having everything about FPS is the wrong approach.
I prefer a reference for FSX users with new systems, in form of a database of settings and results.


E.g:
You choose your hardware from a list and the database determines of someone has the same or a very similar setup and then lists the person's settings and tweaks, e.g in game settings, fsx.cfg tweaks, drivers, traffic, installed texture mods (any resized textures?), degree of smoothness (from 1 to 5) and blurries (1 to 3).
You now have a starting point for further tweaking and as soon as you're done, you submit your results to the database.

By that, we could have an ever growing reference on how to set up FSX properly for the best (looks, smoothness, sharpness) result.

[Brett_Henderson]

I think we're agreeing and not knowing it

[justpassingthrough]

I think Brett is onto something but it may have some issues to think out

The only way to truly benchmark the software for hardware compares is to have base flights set up in which the default scenery must be used and a predetermined set of slider and checkboxes laid out

I think the flights would need to be SAVED and a posted to ensure all conditions are met EXACT including the aircraft, lighting, time, airport, weather, etc, etc


It would also mean the FSX.cfg file COULD NOT have ANY edits in it, AT ALL. In order to benchmark the scenery.cfg and fsx.cfg must be default and allowed to be written without any tweaks

Now for the bad news

We all know FSX needs tweaks to run right, including driver tweaks. So the benchmark would still not be right. Each system (proc, mem amount and speed, FSB, video card, HDD system) is different and I know from personal experience I can get DOUBLE the performance out of FSX by making changes to the config so I am not sure if a benchmark would really show truth in what a system can and can not do.

[ATI_7500]

We all know FSX needs tweaks to run right, including driver tweaks. So the benchmark would still not be right. Each system (proc, mem amount and speed, FSB, video card, HDD system) is different and I know from personal experience I can get DOUBLE the performance out of FSX by making changes to the config so I am not sure if a benchmark would really show truth in what a system can and can not do.

That's why I'm all against FPS values which are quite subjective as well. For me, 18 FPS are more than okay, for someone else, 18 FPS are disastrously low.

Then again, drop the smoothness and blurriness rating. Let those into the database via remarks.

[Brett_Henderson]

Exactly  :)   But 18FPS is a number, not an opinion. It's a very real indicator of how the computer is handling the load.

My whole point is...  a FPS number is real, actual, tangible...  You can't say, "In my opinion, that's 18fps", because it IS 18fps.

I wouldn't want a data-base to help people hit a target frame-rate..  I just think FPS are a useful barometer, and with a situational, consistant benchmark for reporting them, you have a real reference.

If your computer runs the benchmark at XX FPS and you see in the data-base where a very simiar machine benches at XX+5 FPS, you've got something to work with... investigate their tweaks, drivers or slightly different hardware..

The whole key is that as many variables as possible be eliminated, before a person reports their FPS.

[ATI_7500]

It's a very real indicator of how the computer is handling the load.

It's 18 FPS LOCKED, not 18 FPS peak.

You can't say, "In my opinion, that's 18fps", because it IS 18fps.

It IS 18 FPS, but while person a says "that's way too low!", person b says "perfectly fine".

I wouldn't want a data-base to help people hit a target frame-rate..

I've learned two things from FSX: A sim is about smoothness and eyecandy. Get both to the maximum extent possible for you and you're fine. No matter if your FPS counter shows 18 locked or 25 locked.

If you still think, 18 locked is too low, fine. Drop your settings a bit and up the lock to 22 or more.

If your computer runs the benchmark at XX FPS and you see in the data-base where a very simiar machine benches at XX+5 FPS, you've got something to work with... investigate their tweaks, drivers or slightly different hardware..

Settings!

The guy running xx+5 FPS problably hates autogen...or AI traffic...or water...



You know what would be creepy?
If that idea really turned into reality...

[Brett_Henderson]

Let's experiment..

[aussiewannabe]

I've learned two things from FSX: A sim is about smoothness and eyecandy.

Boy, ain't that the truth!!!

[Brett_Henderson]

If you still think, 18 locked is too low, fine. Drop your settings a bit and up the lock to 22 or more.

Agreed !  This not to tell someone what they like.. it's just so you can compare your hardware and settings to others .. See how a similar computer might do better with a tweak or a change... and go from there ..

[Brett_Henderson]

It's not so we can say, "18FPS is good or bad" .. This isn't about what a good frame-rate is.. it's about comparing hardware configurations by a meaningful standard...

[Brett_Henderson]

For example.. I'm looking to upgrade..

[ATI_7500]

My FPS won't be better in KORD, but then again, I run FSX nearly maxed out.


Edit: No scores in the database!

[Brett_Henderson]

My FPS won't be better in KORD, but then again, I run FSX nearly maxed out.

Yes... but if you set your settings to the SimV standard and run the test.. I'll have a good idea of how much better your hardware will be.. compared to what I have now..


It's not to see who can score higher.. just to see how different hardware perform under the same test.

[justpassingthrough]

[quote]Exactly