But how do they know rain effects will affect performance by 50% unless they've actually built the model?

That's what I couldn't figure out. lol

Ok, so I thought, just for a few laughs, I'd see what sort of an effect on framerates FSX VC rain effects have.So I took the Tecnam Sierra, deleted the rain effects and recompiled the model. Here are the facts:Model with rain FX : 71351 trianglesModel without rain FX : 48183 trianglesFor those that know the Sierra it has very curved windows which uses quite a few triangles. The Sierra has a total of 5 windows, the front window, two on the canopy (which are separate from the others as the canopy opens and shuts) and two small rear windows.Framerates: Model with rain FX, no weather: 43.7fpsModel without rain FX, no weather: 45.0fpsModel with rain FX, gray and rainy: 34.0fpsModel without rain FX, gray and rainy: 35.5fpsConclusion: Adding rain FX results in about a 3% drop in framerates.Question remains: Where does Aerosoft get their 50% figure from?

Wow Ant, that's a good comparison. I still need to learn about rain effects, I have one question, are the effects created in the 3D model software eg. Gmaxor are they transparent bitmaps that are placed on the model of the canopy, now knowing their effect on an aircraft such as the sierra gives me inspiration to look deeper into this effect and play around with it.

Question remains: Where does Aerosoft get their 50% figure from?

Where Aerosoft gets all their other figures from... :wink:BTW Thanks again for your LSA. It is absolutely epic! :( :( :(

Wow Ant, that's a good comparison. I still need to learn about rain effects, I have one question, are the effects created in the 3D model software eg. Gmaxor are they transparent bitmaps that are placed on the model of the canopy, now knowing their effect on an aircraft such as the sierra gives me inspiration to look deeper into this effect and play around with it.

Tim, first you make 16 transparent bitmaps which have the rain drops on them (takes forever)Then you apply each bitmap to a copy of the window so you need 16 copies of the window.You then apply attachpoint visibility conditions to each window and create a counter that counts from 1 to 16 when it is raining. Using the counter you control when each of the 16 windows appears giving the effect of animation. Repeat the process for the fast moving raindrops.In your model you will have 33 windows. 1 for the normal window, 16 for slow moving rain effects and 16 for fast moving but with the visibility conditions only a maximum of 2 will ever be drawn at once.

Thanks Ant, I will play around with this oneday and see what results I get. It dosent seem too difficult to achieve code wise just time consuming creating the textures.

As it has been pointed out, VC rain effects are one of the features that have been removed from FSX SDK, so that's not a good example for "things that are easier to do in FSX than in FS9". Actually, it's harder to achieve this particular effect than in FS9.The real answer for the original question is: each extra feature costs time and money to implement. With every product, developers have to decide what to include in the package, given the time and financial constraints they have. No addon is perfect, and none has all the features possible. Performance hit is a good excuse, quite probable, but sometimes a bit exaggerated. :(   

A VC with a classic steam six-pack will always get far greater performance than a full glass VC.  Doesn't matter what system you run them on.  Yet people want to take something like the J41 which is a relatively simple panel setup and compare it to something that's got a great deal more going (glass, lots of it).

I wouldn't call J41 cockpit a classic steam six-pack. It has 5 glass displays, FMC, plus full set of analog gauges and plenty of buttons. Compared to Airbus A320 VC, there is roughly the same number of glass, clickable and animated parts. Not to mention very high resolution textures. They are both in the same league of modern, two-seater cockpits, plus PMDG is more detailed simulation.

Bill, meet Anthony aka Ant. He made the Tecnam you liked? :( And a Tiger Moth, too!

Oh, I know Ant... I was guilty of not actually paying attention to the username... :(

Bill, thank you for the clarification regarding the vc rain effects. As a person not actually developing add-ons, but rather just seeing the finished work, it is easy to take these additions for granted.

No problem. I had figured out the "system" within a few minutes of installing the marvelous Lotus model. As Anthony wrote, the "hard part" is creating the 32 required textures. But, like the XML script required, it need only be done once, then may be implemented on any subsequent models one creates.Modelers and programmers who don't pay attention to their fellow modeler's and programmer's work are doing themselves -and ultimately their "customers"- a great disservice. I highly appreciate Ant taking the time to post some relative comparisons of the impact of "vc rain" on his Tecnam, as it does provide some sort of baseline data. Of course, that doesn't necessarily mean that another model can or will achieve a similarly low impact, as there are many other variables at work. Any model's performance is ultimately an synthesis of many, many factors, not all of which are easily predictable -much less measurable- until the alpha testing begins...Realistically, the major and most fundamental difference between an FS9 and FSX model is that for FS9 the emphasis is on keeping the total triangles count as low as possible (polygon loading), and for FSX the emphasis is on keeping the total number of drawcalls as low as possible (texture loading).