Aircraft performance fall off with Win 7? last call for explanation

[b5523482]

I've noticed in another beta that there can be lots of interrupted functions and exceptions. IL-2 also manages to run when the screen is stuck. So there's a lot going on in the background even when nothing happens on screen. A few games lose functions on newer, faster hardware. So indeed it is possible that you now play a more complete game or even a gimped one.

My advice is to test for this by throttling down your CPU either in Win 7 settings, via BIOS or 3rd party software. It might be worth a try.

[sniperton]

Originally Posted by Pugo3

I installed 4.07 from the same install DVD for both computers, and after some time finally patched my later Win 7 set up up to 4.10.1. Being that I noticed the drop off early on while in the familiar 4.07, it doesn't appear to be a error or corrupt install, but thank you.

Is it not due to the changed AI and FM since 4.07 and 4.10.1? Just a question. The 'feel' of the game has changed a lot in the past few years.

[K_Freddie]

Originally Posted by IceFire

The physics might be calculated faster leading to faster frame rates only. You can't take shortcuts here with a slower or faster CPU... otherwise multiplayer would be a write-off. So no... not likely to be the case.

There is a difference between the different maths units on the processor. Intel documents this (1000s of pages). if IL2 chooses FPU/GPU on an XP laptop, and MMX/XMM/SEEx on the new laptop, the results depend on the Options used for each maths unit.

I doubt whether UBI or TD have worked on this yet, with regard to C or C++ compilers as these are not the best solutions, other than absolute control of your compilation process - not offered with C or C++ compilers as of yet.

[ElAurens]

Guys, I think Pugo3 is comparing V4.07 with V4.10.1, yes there will be a rather pronounced change in the way that aircraft behave if that is the case.

Pugo, on the deck dogfighting a Ki 43 with a much heavier Corsair is a formula for failure.

[Woke Up Dead]

Originally Posted by ElAurens

Guys, I think Pugo3 is comparing V4.07 with V4.10.1, yes there will be a rather pronounced change in the way that aircraft behave if that is the case.

That's what I thought too, but he also says this: "Being that I noticed the drop off early on while in the familiar 4.07,..."

It's definitely odd, especially when he mentions that it also affects another flight sim on his computer.

Pugo, there was a recent thread here by Horseback with acceleration charts, see if you can replicate some of his results or if you really are 10-15% slower. It won't change your perception of the aircraft feeling worse than before, but at least you'll know you don't have a real disadvantage.

[Pugo3]

Originally Posted by b5523482

I've noticed in another beta that there can be lots of interrupted functions and exceptions. IL-2 also manages to run when the screen is stuck. So there's a lot going on in the background even when nothing happens on screen. A few games lose functions on newer, faster hardware. So indeed it is possible that you now play a more complete game or even a gimped one.

My advice is to test for this by throttling down your CPU either in Win 7 settings, via BIOS or 3rd party software. It might be worth a try.

Thank you, please advise on how I might try this or perhaps a recommended software to use? Hopefully a simple first attempt as you mentioned by adjustments to the Win7 settings via BIOS - can you walk me through or direct me to a posting/forum, etc?

p3

[Pugo3]

Originally Posted by K_Freddie

There is a difference between the different maths units on the processor. Intel documents this (1000s of pages). if IL2 chooses FPU/GPU on an XP laptop, and MMX/XMM/SEEx on the new laptop, the results depend on the Options used for each maths unit.

I doubt whether UBI or TD have worked on this yet, with regard to C or C++ compilers as these are not the best solutions, other than absolute control of your compilation process - not offered with C or C++ compilers as of yet.

Way beyound my understanding, but basically that different processors can produce different flight interpretations?

[Pugo3]

Originally Posted by ElAurens

Guys, I think Pugo3 is comparing V4.07 with V4.10.1, yes there will be a rather pronounced change in the way that aircraft behave if that is the case.

Pugo, on the deck dogfighting a Ki 43 with a much heavier Corsair is a formula for failure.

Actually, I made my comparisons with 4.07 on both machines, as it was some time before I jumped to 4.10.1.

It was a learned process, mostly not panicing after getting nailed many times, but you can use those beautiful flaps as you begin a low to medium speed dogfight with a Ki-43, along with the throttle inputs to pull you through. It would most often be a series of roll over/under, canopy rolls, rolling siccisors with sprial towers. You could look for an opportunity and gun the trottle to get the H... out of there if things began to look bleak, then come back in for another go. It is this particular dogfight I most lament losing in the performance difference between computers. [I did mention that the AI are not as agressive and nuanced either, somewhat lackluster and not nearly as aggressive and challenging] Perhpas with Win8 in my coming desktop - I can hope...

[Pugo3]

Originally Posted by Woke Up Dead

That's what I thought too, but he also says this: "Being that I noticed the drop off early on while in the familiar 4.07,..."

It's definitely odd, especially when he mentions that it also affects another flight sim on his computer.

Pugo, there was a recent thread here by Horseback with acceleration charts, see if you can replicate some of his results or if you really are 10-15% slower. It won't change your perception of the aircraft feeling worse than before, but at least you'll know you don't have a real disadvantage.

Thank you, I will be reloading my Win 7 Prof and all software next week, so I will try this as soon as I'm up & running.

p3

[K_Freddie]

Originally Posted by Pugo3

Way beyound my understanding, but basically that different processors can produce different flight interpretations?

Not that difficult..

If the game sticks to the FPU maths unit through all cpu's, the difference will be slight, maybe barely noticeable, as even the FPU unit is upgraded in next generation chips.

BUT if it now decides to dump using the FPU for faster XMM maths units, the differences might become apparent, as the XMM units (ect) do not conform completely to the original FPU specifications.

The game is written in C/C++ and Java (if I'm not mistaken) and the maths units underlying features would be hidden from the developer, unless he/she specifically gets into the details of each.

This is just speculation... as I think TD does not have access to the engine, and probably only Oleg and the original devs might have these details of what the engine does.