The "fix" was to allow some (way) older GPUs to load more than 512Mb textures. What it does now is allow a flood of texture information rather than a smooth feed of textures, that's why we see the drop in frames and the stuttering with the fix with the lightening example I showed. There is a "small" increase in VAS used as Christopher found out, but unfortunately his system is operating near the VAS limit and so the results will be inconclusive. What you get is stutters and run into the VAS limit earlier with HIMEMFIX=1 - it's that simple.

Not sure I agree with that. Have not had stutters for the last 2 years. And no freezes with the t7 either

Nor have I! No stutters, no missing textures, no freezes - nothing.

 

Let's not obfuscate the reason I came onto the thread - I am still saying testing requires a logical process of elimination.

I did have this annoying issues, I resolved it by unselecting "Allow Applications to take exclusive control of this device" in my realtek sound drivers settings.

I know it's strange sound driver and video screen freeze, but it did fix it. 

Regards

 

Chris

Thanks for your input Chris.

 

If one comes onto a support forum with a problem, than it is only fair to be sure of eliminating all possibilities. The product should be tested only with fundamentally necessary components. Setting system wide affinity, running weather engines, adding scenery, or making GPU modifying settings, all have to be eliminated. That’s the key. Some people have difficulty with that concept, and it does seem at a glance that they are also the ones with problems. None of my machines performed better with the HIGHMEMFIX, and usually showed graphics anomalies of some kind. I wouldn't use it.

Steve,

 

To be fair, adding the "HIGHMEMFIX=1" statement to my fsx.cfg file appears to have solved at least one graphics anomaly that I was experiencing, so it isn't clear cut.

It is a fix.

 

Microsoft FORGOT to put HIGHMEMFIX in there.

 

 

 

You might as well delete everything from the .cfg......if you think everything in there is just a tweak!

 

 

 

 

But anyway, this thread is about freezes!

It would be nice if we can get back to that.

 

 

I would like to know if anyone else has been able to fix freezes by reducing the load on core 0.

By whatever means....(affinity of Utility Software, lowering AI traffic, lowering autogen, whatever.)

Retested with/without FSUIPC and it’s around 15Mb VAS with the unregistered version. I had to be sure to cold boot and start up in the exact same way, and avoid changing views or touching anything to get a proper reading. The casual with/without test result was completely flawed and way-out. If the time has moved on a different set of circumstances arises, different AI schedule, different weather etc. These and other inconsistencies bring a varying collection of results. Christopher’s initial test of highemefix brought in VAS Limit Chimes and missing textures, but then the subsequent test without, revealed some chimes and some other slow to load textures. I think it’s too early to say what in fact is really going on there, and a conclusion can't be drawn from it early on.

Nothing new that FSX constantly changes.

 

It is dynamic and thus the same flight twice under the same condition on the same PC will not result in the same fps, VAS use, etc.

 

That is why FSX specialist say: it is not comparable to a game and you cant benchmark it.

...So that can be minimised by paring back the sim and starting from a .FLT to improve repeatable consistency.

Please guys, when will SP1 be ready, we're impatient

no idea....but in the mean time, by going back to the basics I was able to pinpoint the bottleneck on my system.

 

Basics for me means:

 

- a clean Win7 install and

- clean FSX install with

- no tweaks in the .cfg other than Wideviewaspect=true (removing this is senseless because even if it were the cause of freezes I would never fly FSX again if I had to leave it out)

- The required by PMDG, HIGHMEMFIX=1 entry in the .cfg (removing this is senceless also because even if it were the cause, I (and many others) could not fly the PMDG777 (and other complex aircraft!) due to missing textures if I would have to leave it out)

- minimal addons ; only PMDG777 and OPUS weather engine.

 

Leaving the 777 out during testing is senseless because IF PMDG were to blame for the freezes, I would quit using FSX if I could not use PMDG aircraft. My main goal IS PMDG aircraft. Testing without OPUS was done and although that stopped the freezes, so did disabeling AI traffic. So the goal became....how to stop freezes WITH the PMDG777 aircraft AND the OPUS weather engine?

 

I was able to solve my freezes (it seems) by moving things (OPUS) away from core 0 with the task manager.

 

If you are also happy to just fly FSX with the default C172 and an FSx weather theme, then by all means test like that.

 

Just try moving things from core 0 it at least.

Basics for me means:

- a clean Win7 install and

before you go that path  might try give this  a go might save  you some time if  it   doesn't  than you might as well go  for  a clean  re install than

http://forum.avsim.net/topic/412561-reinstalling-win7-without-re-installing-fsx/

My freezes seem to happen when the hydraulic system is being challenged: starting up the engines, flap change, ...

hopefully  with the upcoming  release of the sp1  it might  solve the freezes  that some users  are having one  can hope

Spinning Displays bug, Episode 4 last night <_<