PMDG 777 SP1 VAS usage

[voodoo101 21]

Before anybody jumps, I know that the PMDG is not the cause of VAS, I also know how to limit the problems with OOMS. As I explained on another forum, FSX is like a tub of water (VAS memory) and the addons are buckets of water. As you add each bucket of water, the water rises and eventually the bath is full and can't cope, read OOM. People normally assume that the last addon they installed is the cause which is why so many people are moaning about different addons.

 

Now with the advent of OrbX global, OrbX vector, HD airports and such wonderful high quality textures in the 777, this limit is quickly reached. Now you can turn down the textures but after all the years of trying to get the textures up by tweaking and buying better hardware, it seems a shame to go back down again. Yes you can turn off certain airports sceneries but again it's a pain. FSX is a 2006 program that we are trying to make look like a 2014 program with all the bells and whistles.

 

So now to my question which is quite a basic one, although the PMDG 777 is not the cause, it does have a high VAS usage, has anything been done to lower this is SP1 or has the beta team noticed a difference. Unfortunately the PMDG 777 does tend to tip the scale, partly because of the high textures, partly because it is a long hauler and so fly's over more scenery for FSX to save up.

 

Please don't let this get into a heated debate on why PMDG has not fixed this or that it's PMDG's fault. If the answer is no then fine, like I said FSX is a 2006 program and was never designed to run this long and never really finished.

[Simicro 247]

Hi Craig,

 

You will find some answers there http://forum.avsim.net/topic/444846-sp1-beta-testers-performance-differences/

[scandinavian13 5,192]

In the interest of not leaving it so open ended such that someone is wondering "why are there so many views but no answers," I'll provide this:

 

I saw the thread.  Read it earlier.  Decided not to respond.

 

The reason I decided not to respond earlier stems from basic stat and scientific method.  I don't have hard numbers from before SP1, and and don't have any hard numbers afterwards.  I simply don't watch VAS that much unless I get the VAS Ding, and by that point, who knows what the issue is.  It could be the route of flight, but there's no way to pin that down, either, unless you've been tracking that for the flights that went off without the VAS Ding.

 

The rest of it would simply be notional, but that's not reliable.  It wouldn't be fair for me to say "VAS use has been improved in SP1" unless I have hard evidence of that fact.  Even a notional drop in getting the ding wouldn't necessarily mean it came from SP1 unless you flew the same route, with the same weather, at the same time of day, and remained in one view the whole flight.  To truly pin it down, you'd need to be testing specifically for VAS use.  I don't think any of us have done that.  As such, I don't think it would be appropriate for us to speak to it.

 

On the next build, I might set the RTM up somewhere and test it, and then the SP1 build, but only if I remember and have time.  I doubt it's changed significantly enough to matter, really.

[voodoo101 21]

Thank you Kyle for your answer. I don't normally worry too much about VAS but a recent reinstall made me wonder on what I could lose to help with the arrival of new and improved scenery and extra's in the past. The only time I have had a problem was once when I tried a flight from OMDB to KLAX, both had quality sceneries.

 

As I said before, it is not really a case of a certain addon but a whole general thing added together, the same as frame rates, each setup is different.

 

I was just wondering if there was any mention from PMDG or anybody knew if they had maybe cleaned up the textures or maybe lost a line of code that was not needed but left in. In my old days of programming on the ZX Spectrum in basic, I often wrote something then found out that I could take a chunk of code to be able to do the same thing with less. I'm sure when it is released there will be a group sharing their view one way or another.

[Boeing or not going 578]

My question is are the testers getting OOM errors when flying? I don't care about graphs and stats, I'd like to know if the testers have to chase OOM errors like we do using the release version of the 777X.

[scandinavian13 5,192]

My question is are the testers getting OOM errors when flying? I don't care about graphs and stats, I'd like to know if the testers have to chase OOM errors like we do using the release version of the 777X.

 

Read my earlier post...

Unless we're flying the same exact routes, with the same exact situations, any answer here would be anecdotal at best.  I don't care much for "butt dyno" answers ("butt dyno" is a term used in the car modding group to basically say "it felt like it was performing better, but I have no hard evidence to this fact," as would come from a real dynamometer).  You may not want graphs and stats, but I'd prefer not to give answers that aren't backed by data somehow.  

[voodoo101 21]

I would think that if the testers where getting OOMS when testing then PMDG would not be due to release the update. The same as the original release. Some things are not picked up in beta testing because it is on a relatively small scale and everybody's FSX configuration and loading is different. What you have one your system, what your config, what you PC set up is going to be different to mine or Kyle's, because FSX is so modular. I might fly from stock airports over Orbx textures with no weather, You might fly from Aerosoft airports over UTX textures in bad weather. It would be different if PMDG did some texture editing to reduce memory, or maybe suggested to the beta team that they had done something to stop a memory leak. It appears that they have not or have and not mentioned it, which is not surprising considering the special extra's they have added. Only time will tell how it runs on your own system with your add-ons and flying the routes you do.

[Tabs 2,790]

There have been some optimizations made both on the texture/model and code sides, but it's not going to be anything huge. There was an option added to stop sounds from precaching into memory that may be useful for people having OOM issues too.

 

VAS usage isn't caused by one particular texture or line of code or anything like that though - it's a cumulative thing and the big drivers are the functions in the airplane that have to store large amounts of temporary data in memory and perform calculations on it. This is stuff like the ECL, the EICAS, the VNAV and LNAV solvers, the FBW system etc. There's only so much we can do to reduce memory usage in those systems before functionality necessarily suffers.

[777simmer 117]

There have been some optimizations made both on the texture/model and code sides, but it's not going to be anything huge. There was an option added to stop sounds from precaching into memory that may be useful for people having OOM issues too.

 

VAS usage isn't caused by one particular texture or line of code or anything like that though - it's a cumulative thing and the big drivers are the functions in the airplane that have to store large amounts of temporary data in memory and perform calculations on it. This is stuff like the ECL, the EICAS, the VNAV and LNAV solvers, the FBW system etc. There's only so much we can do to reduce memory usage in those systems before functionality necessarily suffers.

Thanks Ryan, that is the answere I was waiting for.

 

"Improvements have been made.

But nothing huge."

 

I will take any VAS improvement, no matter how little :-)

 

Thanks for taking our VAS problems serious and thx for trying!

[hsung 3]

Can the beta testers confirm if exiting from PMDG 777 to the main menu before exiting FSX still cause a freeze and crash OR is this fixed?

 

This problem also currently occurs when loading another saved game from a current PDMG 777 session.

[voodoo101 21]

There is a work around for the crash on exit and it is noted in the "PMDG 777 tracking thread". It states to delete runways.csv from the FSX folder, if it is there. Also ARPT_RWY.dat in the FSX\PMDG\NAVDATA folder.

 

This post has been up since the 777 was released, so it might be worth a quick read as it also a sort of changelog for SP1, you will see what has been fixed.

[hsung 3]

There is no runways.csv, but found and deleted ARPT_RWY.dat.  Tested this afterwards, but am still getting a fatal error on exit from main menu ... except this time with some problem details below.

 

Problem Event Name: APPCRASH

 

Application Name: fsx.exe

 

Application Version: 10.0.61637.0

 

Application Timestamp: 46fadb14

 

Fault Module Name: PMDG_777X.DLL

 

Fault Module Version: 1.0.0.0

 

Fault Module Timestamp: 52256029

 

Exception Code: c0000005

 

Exception Offset: 000a467a

 

OS Version: 6.1.7601.2.1.0.256.48

 

Locale ID: 3081

 

Does anyone recognise or know these problem details?

 

Cheers, Harry

[CaptCraig 226]

Ryan's post was very welcome when it comes to this issue of VAS usage and the 777.  I think those of us (so far in this thread) can appreciate the topic.  It really is going to come down to what other candy we are running, and what our system can handle, as it always has.  It's just in these times we have more candy dedicated to the actual aircraft.

 

By the way - if we are already turning off scenery that we're not flying to, how hard is it to turn off HD textures if that's what it takes to fly the 777 for you?  Or maybe turn some "sliders" down.  We really need to learn to balance it out.  

 

Kyle, your post really said nothing you political devil you  :lol:.  I do like how you admitted to getting some OOM dings.  I'd love to hear the story behind yours.  

[James Bennett 52]

There have been some optimizations made both on the texture/model and code sides, but it's not going to be anything huge. There was an option added to stop sounds from precaching into memory that may be useful for people having OOM issues too.

 

VAS usage isn't caused by one particular texture or line of code or anything like that though - it's a cumulative thing and the big drivers are the functions in the airplane that have to store large amounts of temporary data in memory and perform calculations on it. This is stuff like the ECL, the EICAS, the VNAV and LNAV solvers, the FBW system etc. There's only so much we can do to reduce memory usage in those systems before functionality necessarily suffers.

 

Any improvement is a huge improvement. After all, it's the last 1mb that causes the OOM.

[ZKOKQ 984]

Mitch is on the case :lol:  http://forum.avsim.net/topic/445555-try-setting-your-graphic-rez-to-4096-as-your-new-default-setting/