fsx-ms Captain Sim L1011 vs Justflight Pro

[Bjoern 1,153]

Just tried a short lap around BGR in the L-1011 Pro. Didn't even get to the turn on final and received an OOM.

With bog standard default scenery.

And FSXSE.

In (fixed) DirectX10 mode.

 

My disappointment is as huge as the lack of commercial success for the real thing.

 

 

 

But with 400+ MB of textures to process per session and 700 drawcalls for the VC model (= lots of VAS use), I was a bit of a fool to expect a different outcome.

 

Too bad. This bird had a bit of promise.

[haskell 298]

Thanks for the report Bjoern. I was JUST ABOUT to buy it. 60% off is tempting. But I haven't had good luck with JF so I'll take your experience as a nice warning.

[Bjoern 1,153]

You're welcome.

 

I would have requested a refund, but this is sadly not possible for downloaded software.

[Martyn JF 388]

Hi Bjoern. We'll definitely investigate the issue but we've only received a couple of reports of performance-related issues in the few years that we've been selling it, so I can only assume that not everyone encounters OOMs (as the developer of the 'Pro' upgrade, I spent many many hours flying the aircraft and I didn't encounter a single OOM with a relatively modest PC). The 'Pro' upgrade added a ton of animations and visibility tags to the engineers panel so that would have increased draw-calls significantly, but obviously it shouldn't be to the point where the product is not usable. 

 

Regarding refunds, please contact our customer service team and they'll be happy to provide you with a refund.

[Bjoern 1,153]

Hi Bjoern. We'll definitely investigate the issue but we've only received a couple of reports of performance-related issues in the few years that we've been selling it, so I can only assume that not everyone encounters OOMs (as the developer of the 'Pro' upgrade, I spent many many hours flying the aircraft and I didn't encounter a single OOM with a relatively modest PC). The 'Pro' upgrade added a ton of animations and visibility tags to the engineers panel so that would have increased draw-calls significantly, but obviously it shouldn't be to the point where the product is not usable. 

 

Regarding refunds, please contact our customer service team and they'll be happy to provide you with a refund.

 

Hi Martyn, thanks for replying.

 

The computer hardware isn't the key to the issue as framerates in the L-1011 were good. It's the amount of memory FSX can manage and this has a fixed upper limit regardless of computer hardware installed. So if you run FSX with default scenery, no AI traffic and medium to high slider settings, you're perfectly able to run the L-1011 Pro all day long, but as soon as there's other stuff that requires a chunk of managed memory, such as autogen, AI and a custom scenery or high resolution environment textures, the margin of available memory gets narrower and chances are that the memory limit will be encountered.

 

With complex models featuring a lot of animations, the problem is that FSX creates a new draw call for the animated part, with associated texture files. So with a 4 MB texture file for the diffuse, specular and bump texture each and a lot of animated switches in the cockpit, the amount of memory claimed by the cockpit model alone will be huge.

 

For me as a end user, the only workaround is deleteting any specular and bump maps from the texture folder, shrink the VC diffuse textures from 2048x2048 px to 1024x1024 px and then convert them to DXT1 format for maximal memory conservation. Not a pretty solution though and the model file will still look for the missing textures, meaning some excess CPU/GPU cycles.

 

So if you're still in touch with the model developer, providing a "lite" interior model version without any specular or bump maps in the materials and textures at half resolution could help lower the general memory use for users with problems.

 

 

And since we're already talking, here's some minor things I've found:

 

- There are some draw order issues in the VC with switches hidden below transparent covers. The switches will not be drawn when the plexiglass cover is in front of it.

- The main annunciator's "electrical system" light stays illuminated despite all systems being a-ok and running. In a short test, I could fix this by replacing the "||" in the associated XML gauge file with an "or".

- I don't think that the anti-skid system may only be engaged when the parking brakes are off. In other aircraft with such a system, the switch merely arms it, with the actual anti-skid system engaging and disengaging automatically as long as it's armed. Granted, this could be a quirk of the real L-1011, but I'd be very surprised if the crew had to toggle the button each and every time before or after they used the parking brake.

- The EPR bugs are very clumsy to operate since every single one has to be dialed from 0 up to the target value. Starting the range of valid values at 1 and implementing a "match the other two bugs to the current one" function (via middle mouse click) would help here.

- The installer refuses to install the L-1011 to an outside directory (unless it finds a fsx.exe in it).

 

 

And a general question:

- Does the L-1011 feature a save system for the custom variables used by the switches to facilitate loading a flight? I could not find any reference to this in the manual.

 

 

Regarding the refund, I'll wait and see how this conversation develops.

[Bjoern 1,153]

Deleting the spec and bump maps from the texture folders frees up some 200-300 MB of VAS. Was not able to try the DXT1 conversion for the remaining textures just yet.

[Bjoern 1,153]

Spent the last few days testing the 1011 Pro on my laptop trying to cut down on VAS usage. Converted all textures to DXT1 and replaced the bump and spec maps with very small placeholders to avoid FSX complaining. The available VAS still dropped from 2.2 GB initial to 1.7 GB after fifteen minutes of flying the pattern. Too little if used in a "real", add-on equipped FSX enviroment (the one on the laptop is default).

Then I tried DirectX 10 mode and regardless of texture format and size, available VAS stayed around 2 GB during my test flight and I could "complete" (crash land short of the runway due to a lack of suitable input controller) it. This begs the question why I did get an OOM error on my desktop PC in DX10 mode. Does video memory (or lack thereof) contribute to OOM issues?

 

(The pendulum swings towards keeping the TriStar at the moment.)

[Martyn JF 388]

Hi Bjoern,

 

Does the L-1011 feature a save system for the custom variables used by the switches to facilitate loading a flight? I could not find any reference to this in the manual.

 

Unfortunately I don't think that it does, unlike our newer aircraft.

 

I have begun looking into the issues that you've reported but unfortunately I will now be away for a week, so I'll be unable to continue until I get back. I'll provide you with an update on the progress on my return.

 

Thanks

Martyn

[Bjoern 1,153]

Unfortunately I don't think that it does, unlike our newer aircraft.

 

I have begun looking into the issues that you've reported but unfortunately I will now be away for a week, so I'll be unable to continue until I get back. I'll provide you with an update on the progress on my return.

 

Hi Martyn,

 

the lack of load/save functionality isn't much of a problem as long as the major systems stuff is handled by XML gauges whose variables are pretty easy to access. It would just have been nice to have it out of the box since creating one myself is quite time consuming.

[ThomasAH 61]

Bjoern, I'm looking forward to your creations, like you did for the Metro 3

[qqwertz 770]

I finally was too curious and bought the upgrade for the JF Tristar Pro. To be frank, I am disappointed. There are still many switches that don't work, and the textures of the flight engineer's panel are so low-res that one cannot read the text of many gauges. IMHO, the CS Tristar is ahead by a margin.

 

Peter

[theophile 2]

I own both the JF L1011-500 and the CS L1011-100.They are both decent aircraft,but my preference is to the

 CS version.Let me first point out the positive things on the JF with respect to the CS:

There is a galley smoke test

There is a brake temperature test

The altimeter must be set manually below 18000 feet

You can effectively detach the DLC

The chronometer is good:up to many hours of flight can be recorded (no chronometer on the CS)

The oxygen masks can be tested

Now on the down side (JF vs CS):

The slat monitor test is incorrect

There is no side window heat

There is no external power

There is no alt antiskid test

The lighting in the dawn and dusk zones is very bad

The INS lines up extremely slowly

Fire test lights do not light up

There is no turb cooling air light

Wheel well fire test is incorrect

No oil quantity test

APU works without fuel supply

COM frequency logic ( if you go up from 124.95,you get 125,i.e. the integers and the decimals are not independent)

No functioning weather radar

the AP VS switch goes in the wrong direction

no voice recorder test

only the main door opens (in P3D v4.5,which I use)

incomplete light test

you cannot disconnect the AP from the yoke

No sunscreens 

No payload/fuel configurator

 

As far as the flight model is concerned,both are probably not very correct,but the JF version of a landing with

DLC is totally wrong.I was a real life pilot myself,and I would never ever fly a plane with these landing characteristics.The CS version is landable with DLC,although it feels somewhat rigid.

I love the Tristar,and in my opinion CS has succeeded better in simulating this wonderful aircraft.

[edetroit 933]

....... and neither one of them features the distinctive Hamilton Sundstrand FMS. Amazing! 😲

[Doug47 574]

FMC? Ugghh

[edetroit 933]

6 hours ago, Doug47 said:

FMC? Ugghh

Pretty sure this thread is about L-1011 simulation. At least Aerosim attempted it 15 years ago. 😄

Edited July 3, 2020 by edetroit