GTN Trainer 6.62.3 vs 6.21 (resolved)

[RXP 2,338]

21 hours ago, Kaan KASIM said:

for GSYNC to work if I remember correctly.

This might be the key then: what about creating a NVidia Profile for "GTN Simulator.exe" and disable GSync and maybe any 'optimization' setting?

[Kaan KASIM 54]

GSYNC was my first thought too, already tried that before but it doesn't help. Tried disabling both systemwide and app only.

[Kaan KASIM 54]

I bought rxp gtn for x-plane and the clickspot issue I had on p3d with v2.5.23 happens in both virtual cockpit and the popup window, same as in rxp for flightsim.

Edited March 17, 2020 by Kaan KASIM

[RXP 2,338]

I get on your system, GTN trainer 6.21 is smoother, and I'll continue investigating this afterward.

But for the clickspots scaling issue, I'm lost now with your post in the GTN XPlane forum and I'd like you to help me understand.

You current setup now is using 2.5.23 gauges and plugins, but 2.5.22 rxpGtnSim.dll (ProgramData) ?

In other words:

1. You've solved the clickspot scaling issue just using the previous rxpGtnSim.dll version (2.5.22)?
2. The previous rxpGtnSim.dll (2.5.22) also solves this clickspot issue whether using GTN Trainer 6.21 or 6.62.3?

The file versions print on each RXP log file first line and reflect exactly which file is in use. Can you please confirm all these version numbers reading the RXP log files in "Documents"?

What is puzzling me even more is that for now I fail to find any possible cause when looking at the changes between 2.5.22 and 2.5.23. Also, there is no other similar report so far which leads me to think it is something only on your system which is doing this. This makes it very hard to point you in the correct direction for now but we're looking. 

 

 

Edited March 18, 2020 by RXP

[Kaan KASIM 54]

2 hours ago, RXP said:

• You've solved the clickspot scaling issue just using the previous rxpGtnSim.dll version (2.5.22)?
• The previous rxpGtnSim.dll (2.5.22) also solves this clickspot issue whether using GTN Trainer 6.21 or 6.62.3?

1. That is true but the gauges were from v2.5.22 too

2. Clickspot issue only happens with 6.21, I had no problems with 6.62.3 trainer except the performance. 

From the logs,

"rxpGtnSim64.dll version 2.5.22.2"

"rxpGTN.dll version 2.5.23.0"

"win.xpl version 2.5.23.0"

---

2 hours ago, RXP said:

You current setup now is using 2.5.23 gauges and plugins, but 2.5.22 rxpGtnSim.dll (ProgramData) ?

This was only the case for X-Plane, since I couldn't find an older installer. But you are right, after you mentioned this I backed up Programdata folder, upgraded RXP and replaced the programdata files with 2.5.22.0, no issue whatsoever. The gauge files are not the issue at least, just programdata dlls.

 

So to summarize I had been using Trainer 6.21 and rxp v2.5.22 when I purchased it. After the update came out and I installed it I noticed the click issue and downgraded after I uninstalled RXP. The issue comes back after I update to latest RXP, doesn't matter If I uninstall it before or upgrade it. 

I hope I am not being even more confusing, trying my best to explain the situation.

 

Cheers!

Edited March 18, 2020 by Kaan KASIM

[RXP 2,338]

2 hours ago, Kaan KASIM said:

"rxpGtnSim64.dll version 2.5.22.2"

"rxpGTN.dll version 2.5.23.0"

"win.xpl version 2.5.23.0"

Log files are indicating you're using the latest gauge and plugin version (2.5.23) with the previous rxpGtnSim.dll (2.5.22.2).

 

2 hours ago, Kaan KASIM said:

hope I am not being even more confusing, trying my best to explain the situation....

Clickspot issue only happens with 6.21, I had no problems with 6.62.3 trainer except the performance. 

I'm confused because I always thought you had both performance and clickspot problems with GTN Trainer 6.62.3.

If I'm still following:

1. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.23    + trainer 6.21 = clickspot problem
2. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.22.2 + trainer 6.21 = no clickspot problem
3. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.23    + trainer 6.62.3 = no clickspot problem but trainer performance problem.

Is this correct?

 

[Kaan KASIM 54]

19 minutes ago, RXP said:

Log files are indicating you're using the latest gauge and plugin version (2.5.23) with the previous rxpGtnSim.dll (2.5.22.2).

 

That's right, but it wasn't the case before I bought X-Plane version. They were all 2.5.22.

 

20 minutes ago, RXP said:

I'm confused because I always thought you had both performance and clickspot problems with GTN Trainer 6.62.3.

If I'm still following:

1. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.23    + trainer 6.21 = clickspot problem
2. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.22.2 + trainer 6.21 = no clickspot problem
3. gauge/plugin 2.5.23 + rxpGtnSim.dll 2.5.23    + trainer 6.62.3 = no clickspot problem but trainer performance problem.

Is this correct?

 

Sorry If I somehow caused that confusion.

They are all correct, I'll just add one more. I only mixed versions unknowingly after buying the X-Plane version.

• gauge/plugin 2.5.22 + rxpGtnSim.dll 2.5.22    + trainer 6.21 = no clickspot problem

[RXP 2,338]

OK, this is why we were unsuccessful searching nonexistent bug in the code!

Now we can focus our search on what is causing the clickspot specifically with 6.21 vs 6.62.3.

As for the performance problems it is hard to say yet but what is certain is Garmin is implementing their multithreading differently since 6.41 and furthermore, they are using a compiler with a known bug in this area... We're doing our best to alleviate as much as possible already and correct known bugs from the inside out, but I'll continue looking at options for you to further try if you agree.

[Kaan KASIM 54]

Sure, just tell me what to do. 

 

24 minutes ago, RXP said:

We're doing our best to alleviate as much as possible already and correct known bugs from the inside out

I guess this is why the newest trainer runs better in the sim, I still can't wrap my head around how the standalone trainer can run at 2 fps... I tried contacting Garmin but they just told me to reboot. 😄

[RXP 2,338]

update: we've found the culprit for the clickspot problem you're having with Garmin trainer 6.21 and this will be fixed in the next udpate ASAP.

NB: this is affecting every customer using trainer 6.21.

As for performance, I'll update this discussion if I find anything meaningful but I strongly suspect the effect is due to the use of a different threading architecture introduced with trainer 6.41 and it is inherent, although I'm looking whether there could be external (i.e. Operating System) ramifications too.

[Bernhard777 1]

Hello,

is there a solution in the meantime for the slow running GTN trainer 6.62? Coz I have the same problem as Kaan, my trainer is also soooooo slow and sluggish. It is the trainer itself which is so slow - without runnig P3D. Using Trainer V6.2 is ok but not 6.62. 

Sorry to misuse this thread, I'm a F1 user, but this speed problem here with the trainer is like my problem and F1 has no interest in helping with the trainer.

i7-8700K , 16GB , GTX1080Ti, 11GB VRAM, Win10 V2004, P3DV5

Bernhard

Edited April 27, 2020 by RXP

[Bernhard777 1]

Ok, I just found a discussion in the simforums of F1. Must really be a problem with the 2004 version of win10.

Edited April 27, 2020 by RXP

[RXP 2,338]

Hi,

Can you please follow up and let me know whether the latest RXP GTN 2.5.24 update is running better on your systems?

Edited May 5, 2020 by RXP

[Kaan KASIM 54]

Hi,

Been trying the latest RXP on both p3d and xp the last few days and compared to https://imgur.com/b5IUmKR this, It's much better now.

Still not as smooth as trainer Trainer 6.21, but I am happy with the latest trainer now.

Edited May 10, 2020 by Kaan KASIM

[RXP 2,338]

@Kaan KASIM Hi, you might want to try the just released 2.5.25 which further includes specific GTN trainer 6.62.3 enhancements: