Here's an interesting discovery I made whilst replying to joshmeister in another forum thread...
In the latest TRS2019 beta build 96191, go to Content Manager and select the built-in SP SD40T-2 locomotive and then view it in Inspect Mode in Preview Asset...
Looks good, huh?
Now switch the mode to 'Performance Analysis' and see if your rig "canna handle the strain, Jimmie".
In this build and 96000, plus the latest SP3 HF1 beta build in T:ANE, my physical memory use rockets up to >16+ Gb; System Commit to over 63 Gb and CPU utilisation on my i7-4790K goes off the scale. GPU is not too fussed... only moderate use.
I do get a lot of locos stretched across the screen as expected. And System Commit memory use keeps rising until I run out of virtual/ pagefile memory.
BUT - when I return to Inspect mode, the program CTDs and memory returns to around 4.4 Gb.
Suspect that this might be a recently-introduced scripting issue associated with the auto-numbering script; hence the excessive memory and CPU loads.
I went way back to T:ANE SP1 (Build 84204) and got hundreds of SP SD40T-2s marching across my screen in Performance Analysis mode with only a modest amount of CPU/ GPU and RAM utilisation.
My rig didn't protest the load at all this time around.
No crash upon return to Content Manager or Inspect view.
I ran each test at the highest visual quality and performance settings, which is how I run my sessions.
My question therefore is - Are you seeing similar behaviour with this asset (and its bigger brother, the SP SD45) in TRS2019? Or T:ANE SP3?
Neither of these locos has been modified by me in any way - they are current built-ins...
In the latest TRS2019 beta build 96191, go to Content Manager and select the built-in SP SD40T-2 locomotive and then view it in Inspect Mode in Preview Asset...
Looks good, huh?
Now switch the mode to 'Performance Analysis' and see if your rig "canna handle the strain, Jimmie".
In this build and 96000, plus the latest SP3 HF1 beta build in T:ANE, my physical memory use rockets up to >16+ Gb; System Commit to over 63 Gb and CPU utilisation on my i7-4790K goes off the scale. GPU is not too fussed... only moderate use.
I do get a lot of locos stretched across the screen as expected. And System Commit memory use keeps rising until I run out of virtual/ pagefile memory.
BUT - when I return to Inspect mode, the program CTDs and memory returns to around 4.4 Gb.
Suspect that this might be a recently-introduced scripting issue associated with the auto-numbering script; hence the excessive memory and CPU loads.
I went way back to T:ANE SP1 (Build 84204) and got hundreds of SP SD40T-2s marching across my screen in Performance Analysis mode with only a modest amount of CPU/ GPU and RAM utilisation.
My rig didn't protest the load at all this time around.
No crash upon return to Content Manager or Inspect view.
I ran each test at the highest visual quality and performance settings, which is how I run my sessions.
My question therefore is - Are you seeing similar behaviour with this asset (and its bigger brother, the SP SD45) in TRS2019? Or T:ANE SP3?
Neither of these locos has been modified by me in any way - they are current built-ins...
Last edited: