Elvas Tower: Smoothing out the Player Path - Elvas Tower

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

Smoothing out the Player Path Steadicam? Rate Topic: -----

#1 User is offline   eolesen 

  • Superintendant
  • Group: Private - Open Rails Developer
  • Posts: 1,546
  • Joined: 30-October 10
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 14 March 2021 - 10:13 AM

As I've been debugging some of my own routes by running trains across them, I'm finding places where there's a very visible jerk or jump in the camera view when running across places where there's a very sudden shift left/right/up/down in the TDB vector. I don't want to blame TSRE for introducing these, but I will say I'm finding it more and more on places where I've done editing and where there's an intended vector change that the editor doesn't always retain as cleanly as it could.

The motion is far more noticeable in a hyrail where there's no bogies to pivot, but even on a B-B locomotive it's apparent.

I'm guessing that the actual deflection is around a railhead's width (less than 10-20 cm) left to right and it rarely continues for more than 10-20cm in length, but the camera is sensitive enough for the view to jog suddenly anywhere from 5-90 degrees left or right of the intended path. It's something I'd expect on a roller coaster, but in the real world, the suspension of a locomotive or passenger car would easily absorb the shock from that type of anomaly.

Would it be possible for an option in the viewer logic to smooth out those variations e.g. some averaging based on the next 1M of travel or maybe a max deflection per frame value (degrees of rotation, movement left/right/up/down?)

#2 User is offline   joe_star 

  • Fireman
  • Group: Status: Active Member
  • Posts: 191
  • Joined: 16-January 13
  • Gender:Male
  • Simulator:MSTS
  • Country:

Posted 16 March 2021 - 02:01 AM

Sadly this is indeed a TSRE introduced artifact. I have this happening even at tracks oriented to form a perfect joint. The effect is as bad as having a 10cm mismatching rail joint using the old MSTS RE.

The effect can be seen on the yellow TDB lines in TSRE.

Removing the TDB vector completely and re-selecting track doesn't seem to work either, or not in 1 go at least. I am tempted to run the route thru MSTS TDB rebuild, but it doesn't load anymore there due to another issue introduced by terrain painting in TSRE, as well as my reluctance due to all the linked signals contained.

I would hate to have this bug "fixed" by additional work in ORTS, but if it improves on the issue I would trupy welcome it

#3 User is offline   eolesen 

  • Superintendant
  • Group: Private - Open Rails Developer
  • Posts: 1,546
  • Joined: 30-October 10
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 16 March 2021 - 05:03 AM

Dunno... I'm extremely grateful to have a new editor.

If there are a few anomalies that can be error corrected in the game engine, it's worth it to me to have all the new content in work.

#4 User is offline   vince 

  • Superintendant
  • Group: Status: Elite Member
  • Posts: 1,312
  • Joined: 18-June 14
  • Gender:Male
  • Location:West of the Contental Divide
  • Simulator:ORTS_Running MSTS_Editing
  • Country:

Posted 16 March 2021 - 01:34 PM

View Postjoe_star, on 16 March 2021 - 02:01 AM, said:

Sadly this is indeed a TSRE introduced artifact. I have this happening even at tracks oriented to form a perfect joint. The effect is as bad as having a 10cm mismatching rail joint using the old MSTS RE. The effect can be seen on the yellow TDB lines in TSRE. Removing the TDB vector completely and re-selecting track doesn't seem to work either, or not in 1 go at least. I am tempted to run the route thru MSTS TDB rebuild, but it doesn't load anymore there due to another issue introduced by terrain painting in TSRE, as well as my reluctance due to all the linked signals contained. I would hate to have this bug "fixed" by additional work in ORTS, but if it improves on the issue I would trupy welcome it


No, de-selecting, Z then save does not always work. In those cases you must re-snap the offending track section. All the interactive object precautions apply.

In my own route which had a clean Track Database when I started using TSRE I have had absolutely none of the described line up problems discussed in this thread. BUT, during the rebuild of the PRR-East at Patch level 6 it was a horror show.
I had hundreds of problems like those described.

Being that my route starting with no database problems at all and the PRR starting with several thousand it looks to me like TSRE doesn't like problems such as these.
Goku has been working cleaning several database issues, the EmptyItem comes to mind.

That one screwed over a couple of hundred speedposts by mysteriously linking them as Signals. The temp fix at the time was to add a UseEmptyItem false config in the settings. Goku has since fixed the bug and now using Empty TDB datablocks operates as it should.

Getting the PRR to an beginning editable condition took over a year using chifans magic programs to flush out and eliminate the database errors.
Some errors, like the empty item bug didn't show until the interactives were being added. Maucerio help fix the crash when display platforms/sidings switch was enabled.

So the bottom line is if you begin editing a route that has existing database errors, editing results are unpredictable.

Here's the thread where Goku and I hashed out the described problems. A the the time I had no idea that the corrupted PRR database files were the root cause!
edit:add >>>===> http://www.elvastowe...nment-problems/
regards,
vince

#5 User is offline   eolesen 

  • Superintendant
  • Group: Private - Open Rails Developer
  • Posts: 1,546
  • Joined: 30-October 10
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 16 March 2021 - 06:34 PM

Let's try to keep this on-topic, which is whether or not the game engine can absorb/ignore/smooth out abrupt vector changes.

Feel free to use the TDB Corruption thread I created a few months back if you want to talk about the TDB itself --> http://www.elvastowe...uption-on-route

#6 User is offline   vince 

  • Superintendant
  • Group: Status: Elite Member
  • Posts: 1,312
  • Joined: 18-June 14
  • Gender:Male
  • Location:West of the Contental Divide
  • Simulator:ORTS_Running MSTS_Editing
  • Country:

Posted 16 March 2021 - 07:27 PM

Do you mean the abrupt vector changes brought on by TDB corruption? http://www.elvastower.com/forums/public/style_emoticons/default/bigboss.gif

vincehttp://www.elvastower.com/forums/public/style_emoticons/default/curiousPC.gif

#7 User is offline   eolesen 

  • Superintendant
  • Group: Private - Open Rails Developer
  • Posts: 1,546
  • Joined: 30-October 10
  • Gender:Male
  • Simulator:ORTS
  • Country:

Posted 17 March 2021 - 10:49 AM

Yep. We know they happen. How or why, not this thread's objective.

Can the game engine smooth out the raw input?

#8 User is offline   joe_star 

  • Fireman
  • Group: Status: Active Member
  • Posts: 191
  • Joined: 16-January 13
  • Gender:Male
  • Simulator:MSTS
  • Country:

Posted 26 March 2021 - 10:14 AM

View Postvince, on 16 March 2021 - 01:34 PM, said:

No, de-selecting, Z then save does not always work. In those cases you must re-snap the offending track section. All the interactive object precautions apply.

In my own route which had a clean Track Database when I started using TSRE I have had absolutely none of the described line up problems discussed in this thread. BUT, during the rebuild of the PRR-East at Patch level 6 it was a horror show.
I had hundreds of problems like those described.

Being that my route starting with no database problems at all and the PRR starting with several thousand it looks to me like TSRE doesn't like problems such as these.
Goku has been working cleaning several database issues, the EmptyItem comes to mind.

That one screwed over a couple of hundred speedposts by mysteriously linking them as Signals. The temp fix at the time was to add a UseEmptyItem false config in the settings. Goku has since fixed the bug and now using Empty TDB datablocks operates as it should.

Getting the PRR to an beginning editable condition took over a year using chifans magic programs to flush out and eliminate the database errors.
Some errors, like the empty item bug didn't show until the interactives were being added. Maucerio help fix the crash when display platforms/sidings switch was enabled.

So the bottom line is if you begin editing a route that has existing database errors, editing results are unpredictable.

Here's the thread where Goku and I hashed out the described problems. A the the time I had no idea that the corrupted PRR database files were the root cause!
edit:add >>>===> http://www.elvastowe...nment-problems/
regards,
vince


I read through your thread, if only I knew 1/10 of the stuff catalogued there earlier :furiousPC:

I realized now all my issues are at spots where I had reworked the route & applied manual rotation to align 2 ends of the track, rather than having snapped the pieces together. Although the visual pieces are now aligned perfectly, the yellow TDB lines say otherwise.

While waiting for any ORTS solution, I am going to try to smoothen these out via TSRE or manual TDB manipulation, with lots of backups in hand.

Not going to clutter this thread further, keeping to the topic :)

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users