SG files and new editing tools
(08-26-2021, 01:26 PM)Pavel 69 Wrote: If DYNO strings are for cars then I agree with that. It looks like DYNO string set the area (longitude an an latitude range) where car can appear on current section. And there is no sence to change values in DYNO strings.

More intresting is why trk23d 3d file has FACE function in the beginning of every section? But converted file doesn't have it. Because this FACE function should turn off whole section and cars on it when viewer is below the plain of FACE function.
I have a question out to Noorok to confirm if the tool really ignores FACE.

I had another thought this morning, from thinking about the point where the pit lane and cars pop into view at Nashville. I was thinking that might have to do with which HI sections are being switched on. But relevant to that, which I hadn't considered before, is that we have to think about both the sections ahead and the sections behind.

I think the N3 subsections may be too short compared to those for ICR2 and therefore the unusual things happen when track sections ahead and behind are "loaded" by the game. If indeed this is the case, there is no easy solution since we do not have the Trk23d tool developed for ICR2. However, it is still theoretically possible to edit the .3D text file to combine and lengthen the sections.
Reply
Noorok converter ignores LINE, but not FACE.
Reply
Here's some more information about subsection lengths.

From ICR2 default Laguna, the subsections on curves have DLONG lengths of anywhere from about 100000 to 440000. Straight sections have a maximum subsection length of 1600000 (I guess if the section is over 2000000 long, then it will divide it into 2 subsections).

On an N3 track, the curve subsections that I observed are between 49000 and 500000 (but mostly in the 100000-200000 range), so not significantly different from ICR2. Straight sections, however, are consistently between 160000 and 530000 (my guess is anything longer than ~600000 gets divided).

There will be a bit of work involved but I would like to try to redo the straight sections and I think the way to do it might be to texturize the LO section (I think there is an option to do this in TRK23D) and make that the HI/MED/LO sections. Fences might need to be dealt with separately, but not as important to this question.
Reply
What for? More subsections - more closer track 3do follows trk. Also all default tracks in icr2 came almost unchanged from icr1. And in those times less polys= better fps
Reply
(08-26-2021, 06:28 PM)Pavel 69 Wrote: What for? More subsections - more closer track 3do follows trk. Also all default tracks in icr2 came almost unchanged from icr1. And in those times less polys= better fps

I have a suspicion that the many subsections causes the cars showing through elevation glitch. I will see if there is a quick way to prove or disprove this. In any case, I think I would just combine the straight sections so that actually the track detail should be unaffected.
Reply
Today I had the idea to reduce the overlap in the DATA section, but in .3d file all DATA lines are only once.
In the original .3do files the overlap is incomplete, so some DATA lines are twice in the file and some are only once. In converted .3do all DATA lines are at least twice in the file, but they can be there up to three times.
What is the reason for this? And can we adjust it?
Reply
(08-26-2021, 05:18 PM)Pavel 69 Wrote: Noorok converter ignores LINE, but not FACE.

Noorok confirmed to me that the converter specifically replaces FACE on section roots with the LIST that is the first child of FACE. Keeping FACE in there makes the game crash. So this is consistent with what Dennis and I have observed.
Reply
(08-26-2021, 07:54 PM)Dennis Wrote: Today I had the idea to reduce the overlap in the DATA section, but in .3d file all DATA lines are only once.
In the original .3do files the overlap is incomplete, so some DATA lines are twice in the file and some are only once. In converted .3do all DATA lines are at least twice in the file, but they can be there up to three times.
What is the reason for this? And can we adjust it?

Good question. I wonder if it affects anything though, if the DATA lines are simply pointing to the same things. On the default tracks, there isn't even a full set of the second lap - I checked Laguna and Elkhart where the DATA seems to include the last 15 sections or so, then a full lap, then the first 15 sections.
Reply
(08-26-2021, 06:46 PM)checkpoint10 Wrote: I have a suspicion that the many subsections causes the cars showing through elevation glitch. I will see if there is a quick way to prove or disprove this. In any case, I think I would just combine the straight sections so that actually the track detail should be unaffected.

Unfortunately, reducing subsections seemed to make zero impact on the hill visibility glitch. It's too bad, but at least we can keep the smaller subsections thus improving the detail of the track, and avoid the labor-intensive method of swapping out a HI with a LO section
Reply
Meantime I was thinking why with my center wall fix cars draw properly. And I think the reason is I moved all DYNO's to the LIST at the end. So I want to try this with simple section too. I will relocate single DYNO from BSPA to the LIST. I'm going to test this during weekend.
   
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)