Posts: 50
Threads: 7
Joined: Jul 2020
Reputation:
1
05-18-2024, 10:35 AM
(This post was last modified: 05-18-2024, 10:49 AM by Tony_D 26nz.)
Hi Samsepi
I know it's a bit late, but I downloaded and tried your 1st version of LB 96 today.
In both windows version and dos version I got CTD's before end of 1st lap. With windows version it was at last corner. With dos version it was after the fast right hander leading into the left-right complex at the end. Then I realised I was in "single race" mode instead of preseason mode. Changed to preseason and all was good. There's some learning for me.
I've just had a quick spin around both versions. Looking really great. The one thing I noticed was a quick flicker in Windows version at T2 driving in F10-1 mode.
p.s. I also downloaded 2May version of LB91 today and will have a go in next couple of days. Looking forward to it. I've really enjoyed the earlier versions of LB91 so far
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
(05-18-2024, 10:35 AM)Tony_D 26nz Wrote: Hi Samsepi
I know it's a bit late, but I downloaded and tried your 1st version of LB 96 today.
In both windows version and dos version I got CTD's before end of 1st lap. With windows version it was at last corner. With dos version it was after the fast right hander leading into the left-right complex at the end. Then I realised I was in "single race" mode instead of preseason mode. Changed to preseason and all was good. There's some learning for me.
I've just had a quick spin around both versions. Looking really great. The one thing I noticed was a quick flicker in Windows version at T2 driving in F10-1 mode.
p.s. I also downloaded 2May version of LB91 today and will have a go in next couple of days. Looking forward to it. I've really enjoyed the earlier versions of LB91 so far
The 96 version has a text file with values from 91 so they aren't right. The track was shorter in 96, so the DLONG values will cause a crash. As I type this I am almost done with LPs and TXT values that will allow me to test it this weekend and hopefully get something you can use in "single race" mode.
LB91 should be raceable and fully driveable. If you experience a crash, it is something I need to fix.
LB96 will only be pre-season until I release a better TXT and LP files.
Thank you for your feedback!
Posts: 50
Threads: 7
Joined: Jul 2020
Reputation:
1
Feedback from my runs/tests at the new LB 96 & LB91 this weekend
Long Beach 96 (1st draft) - Best lap 51.6
- The original LB is one of my favourites. This is a step up from it. I love the flow of T1/T2, aiming down the chute between T3 & T4, and the changes to the left hander before the hairpin at the end.
- A couple of times when I was tight on the exit at the hairpin & going fairly fast I got crashed out even though I didn't hit the wall (but was close).
Long Beach 91 (nearly finished) - Best ~ 1:04.0. At 99% opps this has me a fair way off the pace, so there's lots of time to be gained through better understanding of the track and improved cornering/exit speeds.
- Racing the AI is great. Couldn't fault them. As good as I've seen in icr2.
- Tracks is lots of fun to drive and to race the AI on. The extra section compared to LB96 (after T4) is a real challenge to do well, and the fact it tightens right up at the end (narrower track) adds to it.
- With standing start, at start of race all cars started in single file from left of track. I assume this is something still to be corrected.
- Watching race replay there's a few parts where my car disappears a bit.
- I still have a lot of trouble with the brake markers here at end of the 2 straights. This really does feel like an issue to me (not just my old eyes). Why have them if they can't be seen? I did find other markers to use as references for braking. But it would be good if the brake markers were bigger.
For some reason I'm quicker with the windows version than the dos version even though the settings look similar.
Thanks & great work
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
(05-19-2024, 09:13 PM)Tony_D 26nz Wrote: Long Beach 96 (1st draft)- Best lap 51.6 (I was at 51.6, and I tried running lower wings 14 degrees front / rear and it made me a little faster i was able to do 51.4s - also I like to get the car a little softer, roughly 40% suspension I think.)
- The original LB is one of my favourites. This is a step up from it. I love the flow of T1/T2 (the t1/t2 section on papy long beach is 2 distinct corners but in real life it's almost a chicane. t1 is unique for me because I am really just braking and hitting the apex while I continue to slow for t2)
- A couple of times when I was tight on the exit at the hairpin & going fairly fast I got crashed out even though I didn't hit the wall (but was close). There is some variation between a trk file, and a 3do file. Remember a 3do file is just an illusion - we think we are driving on a 3do file, when we really aren't. I have done the exact same thing but I don't know how to make it any better without loosening up the corner some (and I'm sure you've seen from overhead replays - the hairpin is far from a smooth shape- it is almost 2 tight bubbles.) Once again we are "bending" the rules here and there are consequences. In reality the wall on the exit of the hairpin should be about 6ft narrower also. To accomplish this I would have had to do a "nashville" where the centerline is outside of the racing surface. Which would have made other problems.
Long Beach 91 (nearly finished)- Best ~ 1:04.0. At 99% opps this has me a fair way off the pace, so there's lots of time to be gained through better understanding of the track and improved cornering/exit speeds. Try adding more wing (I run 16.8 both front and rear) and stiffening up the car. The entire infield section basically ruins the back straight as you can barely get into 5th gear at the end. Drag will only slow you on the front stretch not the back. Also stiffen up the suspension as the entire track from t1 to the exit of the infield is one corner on top of the next. When you pass under the hyatt overpass, concentrate on getting your line right for getting positioned to be fast exiting to the back straight. If you have to adjust the difficulty - change the RELS in the track.txt, this way it will be dialed in for you better. It's not about being the fastest - it's about having fun and enjoying yourself!
- Racing the AI is great. Couldn't fault them. As good as I've seen in icr2. I have a new formula that I use to make pass1 and pass2 LPs which I think is helping. One thing I love is that it does not seem like the CCs get jammed up in large groups. I have been watching gplaps 89 season and this is something that happens a lot. The thing I do to avoid this, is give a 12 foot difference in the pass LP files from the race LP file. Testing and making LPs for SK's Detroit taught me this. What bugs me (and I'm sure Pavel has noticed this) is the little wiggle I get exiting t4 on both tracks. If you look at the LP here with EDIT45 I had to almost make the shape of the position trace a spike rather than a "bell curve". With positions from TRAFO20 it holds the inside too long and then dives out (this is a problem that manually needs to be corrected and it's a royal p.i.t.a.)
- With standing start, at start of race all cars started in single file from left of track. I assume this is something still to be corrected. Well.... I don't know. Normally I set up the track txt and the PACE.lp to use single file starts- since it's cleaner (less pile ups and accidents at the starts). To fix this I would need to make an alternate PACE.LP (which isn't really hard- as long as 2 rows can make it around the hairpin and I haven't tested that) but personally if it were solely up to me for my personal use, I would keep it single file. For me it was a tough decision to do this but all of my tracks (regardless if they are vanilla papyrus tracks, pavels tracks or even SK's) I use single file starts with basically the position data from my race.lp. It bugged me at first but now that I tried it I like it more. I will make a 2 file track.txt and pace.lp if people want it. (I still won't use it myself though) It's a 1 line edit in the track.txt (I think on the PACEA line). I have to give proper DLAT values for the rows (I think most are -45000 and 45000) I could just the position from a replay driving down the center. It's not that much work. My problem with 2 row starts (as gplaps pointed out in one of his 89 season videos I watched) is that the cars follow too closely. I was just telling SK about this the other day- I wish we could reverse engineer the game to understand what values we can change in the EXE to manipulate the following distance under yellow. There must be some definition- look at how far they stay behind the pace car. I wish I took more computer science in college and maybe I could dissasemble the EXE and figure this out.
- Watching race replay there's a few parts where my car disappears a bit. I'm not sure what to even say about this - my 3d file was done way before the cam and scr but SK knows a TON more about camera angles and replays than I could ever know so maybe he can provide some feedback on what I could do if anything?
- I still have a lot of trouble with the brake markers here at end of the 2 straights. This really does feel like an issue to me (not just my old eyes). Why have them if they can't be seen? I did find other markers to use as references for braking. But it would be good if the brake markers were bigger. I do agree with your here. It is on my "to do" list. Both tracks use the same brake-marker 3do's and they are coded to be 9000 x 9000 (1.5 foot square) For me- the problem I have it on LB96 on the back straight since the posters behind the brake markers sorta "camouflage" them. Right now the brake markers are drawn as TSOs which are supposed to be outside of the racing surface - as a result when the track fence and the brake markers try and occupy the same "viewing space" the FENCE takes priority when it should be the markers. This is partly why I positioned them like I did. The idea was to put the brake marker in place to minimize the chance of this being obvious. I can make them bigger (which is really easy since they don't need to even be re-positioned at LB since the 0,0,0 origin is on the left side) for you. Now- since Pavel schooled me on how to put objects in the TSD section - is this a better way to do this now? Can we change the drawing order of the brake marker vs fence by cheating like this?
For some reason I'm quicker with the windows version than the dos version even though the settings look similar. I have actually never played windy! LOL
Posts: 501
Threads: 4
Joined: Dec 2013
Reputation:
6
05-20-2024, 05:19 AM
(This post was last modified: 05-20-2024, 05:20 AM by Dennis.)
(05-20-2024, 04:49 AM)samsepi0l Wrote: A couple of times when I was tight on the exit at the hairpin & going fairly fast I got crashed out even though I didn't hit the wall (but was close). There is some variation between a trk file, and a 3do file. Remember a 3do file is just an illusion - we think we are driving on a 3do file, when we really aren't. I have done the exact same thing but I don't know how to make it any better without loosening up the corner some (and I'm sure you've seen from overhead replays - the hairpin is far from a smooth shape- it is almost 2 tight bubbles.) Once again we are "bending" the rules here and there are consequences. In reality the wall on the exit of the hairpin should be about 6ft narrower also. To accomplish this I would have had to do a "nashville" where the centerline is outside of the racing surface. Which would have made other problems.
Without looking at it, the difference between the track edge of the TRK and the wall of the 3DO can be theoretically reduced, with a smaller subsector size. This would add more edges to the wall, which would get it closer to the shape, which it should have. To achieve this, an adjustment in the trk23do script would be necessary. The downside of this is, without further tricks, it would cause a shorter drawing distance, because of the higher number of subsectors.
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
05-20-2024, 05:17 PM
(This post was last modified: 06-10-2024, 02:38 PM by samsepi0l.)
updated LBEACH96, with fully raceable AI! They aren't perfect but I just did a 100% race and it was a ton of fun.
I skipped qualifying and started dead last- I finished 4th. I caused 2 yellows because I was a little over-aggressive (my fault) but besides that I did a lot of passing and wheel to wheel battling and I think the AI cars are very reasonable to race with. They give you room when you are on the right line. I think my LPs are getting better since I struggled so much with LB91 (especially in the hairpin).
Also - say "hi" to checkpoint10's whales when you drive by the arena!
ATTACHMENT REMOVED BECAUSE A NEWER VERSION IS AVAILABLE IN A NEWER POST
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
05-29-2024, 04:12 PM
(This post was last modified: 05-29-2024, 04:17 PM by samsepi0l.)
workin' on my mrk'n...!
Posts: 118
Threads: 26
Joined: Nov 2006
Reputation:
3
Thank you for this project, these versions of Long Beach are spectacular and add a lot of feel compared to the default!
I did a race today on the 91 version and had a great time. The AI race really well once you learn where they give space or not!
I had a couple of smaller issues which i wanted to note in case you didn't know of them:
- Cars disappear in the left side mirror after the start line going through Shoreline Drive
- There is a framerate stutter every lap just at about the turn in to turn 6 (this could be my computer/dosbox config)
Excited to try the 96 layout at some point!
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
06-02-2024, 07:05 AM
(This post was last modified: 06-02-2024, 07:09 AM by samsepi0l.)
(06-02-2024, 01:48 AM)jxeboy Wrote: - Cars disappear in the left side mirror after the start line going through Shoreline Drive
I did see this and of course I could not experience it myself until there were other cars on the track. I thought that it may have been due to a PIT line TXT value, but now I'm not so sure. It only does it for a second or so for me, and not every time. I wonder if the mirror view is smart enough to know that were turning right here along the kink, and for some reason it thinks the car should go out of view in your left mirror? I'm not sure- but for me it was not always repeatable. If anyone has any better ideas to explain this I would be glad to listen.
(06-02-2024, 01:48 AM)jxeboy Wrote: - There is a framerate stutter every lap just at about the turn in to turn 6 (this could be my computer/dosbox config)
I get this from time to time on my machine too. I think it is because the majority of the polygons are loading as you enter the infield so everything can be drawn on the way out. I get this sometimes on HEAVY tracks where a lot is loading (for example turning at the end of congress in Detroit downtown 91, and also when the huge grandstand in Chicago loads) The funny thing is I get none of these in Denver, which is really heavy in my opinion.
Super glad that you enjoyed the race and I would love to see Ryan Axelson race! I just got done watching the 1990 vanilla long beach Axleson race and wow it was weird seeing vanilla long beach again. (I am a little behind as I just finished 1989). The videos are a ton of fun to watch and I love how we share the same enthusiasm for the game. Seeing people enjoy the tracks helps my motivation - can't believe it has taken me a year to get where I am on the long beach project.
Posts: 1,374
Threads: 67
Joined: Aug 2017
Reputation:
9
06-10-2024, 02:40 PM
(This post was last modified: 07-31-2024, 03:20 PM by samsepi0l.)
Here is the newest version of Long Beach 96!
This has MRK (wall ads) TSD (surface lines) and replay cams. It also has bigger brake markers (they are 2.5ft square instead of the previous 1.5ft square). I also moved the brake markers down some.
AI is better, still not perfect. I just raced a full 100% length race and it worked good. Tons of fun.
I am pretty proud of my textured TSDs inside the corners of the curb-apex sections. Besides that this still has an issue (like Jake pointed out) with the cars disappearing in the mirrors just barely for a second or so when you cross the S/F line. This only happens when you hug the wall on the right. If you are just a slight bit away from the wall (like the nose of your car right on the TSD dotted line) it will not happen. I know what is happening here. For some reason the game thinks you are inside the pit lane, and hence on the other side of the wall. I know this because once when I drove through here I could see (for a split second) a car in the pits that was exiting the pits. Logic told me that this must be a wrong PIT value in the TXT file - specifically the first one which tells the game where the DLAT for the PIT is.. but this is not the case. I tried a bunch of different values here (even ridiculous ones) and it did not help. If anyone has any ideas please let me know.
(ATTACHMENT REMOVED - NEWER VERSION AVAILABLE)
To do still in LBEACH96:
- It would be nice to fix the "mirrors" issue. CANT FIGURE THIS ONE OUT... ANY IDEAS GUYS?
- Improve AI more so they are a little slower when offline ready to go in next release
- lower brake markers a tiny bit more, maybe move them into the TSD section of the BSP tree so they draw OVER the fence? FIXED WILL UPDATE IN NEXT RELEASE
- maybe manually edit the 3D file to create a second darker road texture in the parking lot section of the track
- maybe improve the SCR and CAM so that they zoom better like the real TV cameras. I'm (of course) using SK's CAM/SCR maker tool, and this would be easy to do but I need to experiment a bit with zoom values DONE- WILL COME OUT IN NEXT RELEASE
Let me know what you guys think, hopefully it does not suck.
|