MTM2 Beta Patch 0.47 - Improving from .46
Re: MTM2 Beta Patch 0.47 - Improving from .46
Sorry for the delay, ran out to the market. Here's quid on 6650 xt hdd in both cases: https://drive.google.com/file/d/1BwVJFn ... drive_link, https://drive.google.com/file/d/1dRCOhZ ... drive_link
Still some hitches with both vsync and 3x buff off, but very smooth otherwise. Again, unplayable with both on.
Still some hitches with both vsync and 3x buff off, but very smooth otherwise. Again, unplayable with both on.
Re: MTM2 Beta Patch 0.47 - Improving from .46
For the toggle question, can you make the vsync button toggle triplebuffer on and off so that both are either on or off simultaneously without adding a separate toggle?
Re: MTM2 Beta Patch 0.47 - Improving from .46
Just for fun, here's 6650 quid on hdd with vsync off and 3x buff on: https://drive.google.com/file/d/1vArqUG ... drive_link
Much better performance than with both vsync and 3x buff on. Maybe similar to both off.
Much better performance than with both vsync and 3x buff on. Maybe similar to both off.
Re: MTM2 Beta Patch 0.47 - Improving from .46
http://www.mtm2.com/~mtmg/misc/Beta0.47_Rudee_test8.zip
Shorter test this time and yeah, only the .dll file is in there.
I coupled vsync and tripple buffer together. But they can be independently turned off in monster.ini so just toggle vsync on and off a few times and it should get them to couple together. (actually, knowing how crappy the reading method in mtm2 is, toggle the button a few times, leave it on, exit the game and start it again. All this cumbersome stuff will get fixed the day i move to fully hardware rendered menus)
Nothing really changed in this one except the vsync on - 3x on. So run vsync and do either bobsled or quid, no need both and see how it feels. It should have a slow down for about half a sec and then go back to a normal behaviour and with very likely hitches. Also curious if it drops to unplayable for like a second or two every now and then.
To sum it up, the code introduced a weird pattern where the vsync + triple buffer would lock into a bad rythm once the game stutters. Which is a guarantee because the game ALWAYS stutters at load time, so the vsync would pick that up and just lock itself. This should be fixed now.
And reading your logs for vsync off on the HDD, the leftover hitches all follow a repeated pattern instead of something you are actually doing. Or at least this is how it reads looking at the log files which makes it look like it has something to do with the HDD itself, or rather, background operations activating the background and affecting MTM2 which is why you don't see any of it on the SSD.
Game working with multithread cores could potentially solve this problem. The game is not meant to use multithreaded cores and doing that would be a significant reimplementation, which is not doable with Dx9 in the first place because Dx9 doesn't know multithread. Something to keep note when I bring in Dx11 and OpenGL. But yeah it does look like I don't think I can do more for this. There are some real bonuses that we got form this goose chase though. There were legitimate issues with the sound reading that were fixed (these always existed but because of your HDD issues, they were made more obvious), so that got fixed. I did some behind the scenes memory management which resulted in a 20 to 40fps increase for everyone (i think, looking at your own logs, you used to run 60fps ish for the lower avg and now it looks like it's around 90-100). So there's real positive stuff here, just that doing away with the HDD hitches might not be possible with my knowledge or lack of it
But hey, it works great on SSD :p
Shorter test this time and yeah, only the .dll file is in there.
I coupled vsync and tripple buffer together. But they can be independently turned off in monster.ini so just toggle vsync on and off a few times and it should get them to couple together. (actually, knowing how crappy the reading method in mtm2 is, toggle the button a few times, leave it on, exit the game and start it again. All this cumbersome stuff will get fixed the day i move to fully hardware rendered menus)
Nothing really changed in this one except the vsync on - 3x on. So run vsync and do either bobsled or quid, no need both and see how it feels. It should have a slow down for about half a sec and then go back to a normal behaviour and with very likely hitches. Also curious if it drops to unplayable for like a second or two every now and then.
To sum it up, the code introduced a weird pattern where the vsync + triple buffer would lock into a bad rythm once the game stutters. Which is a guarantee because the game ALWAYS stutters at load time, so the vsync would pick that up and just lock itself. This should be fixed now.
And reading your logs for vsync off on the HDD, the leftover hitches all follow a repeated pattern instead of something you are actually doing. Or at least this is how it reads looking at the log files which makes it look like it has something to do with the HDD itself, or rather, background operations activating the background and affecting MTM2 which is why you don't see any of it on the SSD.
Game working with multithread cores could potentially solve this problem. The game is not meant to use multithreaded cores and doing that would be a significant reimplementation, which is not doable with Dx9 in the first place because Dx9 doesn't know multithread. Something to keep note when I bring in Dx11 and OpenGL. But yeah it does look like I don't think I can do more for this. There are some real bonuses that we got form this goose chase though. There were legitimate issues with the sound reading that were fixed (these always existed but because of your HDD issues, they were made more obvious), so that got fixed. I did some behind the scenes memory management which resulted in a 20 to 40fps increase for everyone (i think, looking at your own logs, you used to run 60fps ish for the lower avg and now it looks like it's around 90-100). So there's real positive stuff here, just that doing away with the HDD hitches might not be possible with my knowledge or lack of it
But hey, it works great on SSD :p
Re: MTM2 Beta Patch 0.47 - Improving from .46
Okay cool. Excited to try it out. We have guests over now so I will give it a go in the morning.
Re: MTM2 Beta Patch 0.47 - Improving from .46
My god. So Dx9 right now is, visually at least, at a level that's pretty close to what I want so I started working on porting it to Dx11. The thing is porting it but is throwing away all the fine tuning that I did... so I have to redo every tweak and every bug fix all over again O_O
Re: MTM2 Beta Patch 0.47 - Improving from .46
Out of pure giggles, I decided to throw in this big ass boi (Dennis Taft's King Of The Road) onto the Drag tracks just to see what happens, from my tests, the bigger the truck, the more prone it is to red lighting, even when you slam on the gas at the green light mark, as opposed to my technique, slam on the gas on the first yellow light from a shallow stage.Fila wrote: Sat Jul 04, 2026 5:35 pm I think sometimes the game saves the previous state of the truck's movement and when a new race loads it still acts on that movement. Not an issue for regular race tracks, bigger issue for drags. Will look into it.
Will look into the staging issue too. Zoon's lunar lander from zoon moon doesn't wanna stage either![]()
![]()
![]()
https://mtm2.com/~trucks/dl.php?dl=6291
Re: MTM2 Beta Patch 0.47 - Improving from .46
The drag set is sound... but not really. That probably also needs to have a further look into accounting for various size trucks.
In theory it is supposed to do that right now, but it is not doing it very well.
In theory it is supposed to do that right now, but it is not doing it very well.
Re: MTM2 Beta Patch 0.47 - Improving from .46
So, the toggle for vsync never coupled with 3x buff for me. I tried all sorts of combinations. Also, the game would change to triplebuffer=0 on startup no matter whether vsync was toggled on or off. So, I would set triplebuffer=1 in monster.ini, start the game, and next time I checked it would be triplebuffer=0.
Anyway, if editing the monster.ini while in game works, this is with both vsync and triplebuffer on: https://drive.google.com/file/d/1bn0xvd ... drive_link
Still unplayable on the hdd, and no periods of smoothness as you were expecting.
Anyway, if editing the monster.ini while in game works, this is with both vsync and triplebuffer on: https://drive.google.com/file/d/1bn0xvd ... drive_link
Still unplayable on the hdd, and no periods of smoothness as you were expecting.
Re: MTM2 Beta Patch 0.47 - Improving from .46
Alright, so I finally got to play around with this, and I didn't see any major issues in the few races I tried. The "Sonic Trucks" certainly are something, the extra speed is cool and makes them more competitive, but there's obvious times when they are too aggro and end up wrecking themselves. In a brief lap on The Heights I saw the trucks blow the last left-hander before the first river crossing, and on a few different tracks I saw them nose pirouette after slamming the brakes to make a turn. I may keep that turned off even though it's definitely fun to race with.
As to the trucks: well, this is going to be a lot of work and I'm not sure I want to do it, haha. At a bare minimum, I think I'd go through my HD trucks and change the face types of the bed covers, KC lights, and undercarriages (and the tires but not the wheels) to normal, and that would definitely make the trucks look more realistic. Beyond that - has the 1000 vert limit thing been crushed? Because then there's an argument to go full fiberglass and do a full interior, you could presumably do it without major texture work because using shiny faces in targeted areas could let the reflections take care of a bunch of that. I guess the question is how would a truck like Jqoc's Bigfoot 14, a notoriously high-vert truck, work in the updated game. If it runs fine with it (or 8 of them), then the shackles are off in a way. But if there's still issues then half-fiberglass is probably still the bet.
Again, things have been busy and remain busy for the foreseeable future, so my capacity to make the changes I want is limited at the moment. But we'll see where things go.
As to the trucks: well, this is going to be a lot of work and I'm not sure I want to do it, haha. At a bare minimum, I think I'd go through my HD trucks and change the face types of the bed covers, KC lights, and undercarriages (and the tires but not the wheels) to normal, and that would definitely make the trucks look more realistic. Beyond that - has the 1000 vert limit thing been crushed? Because then there's an argument to go full fiberglass and do a full interior, you could presumably do it without major texture work because using shiny faces in targeted areas could let the reflections take care of a bunch of that. I guess the question is how would a truck like Jqoc's Bigfoot 14, a notoriously high-vert truck, work in the updated game. If it runs fine with it (or 8 of them), then the shackles are off in a way. But if there's still issues then half-fiberglass is probably still the bet.
Again, things have been busy and remain busy for the foreseeable future, so my capacity to make the changes I want is limited at the moment. But we'll see where things go.
Re: MTM2 Beta Patch 0.47 - Improving from .46
Pretty sure from what Kmaster told me already, that the vertices limit has already been increased, be it last patch or the Comm Patch #2, and that was when I asked him in regards to truck details equivalent to those found on both Beam and Rigs (Rigs runs on DX9, whereas Beam runs on DX11/12), so start firing on all cylinders when you can, no rush!Cale wrote: Sun Jul 05, 2026 11:42 am As to the trucks: well, this is going to be a lot of work and I'm not sure I want to do it, haha. At a bare minimum, I think I'd go through my HD trucks and change the face types of the bed covers, KC lights, and undercarriages (and the tires but not the wheels) to normal, and that would definitely make the trucks look more realistic. Beyond that - has the 1000 vert limit thing been crushed? Because then there's an argument to go full fiberglass and do a full interior, you could presumably do it without major texture work because using shiny faces in targeted areas could let the reflections take care of a bunch of that. I guess the question is how would a truck like Jqoc's Bigfoot 14, a notoriously high-vert truck, work in the updated game. If it runs fine with it (or 8 of them), then the shackles are off in a way. But if there's still issues then half-fiberglass is probably still the bet.
If I had a new project title for you, it would be something called "Cale's EVO Trucks", that being high detail, and somewhat a similar art style to those Monster Trucks found on 4x4 Evo 2, PM Customs/Nat Gage's Chevy Thunder and Thunderfoot and Outrage 2020 would be prime examples of that. Basically an imaginary if MTM3 did exist as opposed to the 4x4 Evo games instead.
Side note: Remember this thread? That'll be some good inspiration for it.
EDIT: Bloody hell! Similar thread name, but not the one I was after, epic fail!
Last edited by Ford-chan on Sun Jul 05, 2026 1:49 pm, edited 1 time in total.
Re: MTM2 Beta Patch 0.47 - Improving from .46
Been doing some more testing this morning with your most recent hotfix, and it plays very smoothly on my 5600 xt machine on SSD. With vSync on, the game runs buttery smooth.
So, I am now trying the SSD on the 6650 XT, and... vSync is very stuttery now. However, SAM has been turned back on on this machine for some time now. So, that must be the issue killing the vSync. The game runs fine with vSync off and 3x buffer left on.
I am going to take the HDD over to the 5600 xt and test vSync there. Will report back.
Update1: So, yeah game runs perfectly fine off the HDD on my 5600 xt machine with no SAM over all possibilities vSync/3xbuff on/off.
I'll go turn SAM off on my other machine now and test there.
So, I am now trying the SSD on the 6650 XT, and... vSync is very stuttery now. However, SAM has been turned back on on this machine for some time now. So, that must be the issue killing the vSync. The game runs fine with vSync off and 3x buffer left on.
I am going to take the HDD over to the 5600 xt and test vSync there. Will report back.
Update1: So, yeah game runs perfectly fine off the HDD on my 5600 xt machine with no SAM over all possibilities vSync/3xbuff on/off.
I'll go turn SAM off on my other machine now and test there.
Re: MTM2 Beta Patch 0.47 - Improving from .46
If two of those trucks + 6 of your HD trucks don't crash Fila's Adventure Redux with complex scenery at start, nothing willCale wrote: Sun Jul 05, 2026 11:42 am Alright, so I finally got to play around with this, and I didn't see any major issues in the few races I tried. The "Sonic Trucks" certainly are something, the extra speed is cool and makes them more competitive, but there's obvious times when they are too aggro and end up wrecking themselves. In a brief lap on The Heights I saw the trucks blow the last left-hander before the first river crossing, and on a few different tracks I saw them nose pirouette after slamming the brakes to make a turn. I may keep that turned off even though it's definitely fun to race with.
As to the trucks: well, this is going to be a lot of work and I'm not sure I want to do it, haha. At a bare minimum, I think I'd go through my HD trucks and change the face types of the bed covers, KC lights, and undercarriages (and the tires but not the wheels) to normal, and that would definitely make the trucks look more realistic. Beyond that - has the 1000 vert limit thing been crushed? Because then there's an argument to go full fiberglass and do a full interior, you could presumably do it without major texture work because using shiny faces in targeted areas could let the reflections take care of a bunch of that. I guess the question is how would a truck like Jqoc's Bigfoot 14, a notoriously high-vert truck, work in the updated game. If it runs fine with it (or 8 of them), then the shackles are off in a way. But if there's still issues then half-fiberglass is probably still the bet.
Again, things have been busy and remain busy for the foreseeable future, so my capacity to make the changes I want is limited at the moment. But we'll see where things go.
It takes my fps to 100 (from 140 that was the last baseline) with everything turned on. Now I don't know how this will react for others (since my system is beefy, I generally get more fps).
Re: MTM2 Beta Patch 0.47 - Improving from .46
Update 2: Turned the SAM off and still getting choppyness on the 6650 XT with vsync on even running on the SSD. So, who knows!? It runs perfectly fine / smoothly with vsync off on this machine (3x buffer on or off).
Game runs better with vsync on on my 5600 xt on both ssd and hdd.
Separate issue: I noticed when pausing the game during a race, sometimes unpausing results in all of the track models disappearing. Pausing and unpausing again makes them reappear.
Game runs better with vsync on on my 5600 xt on both ssd and hdd.
Separate issue: I noticed when pausing the game during a race, sometimes unpausing results in all of the track models disappearing. Pausing and unpausing again makes them reappear.
Re: MTM2 Beta Patch 0.47 - Improving from .46
Does that pause bug happen on both machines or just 6650? I know why it's happening, and it shouldn't but I'm wondering if 6650 makes it skip an instruction or whether it's just random luck of the draw (i've been doing lots of pausing and not having the issue since I patched it).
Re: MTM2 Beta Patch 0.47 - Improving from .46
Both machines. I just got it to happen of 5600 by starting a race with vsync on, then going into the menu and turning it off during the race. When resuming the race, the track objects were gone. Pausing and unpausing brought them back.
Re: MTM2 Beta Patch 0.47 - Improving from .46
http://www.mtm2.com/~mtmg/misc/Beta0.47_Rudee_test9.zip
Turns out the code did not work as intended, there was a miscalculation in there. Try this, quid pro only with vsync on.
3x buffer is now, by default, tied to vsync without changing the ini value. So if you turn vsync on it automatically turns the 3x buffer too. I'm probably going to make these separate, but for testing reasons now I'll keep it as is since the only thing we are worrying about now is getting the vsync to work as much as possible, I think the vsync off is as far as we can go (in regards to 6650 SAM version)
There's also a rephase.txt log now too, send that one too please.
Turns out the code did not work as intended, there was a miscalculation in there. Try this, quid pro only with vsync on.
3x buffer is now, by default, tied to vsync without changing the ini value. So if you turn vsync on it automatically turns the 3x buffer too. I'm probably going to make these separate, but for testing reasons now I'll keep it as is since the only thing we are worrying about now is getting the vsync to work as much as possible, I think the vsync off is as far as we can go (in regards to 6650 SAM version)
There's also a rephase.txt log now too, send that one too please.
Re: MTM2 Beta Patch 0.47 - Improving from .46
Just tested. Still chopping on 6650 on ssd with vsync and 3x on: https://drive.google.com/file/d/1je5gPT ... drive_link
I checked monster.ini after turning on vsync, and the triplebuffer was still 0. Did you mean the monster.ini value is irrelevant now?
I checked monster.ini after turning on vsync, and the triplebuffer was still 0. Did you mean the monster.ini value is irrelevant now?
Re: MTM2 Beta Patch 0.47 - Improving from .46
http://www.mtm2.com/~mtmg/misc/Beta0.47 ... test10.zip
Same thing again please
There shouldn't be any improvement yet, just one of the logs should be more detailed so hopefully catch something else.
Yeah the value in monster.ini is irrelevant for now.
Same thing again please
Yeah the value in monster.ini is irrelevant for now.
Re: MTM2 Beta Patch 0.47 - Improving from .46
No Improvement. Here it is: https://drive.google.com/file/d/1UnUXwM ... drive_link