Main Forum

side view

rear view

Thanks for the comments. Yes, I agree on all counts and noticed the same things when playing with bigger boats, especially the weight difference. The weight of the boat is computed from the surface area of the mesh as it's generated, then the driver, fuel, and engine(s) are added to that. The surface area/weight relationship (density and thickness, basically) was calibrated for pretty thin fiberglass material which must not be the case for the bigger boats.

Do you know much about boat construction? Do you know how thick the fiberglass is on the big ones compared to the little 18-21 footers? Maybe I can figure out a way to progressively increase the density with size somehow so things scale up better. It is pretty strange to be blowing over in a 40+ foot boat, they are really too light here.

I agree the cockpit looks goofy unless you set it to the rear position, but then you've got an enormous nose that can be tough to see over. It was fine for the smaller boats but obviously doesn't work real well once they get long. I'll need to look at some pictures of the bigger outboards and rethink some of this, maybe I can make it so the cockpit is of a fixed length, and as you move it forward the rear of the cockpit follows along with it so you don't have a 20 foot gap between the front seats and the rear bench with all that open space inbetween.

As for performance, I haven't played much with the big ones yet. Once I got everything working I spent maybe an hour making those two big boats for the update just to give everyone something to jump right into that wasn't too terrible. So you'll be faster than me with just a little bit of tweaking.

Somebody on Facebook suggested adding a ballast option so you could weigh the nose down too, something like moving your batteries or other equipment around. Maybe a bit of left/right shift could be done too. I thought of that long before release, but decided against it just so people weren't doing goofy things. But now that the weight adjustment is in there and it seems like a positive thing we're all enjoying, a little center of gravity shifting with a ballast weight might not be too unreasonable.

With the weight going up to more typical levels on the big boats, we're going to need some big inboards with a lot more power. I've got a really nice V8 sound sample sitting here that's just begging for an excuse to go in the simulator. 😀

Nice pics! I see that even with the position all the way back the cockpit is just way too long or could use another bench in there. I'll chew on all this and see what I can do for the next patch.

as for knowing on how boats are layed up and how heavy they should be completely depends on the boat and the builder, for example you could look at a 47 ft Apache powerboat from the mid 80's and it'll be tipping the scales at around 22,500 lbs. now look at a modern cigarette marauder, its three feet longer but only tips the scales at 14,000 lbs with twin engines and 18,000 with triples. the main thing is the amount of materials used and how they're used. a formula 419 fastech from the year 2000 weighs in at just over 13,000 lbs dry with stock power. once everything gets straightened out in the future you should be somewhere around 10-11k just because your running much lighter engines (1500 lbs vs 4620, the latter number includes drives)

as for the weight issue, its more of an educated guess and it depends on what you want the end goal to be around, a boat using current technology is going to weigh less than a boat from the mid 80's just because the technology has gotten more efficient and effective so they have to use less material to get the same strength. does the current weight increase deal with the weight per area of the boat?

also with the rigging of the boat, a boat that is 50 ft is going to hold a lot more fuel, and gas is heavy. i know that boats in this range are holding anywhere upwards of 200-300+ gals. (250 gals weighs about 1,600lbs) with regards to the engines, what is the average weight between the major outboard producers? (suzuki ~604, mercury ~635, evinrude ~558, yamaha~763) this is what i gathered from their websites, may want to bump that 500lbs up... (i used the lightest configuration, or the shortest shaft length) along with this you could add the seven marine engines to help the bigger boats, this wouldn't be able to be used on the smaller boats because of the shear weight of the engine (1,094 lbs)

i just spent the last half hour or so looking at your progressive audio on youtube and looking into it deeper, really is some cool stuff. and i look forward to seeing that v8 being used!

Good info, thanks. I'll refer to that when I get a chance to look into the weights.

To answer your question, the weight of the hull in the simulator is essentially surface area times some value. So yes, it's weight per unit surface area. The center of gravity position is computed from that too, it's all integrated over the surface much the same way a finite element program would do it.

Different parts of the hull in the sim have different values in order to try to keep the cg low and forward, so it's not all uniform. Basically I had tuned the 0% weight modification setting to fit a range of small speedboats like Hydrostreams and Allisons. Bass boats were significantly heavier though, I was surprised when I went looking for numbers. I was going to cap the weight increase to 50% until I saw some bass boat numbers. There are no striders here though, that was something I was looking at as an excuse to get the cg lower. It doesn't matter much to top speed performance in a straight line, but makes an enormous difference in turning performance.

Good point about fuel and engine weight. The current fuel weight in the sim is 120 lbs, driver is 170 lbs, and as you recall the engines are 500 lbs. Maybe I'll update that or make the different engines have their own weights (somebody asked for that awhile ago, maybe it was you?). The more I experiment and play with the simulator the more I realize how critical all these things really are. A few pounds here or there makes a big difference, especially if it's shifting the center of gravity. Just a couple inches forward/back or up/down makes a noticeable difference.

Procedural audio: Thanks. That's all some highly experimental stuff I was doing before I started the boat simulator. That sound is all generated from a gas dynamic simulation of the exhaust system. It's remarkable that it works as well as it does. A CPU can't do that, the computational requirements are far too high. This was all done in a compute shader on the graphics card which is far more powerful because it's set up to do highly parallel computations which is exactly what that type of simulation needs.

There's still room for improvement in that system, for technical reasons there's no header system in that, it's just a bunch of straight pipes like zoomies, basically. I did try it in the boat simulator early on:

<a href="https://www.youtube.com/watch?v=hAO7SMqUfI0" rel="nofollow">https://www.youtube.com/watch?v=hAO7SMqUfI0</a>

Not quite as pretty looking back in those days. 😉 I was surprised it actually still ran with the boat simulator running on top of it, the engine simulation was doing something like 43 billion computations per second just to simulate the engine and generate the audio. Yikes. It was pushing things too much, there wouldn't have been room left to do much more than what you see in the video, and that was with a $500 graphics card at the time.

In the end I put it to the side and went with traditional samples which still sound better than my system (unfortunately). Maybe some day I'll get back into that and try to do proper headers in it. Right now the computing power isn't really there to do everything I want to on that end.

so with the procedural audio is it possible to get those computations per second down enough to be able to run in the simulator? because in my opinion they sound better. if you added headers would it make it even more strenuous on the computer? i know that they would completely change the sound of the engine and that seems pretty tempting.

along with the length of the boat you'd have to add in a way to increase fuel for the length of the boat, maybe do it by surface area?

it may have been me asking for the weight of the outboards to be increased, i vaguely remember it. also looking at your older videos on you tube is the torque of the engines still the same? because it looks like they top out at about ~475 ft-lbs, which looks a little high to me. i know that its very hard to get the torque curves of these engines because none of the companies advertise them, possibly contact someone? also that HUD that you have seems like it would be a nice addition to the game if someone wanted to more accurately tune their boat, and also see what the sim is doing.

what are striders? do you mean stringers?

Yes, I meant stringers, not striders. 🙂

Procedural audio: I couldn't find any more big cuts to be made there. Simulating gas flow in real time is a tricky business, I'm not even aware of anyone else having done it for something like this. To give some idea, the boat simulator at the moment runs 100 computational steps per second. The real time engine system, because of the extreme lightness of air and the tiny propogations in pressure and so forth through the exhaust system that need to be computed and tracked over small distances in space, needs to run 480,000 steps per second. Massive difference. That was about as low as I could take it too. If it goes too low, the system becomes unstable and boom, the simulation numerically explodes and the audio just stops. So it was pretty much down to the wire already. There are other things that can be done like increasing the size of the computational nodes in the exhaust pipes, but the audio quality drops rapidly as you do that because small state changes (pressure, velocity, etc) aren't handled as accurately. So there's a tradeoff between computational speed and simulation/audio quality.

Headers would add quite a bit to the computational requirements. It's hard to explain unless you're familiar with memory architecture on GPUs and the different types of memory accessible by compute shaders, but in a nutshell, things change considerably once you start connecting those exhaust pipes together into a network. Memory has to be used differently and that in itself can cause a monumental slowdown, like by a factor for 100 or worse. When I hit that wall, I chipped at it for a bit but couldn't figure out an easy way through it so called it a day (a few months, actually) and set it to the side. So there's a big technical hurdle there to doing headers and connecting the pipes together. Meanwhile the rest of the system is as optimized computationally as I could make it. I even managed to handle all the boundary conditions without using any "if" statements, so it's pretty much just raw mathematical computation.

Maybe some day if/when I can get back to it and give it another try, I'll see something that could be done differently. That won't be any time soon though. I'm glad you like the sound though, I personally preferred the sound we've got now with the samples. At least it sounds like a 300XS. The real time system to me still sounded a little bit too synthetic. Maybe some day, but for right now the hardware isn't quite up to the job, and I need to save some room for other compute shaders. I'm working on a water spray system right now that could benefit greatly from bunging things off the CPU and onto the GPU, so that'd be at least a little bit of space that would no longer be available for the engine simulation.

It's always in the back of my mind but right now there are more important things to do, like adding more types of boats and so forth. As you can see by the number of people in the forums, this project isn't doing very well and eventually I'll run out of money and have to stop if I can't turn things around. So I've got to pick and choose each new development task carefully, and the engine audio stuff is near the bottom of that list at the moment even though it's very cool technology.

Fuel increase with length: Yes, something like that could be done, also with the densities of the boat surfaces. Just need to give it some thought and figure out a way to let it all scale up properly. I'll try to get to that sooner rather than later seeing how far off the weights are on the big boats. I actually noticed that the day I was putting out the patch, but decided to go ahead and release it anyway and take another look at the weight handling another day.

Engine torque: The old videos were different engines completely, just a very simple curve system I could adjust while the simulator was running. What we have now are actual outboard engine dyno results for some of the engines. I took a few independently measured (not manufacturer supplied, I wouldn't trust those anyay) outboard curves in the 150hp-300hp range and then filled in the missing bits by scaling things to get the power right. So they should be quite close to typical outboards.

HUD: Originally I was going to do that, but this has got to run on Oculus Rift and a HUD like that is tricky to make work well in virtual reality without it being really distracting and ruining the immersion. In virtual reality, those displays have to be sitting in 3D space somewhere far enough away so you don't have to go cross eyed to read them. Next thing you know, numbers are floating through the dashboard, steering wheel, etc.. You can see that in the design screen with the panel, as you move the boat closer it goes right through the panel. So I made an exception there because it really needed to be that way for this to even run in virtual reality.

In non-Virtual reality this is easy, you just draw the numbers on top of everything else after it's rendered like you see in some of those older videos. So in that case it's just 2D, no matter how close some 3D object is to your face, you can still see the numbers. So whatever I do there has to work in both non-VR and VR the same way.

seems like the procedural audio is quite a handful, and it seems like you've got the outboard engine part handled. with regards to the HUD, couldn't you have a toggle for it in the options menu and recommend that you not use it with the oculus? that HUD that was in those videos seems like a pretty valuable tool to have.

what'd you say about adding new boats? seems like that would be quite the handful at this stage when there are other things that could use some more polishing. have you put much thought into changing props on how they work and are customized (pitch, rake, cup, blades, ect) or is that something that just makes your head hurt (because it makes mine). also is the prop just running off of the simple triangle that's in the game right now or no?

also what about the props rotation, i read the change log and wanted to know if changing the props being in or out would really change much to the boat handling and speed wise.