Thanks for watching guys. Learning a little each day using Brilliant will help you avoid becoming the guy in the last clip. To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/AngeTheGreat . You’ll also get 20% off an annual premium subscription.
@@Iliedavidrealchannel if you mess with the exhaust piping in your car and separate the tubes for each cylinder you can make it sound amazing. this simulation shoes why things like a Bugatti spend so much time and money on that
For sure! Older Japanese bikes with indivudual pipes for each cylinder do exactly the same thing! I actually LOVE it when they sound like that. CBX 1000 does it as well
I wonder how much of the actual exhaust system behaviour could have been predicted at that time during the design phase. It sounds like the beating between the pipe resonance frequency and the actual exhaust gas pulses might affect the engine by periodically aiding or impairing the scavenging (just a wild guess). Aside from reducing the Q of the pipe to "spread" the resonance, could this be controlled somehow? Perhaps mixing in the pressure wave from another cylinder is enough?
I was going to comment on this as well, I was especially impressed because that is a sound I’ve not heard in a long time. This makes me want to spend money I don’t have just to access the patron content. (Why I stick to Steam.) Edit: Autocomplete thought “Steam” was “Stream.” Fixed.
i noticed something, some race cars actually had similar shapped exhausts and they made the same wobble sound, especially decelerating... for example lancia 037 if you listen to recordings of it youll recognise the similar sound
The old simulator sounds like an electric toy in comparison, whereas if you close your eyes, you're absolutely riding that bike with the new simulator. Very well done.
I was one of the people who was a little disappointed at the single cylinder sound. This completely obliterated any concerns I had back then. Honestly this already sounds absolutely amazing, especially the comparison with the unprocessed sound from ES2D shows how far this project has come. This completely restored my faith and I will worship at the altar of Tubular until this gets released!
This is one of those things that, if you know what's actually going on, feels criminally oversimplified as a model. It never had any right to sound remotely plausible at all. But as more fidelity gets added over time, it'll get closer and closer, until at some point it becomes indistinguishable from recorded sounds of the real thing. I wouldn't be surprised if at some stage this becomes mature enough to be picked up by the engineering teams working on real engines as a first pass to weed out the silliest ideas before any real CFD gets done.
Really cool to see that the team behind BeamNG is one of your supporters. I guess that means they intend to incorporate engine sim into their game in the future. THAT would be sick.
If I'm not mistaken, the wobbling you spoke about at 15:46 is a result of exhaust air resonating in the system. I think that sound happens irl too with some exhaust systems. You'll of course want to check and verify but if it is because of a physical property and not an issue with the sim, it means you're doing very good with the exhaust side of the simulation
I haven't heard of this phenomenon before, but I know in my NA 2.0l focus, there's a noticeable shift in tone between 3-4k rpm and it changes in almost exactly the same way as the simulation. Sounds throatier, almost like there's a lower octave that creeps in and out as you go through the rpms. Again not sure of the mechanics of it, but it seems the simulation matches reality.
@@themonsterodub I'm not entirely sure if that is exactly the case either. Going off of memory, when the air is being compressed by the piston going up in the exhaust stroke (this increases the pressure in the exhaust) there's a return shockwave that can constructively interfere with the exhaust air that is being expelled. I am curious if that is indeed how it works but I wouldn't be surprised if it was some other mechanism
@@themonsterodub In that part of the simulation he's using 4 individual pipes, in a real car the individual header tubes would link together somewhere in order to help pull the exhaust gasses along using scavenging. This also somewhat reduces audible resonance, but if you listen to single cylinder engines or v-twins with 2 exhaust pipes you will sometimes get bizarre resonance out the exhaust.
The audio at 15:50 literally gave me goosebumps. This is the future of racing video games, sim and arcade. You deserve millions for this masterpiece of coding. You've done it, minus a few tweaks you said yourself you want to do, and with another head or two on the project in the future, you are well on your way to making some SERIOUS cash renting this software to big corporations for both fluid simulation purposes and game dev. You are brilliant.
A simulator this complex would never make it into a racing game, you'd need a core i12 69420k to run this and the already intensive game at the same time
@@theairaccumulator7144 That would depend on if the engine sim could be off loaded to some spare threads, most games still aren't using. I imagine most interaction with the sim once its set up is very little data (fuel + air in, and some resistance on the main shaft and then power + audio out), with audio likely being the overwhelming majority of cross talking.
I still come back to this video every few days just to listen to that sound at 15:46. It's incredible to the point where it's not the nerd in me that is getting excited, but the petrol head
The trick to having multiple things operating on a single state is to double-buffer it so that CPU is reading from previous frame's output, and GPU is rendering with it, while CPU (or whatever device is performing the simulation step) outputs the simulation increment to a back buffer. Then you swap the buffers' roles. I'm sure you're well aware of double buffering, but sometimes we forget things!
@@gabrielv.4358not really? You can buy a high core count 10 year old xeon with 16gb of ram and a decent enough motherboard for 100 dollars off of aliexpress. It's literally the only way to go Imo. Like for example if your budget is like 300 dollars, you can buy a rx580 or 5500 for 100 dollars, and the other 100 dollars you can spend on a cheap case, cooler, power supply and m2 sata ssd. It might be a stretch but you can do it in this budget range. But if you're upgrading than you don't need to worry about those.
16:11 the resonance is a real thing, its good that its simulated. the reason its so prevelant is because usually resonance is a lot tougher when there are multiple cylinders going into one exhaust (im pretty sure at least) either way, it sounds very very good!!!!
That kind of resonance is usually cancelled-out by exhaust mergers, if it exists at all. There are NVH factors in real vehicles, but the one in the sim sounds a bit odd.
that "wobble" is actually very realistic as some engines do that at high rpms under engine load (accelerating) I was chocked when I heard that and actually sounds a lot like my real bike at those rpm, Amazing job 👍 edit: while looking at the video again, the wobble appears when the presure in the intake valve increases, due to the high rpms/load ratio, some vehicle's get rid of that with some exaust desing but definetly this is prof that your simulator works beautifully
The smile I had on my face after hearing the acceleration through the gears and that growl at 15:48 Excited for the future where we can test and hear the differences in exhaust layout for different engines. Imagine being able to find a good sounding exhaust for the S55 by building the engine and exhaust in this simulator and figuring out what makes it work digitally first
literally laughed out loud FOUR times with the degenerate engine creations, abducted by aliens, core 2 duo seek help, and 3am Wal-Mart parking lot comments
You bet. BeamNG being one of the top patreons of this channel and subsequently of the engine sound project might have given us a subtle hint. If BeamNG doesn't have first dibs on that I would be surprised.
I predict luxury electric cars will get it soon, followed by all electric cars on a long enough timeline, to make up for a lack of engine sounds. Those companies have huge amounts of cash to throw at something like this and clearly want it given how common and tragically forced fake engine noise is already. Even current gasoline cars like the BRZ have garbage fake engine noise from the factory
Eh. Idk. Simulating an engine on this level of detail uses way more resources than whats necessary just to properly emulate the sound of an engine. Maybe car sandbox games or mechanic games, but certainly not every game in the near future. You need a decent rig just to run this simulation. Just imagine what that would cost if you have multiple cars running the same simulation.
Something i've noticed with car's exhaust is what i call the drainpipe effect. When the exit of the exhaust has a long unmuffled pipe section beforehand, for example vehicle with 2 mufflers but with rear muffler replaced with pipe, there's a certain warbling resonance that can be heard. Engine Simulator 2 has it, that sound was there around 15:17. Superb work.
I've got a 750cc GSXR, and that resonance sound you mentioned about the exhaust in the new simulation really does happen at certain RPMs. Your work blows me away. Intellect can take a man far, but your work and determination on top of it is what leaves me marveling at this.
The non processed comparison of OG vs current iteration ES is absolutely wild. What a staggering amount of process. Not to mention, it's still a WIP. I can't wait to hear the final version
I hope this method becomes the industry standard. Engine sound is honestly one of the most important things to me in a car, and this will get the realism so much closer than it's ever been. Not to mention all the possibilities of simulating really uncommon/non-existent engine types. Like, I'd love to hear a 14 cylinder radial without a prop on it, or that Chrysler six bank 36 cylinder.
Ange, seriously dude. Every single time you come out with something that blows my mind even more. your creation gives me GENUINE goosebumps with how good it sounds. But not only the quality of the audio, the visuals and intricacies you put into your work are stunning and jaw dropping. PLEASE DONT STOP! No lie, every time you demonstrate a new type of motor layout and you get on the gas, banging through the gears, it makes me shiver. Its amazing what youve acomplished.
I had no idea anyone complained about the new version sounding less realistic, and I'm giggling now because OF COURSE it initially sounded less realistic, it was really early in development. It's STILL really early but SO GOOD.
I don't know if you have any musical inclinations, but it'd be awesome to be able to design your own brass and wind instruments with such an accurate simulation.
Sounds fantastic dude. The deepening with exhaust length sounds really nice. The resonance "wobble" is accurate: There's always an RPM range where an exhaust sounds raspy and terrible. Also, changing cam profiles changes the sound of an engine dramatically. Very cool! Love your work
That's incredible. I can't believe that sound is being generated in real time, all from code. That is really crazy. I wish I could contribute. I have a very sharp ear for details. It would be very cool if the model would take into account how different cam specs would negatively impact the idle and light-load scenarios (rough idle / misfires / partial combustion) to give it a little extra life. And/or, when the engine hits cutoff, how it will often pop or bang as ignition / fuel is re-enabled. It would be cool if these were not necessarily just options, but the result of fuel / ignition tables and scalars. Either way, this is very impressive work. Thanks for sharing it with the community.
This is insanely cool. The future of sim racing and sound generation lies here. Next step ideas (as a car guy I guess): simulation of variable intake headers (like the mazda 787B), carburetors (the unperfect flow sound), and turbocompressor sound (recreating the whistle would be amazing). thank you for your time invested in this, you're a genius
15:44 sounds like it's on an engine dyno and immediately makes me think of the Honda and Renault F1 V10 engine test videos! This sounds incredible and I cant wait to hear what you'll achieve at those higher cylinder numbers!
I was in awe when Doom came out with 2.5D graphics. Amazed when 3DFX cards appeared. Poor when I bought a direct drive force feedback wheel in 2011. Blown away by VR. This is the closest I've felt to those experiences for years!
What a huge update! One thing I can't wait to look forward to with the flow simulation idea exhaust tubing, I can 100% see this program being used by exhaust shops for pipe diameter, exhaust shape, tone and reverb!
how could have people said that his was a worse upgrade, like bruh this new hayabusa engine that uses fluid dynamics to make sound is so much better sounding. in this demo stage of the new engine simulator seems very damn good already omo
I have watched these videos because I liked programming, and at some level acoustics. Seeing you now just straight up have sponsors and seeing people using your stuff in random youtube videos has been so cool! Keep at it, and good luck! I know at this point you would never change (and rightfully shouldn't) but I would like to say that I have been doing my personal audio analysis project in go(lang) and the multithreading / concurrency model made my life so much better and I can consistently keep all cores fully saturated and working with lower overhead than I thought possible. Hopefully you can figure out a good way to handle thread sync as it has the potential to do a lot, especially as core counts keep rising and you become due for a new pc.
I'm impressed at how good the simulation is, running on a 6 core CPU that's half a decade old. I had assumed your rig was much beefier when you were doing complex simulations. Then, you pointed out that it's only single threaded. That's crazy. Sometimes I forget how fast modern CPUs are. It's almost unfathomable to realize how fast these things are. Also, gurgling snorkele engine says hello. 17:42
This video makes me realize how the old engine simulator sounded like you were in a cabin of a car or something. Sounds muffled. This new version sounds fucking sick, can't wait to see what comes with it!
Never had any doubts that this would sound far superior. Can’t wait to hear it after a bit more polishing and support for exhaust junctions. As it is right now it would probably fool me if I didn’t know. Especially if there was some background noise and synced to a video.
The downshifts… this project is absolutely materializing into more than the expectations set by the earlier progress videos… and those videos set HIGH expectations. I love the comprehensive of this simulation, and the result is SO good
The fact that I'm playing this video in the background and had to look over if a video of me on my busa was being played somewhere really says something about this program
The pull at 15:45 was astonishing, I have a personal love for the downshifts. Only thing it's missing is some of the afterfires. Mine has some fat two brother exhausts so it's nothing obnoxious but it's definitely there.
I am an F1 PU Systems & 1D simulation engineer, and your job is certified as completely incredible. GT-Power at the door of everyone. Congratulation dude 🎉
One thing you mentioned were junctions. I think those might help out with the simulation because instead of having to simulate multiple long pipes, there would just be the one pipe leaving the engine. I do know that exhaust systems tend to split and merge multiple times depending on the vehicle so it seems like something that is quite important to the sound. But regardless, this is a project I am deeply interested in. I have spoken to many people about this irl including my college professors. One of them even wanted to see about having you come in to present ES2D and ES3D.
He did say not supported YET! So it is a feature he is going to implement. But what you have mentionned is true engin sound does matter depending on the exhaust configuration, just search up x pipe vs h pipe that already makes a big difference in sound depending on the car.
Man every single time you make a video, you play it down and make it out as if it's not one of the coolest individual creations ever. The love you put into it shows through with just how good this sounds, and it's clear just how much work you put in. It sound absoutely insane and it's still a work in progress! I've heard worse audio on other videos of REAL CARS on UA-cam! If you blindfolded me and had me listen to @15:50 onward and a real exhaust camera (like gopro stuck above exhaust in supercar videos), I'd legitimately have a hard time telling you which one was real.
16:15 I used to have an Acura CL Type S that actually had a similar type of resonance at part throttle at a certain rpm range, I think its all in the exhaust design.
Clearly a lot of hard work has been put into this and the results are pretty convincing. Seems that you got a solid grasp on some of the real math happening behind physics of a real combustion engine. This is appoaching a model that could be used in automotive industry for designing better exhaust systems, for example. I see much potential in procedural content generation. In a racing game, it might be impractical to simulate engine sound and dynamics constantly using a model like this one, but one could "pre-render" samples for wild custom engines that the player could build and tweak like a real engineer.
When you showed the exaust noises getting lower when you made the pipes longer, I thought "Of course, this is how a trombone works." I dont know if it would come across, but I find the idea of the moveable bend of a trombone as part of some engines exaust hilarious. Maybe you can use that in some semi-serious testing. :D Edit: Also thank you for that "degenerate" part, thats totally my alley. lol
I would love to see a visual representation on where you put the microphone if there is one, would also like if your demo have the microphone moved dynamically. Imagine hearing the intakes too.
@AngeTheGreat I do hope this reaches you. Allow me to contextualize the following: I am a 20+ year master mechanic, who has personally owned, worked on, and operated 13 different cars/trucks, and 13 different motorcycles, and worked on countless others. I am mildly autistic, and thoroughly fascinated with the operation, and the SOUND of each and every component of an engine. With these traits in mind, I trend towards objectivity, and less towards overlaying my preferences and emotions (remember: autism. How to emotion??). I tell you that to tell you this: I'd like to express my observation, that those who seem to criticize and question the thing they have no control over, seem to react as such (YOUR project vs THEIR experience - come on, people. LET HIM COOK.). That being out of the way, I hope it provided context for my next statements. I have been following your progress on your previous projects, and I am thrilled that the almighty algorithm has favored you - YOU. HAVE. EARNED. IT. The path, and every milestone along the way, has lead you here. And I am ecstatic that it has. I am thrilled to be able to observe the fruits of your labor grow! I have trust and faith in the way you are tackling this project. I have a DEEP appreciation for everything you've done to create as accurate a simulation as possible. I'm not a programmer, nor am I an engineer. But I'm thoroughly impressed at the now raw, procedural sound you've managed to replicate - I was really tickled at how the old sim had additional audio processing to achieve SOME level of real-world sound, and how the new sim doesn't need it at all to sound just about perfect to real world! I am particularly interested at the content after the 16:00 minute mark - the 'wobble'. My intuition on the subject, and the reason I've decided to write and share, is thus: I have experienced this resonance, especially on motorcycles, where the experience is very raw, direct, and visceral. I think your simulation is extremely accurate! Resonance is a very big deal in the world of engineering an engine - it is absolutely crucial to scavenging, efficiency and performance. I know you already know that. So on to my intuition. If it's simulated, rather than addressing your code, which I feel the results would suggest is very accurate, I'd point you towards the area and angles of the intake and exhaust ports (particularly the exhaust, but both matter) and see if it does not directly correlate with a change in the behavior of the resonance 'wobble.' If these parameters are simulated, and changes to the design of the ports results in a direct change in the behavior of the resonance, I think we can reasonably conclude that you're 100% on the right track! And if these parameters are not simulated, then I am talking directly out the wrong cheeks and I would be embarrassed to have written this college paper and consumed your time in vain. Either way - I love what you're doing. I'd hazard a guess that most people who view, but do not comment, feel similarly, and trust and understand the kitchen is properly staffed - and those who criticize are the exact same people in starbucks saying their drink needs more ginger, and when the employee takes the cup, turns around and turns on the grinder for 8 seconds, then hands the customer the exact same drink, unchanged, who then take a sip and say, "perfect"!
I paused your video thinking "thats pretty good. its not perfect though" then i got distracted, heard the 4 cylinder revving, thought "howd that video start again?" then realized it wasnt your video, it was a car outside that sounded identical... so good job.
Ah now that's more like it! I was the audio engineer in one of your early videos asking why you felt the need the use convolution for the pipes. Now I hear why haha. This one sounds very close. It seems to be missing some inherent resonance properties of metal material but the reflections inside the pipe as a general solid body sound great. Incredible work, and thanks again for sharing your journey and knowledge 🙏
This is amazing! The first time you ran the ‘Busa engine in ES3D for the video reminded me of what it might sound like recorded on an old microphone. For the 3am Walmart engine, only thing missing is the “crackle tune” (where it is just random pops and bangs doing nothing, as opposed to the metered and consistent sound of anti-lag 😂😂) Amazing work, Ange. I may not be able to support right now, but I look forward to at least being able to purchase this in the future. Best wishes!
Thanks for watching guys.
Learning a little each day using Brilliant will help you avoid becoming the guy in the last clip. To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/AngeTheGreat . You’ll also get 20% off an annual premium subscription.
car go speed
Is your simulation software able to run a Helmholtz resonator? (I assume it can't yet do so, since you mentioned it can't do splits yet.)
Some of the videos in your playlist are not correctly arranged in chronological order. For example, Devlog #4 comes before Devlog #3.
@@AmaroqStarwind I'll fix it, thanks for letting me know
People of Walmart will never recover from this sick burn.
the sound at 15:46 is absolutely insane
how do you reply before the video even releases??
Litrally it turns into V8 (actully)
@@SirLaser601 pamtreon :3
@@expired_esoup pamtreon? yes. pamtreon
@@Iliedavidrealchannel if you mess with the exhaust piping in your car and separate the tubes for each cylinder you can make it sound amazing. this simulation shoes why things like a Bugatti spend so much time and money on that
that wobble sounds perfectly reasonable and realistic, it's probably a bad exhaust design and not a simulation issue
For sure! Older Japanese bikes with indivudual pipes for each cylinder do exactly the same thing! I actually LOVE it when they sound like that.
CBX 1000 does it as well
It honestly makes the simulator more impressive to me because it can take those sort of design flaws into consideration and accurately reproduce them
I wonder how much of the actual exhaust system behaviour could have been predicted at that time during the design phase. It sounds like the beating between the pipe resonance frequency and the actual exhaust gas pulses might affect the engine by periodically aiding or impairing the scavenging (just a wild guess). Aside from reducing the Q of the pipe to "spread" the resonance, could this be controlled somehow? Perhaps mixing in the pressure wave from another cylinder is enough?
I was going to comment on this as well, I was especially impressed because that is a sound I’ve not heard in a long time.
This makes me want to spend money I don’t have just to access the patron content. (Why I stick to Steam.)
Edit: Autocomplete thought “Steam” was “Stream.” Fixed.
i noticed something, some race cars actually had similar shapped exhausts and they made the same wobble sound, especially decelerating... for example lancia 037 if you listen to recordings of it youll recognise the similar sound
Hell yeah, man's been cooking.
FOR REALLL
Who let him cook 🗣
@@blue_mate_six5639 who lot him cook ama raihgt
oh dude, the audio is sounding so awesome now! Your progress feels hard earned, and is showing fantastic results
Thanks, I'm pretty excited to see what you make with it 🙏
@@AngeTheGreat I'm excited too :)
Holyyyyy the pull at 15:46 gave me chills
Bro samee
The overrun was the most surprising part to me, as that's usually the part of an engine's sound that games never get right.
The downshifts are soo good🤤
insanely good
@@zenkoz3158I guess BeamNG can ask for the code now lol
The old simulator sounds like an electric toy in comparison, whereas if you close your eyes, you're absolutely riding that bike with the new simulator. Very well done.
I was one of the people who was a little disappointed at the single cylinder sound. This completely obliterated any concerns I had back then.
Honestly this already sounds absolutely amazing, especially the comparison with the unprocessed sound from ES2D shows how far this project has come.
This completely restored my faith and I will worship at the altar of Tubular until this gets released!
It really is some amazing work and a huge leap from before.
This is one of those things that, if you know what's actually going on, feels criminally oversimplified as a model. It never had any right to sound remotely plausible at all. But as more fidelity gets added over time, it'll get closer and closer, until at some point it becomes indistinguishable from recorded sounds of the real thing.
I wouldn't be surprised if at some stage this becomes mature enough to be picked up by the engineering teams working on real engines as a first pass to weed out the silliest ideas before any real CFD gets done.
@@233kosta i would say it already sounds better then recorded sounds coz of the compression and shit most games have mid audio
Really cool to see that the team behind BeamNG is one of your supporters. I guess that means they intend to incorporate engine sim into their game in the future. THAT would be sick.
Engine go vroom vroom am I right 😁
Yes
Very vroom vroom
Real (fake)
Sometimes engine go brap brap
You're god damn right
If I'm not mistaken, the wobbling you spoke about at 15:46 is a result of exhaust air resonating in the system. I think that sound happens irl too with some exhaust systems. You'll of course want to check and verify but if it is because of a physical property and not an issue with the sim, it means you're doing very good with the exhaust side of the simulation
I haven't heard of this phenomenon before, but I know in my NA 2.0l focus, there's a noticeable shift in tone between 3-4k rpm and it changes in almost exactly the same way as the simulation. Sounds throatier, almost like there's a lower octave that creeps in and out as you go through the rpms. Again not sure of the mechanics of it, but it seems the simulation matches reality.
@@themonsterodub I'm not entirely sure if that is exactly the case either. Going off of memory, when the air is being compressed by the piston going up in the exhaust stroke (this increases the pressure in the exhaust) there's a return shockwave that can constructively interfere with the exhaust air that is being expelled. I am curious if that is indeed how it works but I wouldn't be surprised if it was some other mechanism
@@themonsterodub In that part of the simulation he's using 4 individual pipes, in a real car the individual header tubes would link together somewhere in order to help pull the exhaust gasses along using scavenging. This also somewhat reduces audible resonance, but if you listen to single cylinder engines or v-twins with 2 exhaust pipes you will sometimes get bizarre resonance out the exhaust.
The VQ35 displays this behavior on quite a few exhaust setups. I hear it on my own 350z.
17:41 POV: You lost GPS acces and went too far off into the texan countryside when you hear a tuned tracker from behind start to overtake you lmao
Initial D: Dual Corn-Row Drifting
tractor
Ah yes, a tuned Geo Tracker.
Everyone is saying this, and i'll say it again, the sound at 15:46 is CRAZY. I cant wait for more updates, this is looking incredible.
15:45 Jesus christ this sound is insane! I think I just had a eargasm
And it's not even a V10 or a V12 or a W8 ect
Sounded like Honda VTEC but with propper exhaust, instead of that fart can one.
@@poklianon At that point imma go wild I think
Same here! Oh yeah, he definitely gonna make us wait for this one!
The audio at 15:50 literally gave me goosebumps. This is the future of racing video games, sim and arcade. You deserve millions for this masterpiece of coding.
You've done it, minus a few tweaks you said yourself you want to do, and with another head or two on the project in the future, you are well on your way to making some SERIOUS cash renting this software to big corporations for both fluid simulation purposes and game dev.
You are brilliant.
license*
thats why beamng supports him as they do want to implement this idea into their simulation
@LoserSnoozer beam is already cpu heavy now imagine with this implemented youre gonna need a 10950x3d or something
A simulator this complex would never make it into a racing game, you'd need a core i12 69420k to run this and the already intensive game at the same time
@@theairaccumulator7144 That would depend on if the engine sim could be off loaded to some spare threads, most games still aren't using. I imagine most interaction with the sim once its set up is very little data (fuel + air in, and some resistance on the main shaft and then power + audio out), with audio likely being the overwhelming majority of cross talking.
The fact that this is absolutely pure output of the simulation and that it sounds so real is astonishing
I still come back to this video every few days just to listen to that sound at 15:46. It's incredible to the point where it's not the nerd in me that is getting excited, but the petrol head
15:50 hearing that wind up and gear up is just magical to me - keep it going man
ES3D sounds so much better than 2D, I cannot wait until I can play with this
I can't wait to see what engine simulator 4D sound like
@@komodoensis-rexyour screen blows exhaust gases in your face
2D sounds like good video game sound design, 3D sounds like a real recording from real life.
The trick to having multiple things operating on a single state is to double-buffer it so that CPU is reading from previous frame's output, and GPU is rendering with it, while CPU (or whatever device is performing the simulation step) outputs the simulation increment to a back buffer. Then you swap the buffers' roles. I'm sure you're well aware of double buffering, but sometimes we forget things!
Of course he is using double buffering, which has nothing to do with the issue he was describing.
19:20 actually sounds like my miata with the muffler delete
Ah, so YOU'RE my neighbor.
Some of us are trying to sleep, Todd, and nobody finds your car impressive no matter how loud it is.
@@Deathmonkey7I find it impressive. leave my homie todd alone 😠
Stop it, get some help
🤣
Dont stop it, dont get help. Get more gas
"If you're still running a core two duo rig in 2024 I'm going to tell you to seek help." That had me laughing way more than it should have.
This is a havock with low end contries
@@gabrielv.4358not really? You can buy a high core count 10 year old xeon with 16gb of ram and a decent enough motherboard for 100 dollars off of aliexpress. It's literally the only way to go Imo. Like for example if your budget is like 300 dollars, you can buy a rx580 or 5500 for 100 dollars, and the other 100 dollars you can spend on a cheap case, cooler, power supply and m2 sata ssd. It might be a stretch but you can do it in this budget range. But if you're upgrading than you don't need to worry about those.
16:11 the resonance is a real thing, its good that its simulated. the reason its so prevelant is because usually resonance is a lot tougher when there are multiple cylinders going into one exhaust (im pretty sure at least)
either way, it sounds very very good!!!!
That kind of resonance is usually cancelled-out by exhaust mergers, if it exists at all. There are NVH factors in real vehicles, but the one in the sim sounds a bit odd.
that "wobble" is actually very realistic as some engines do that at high rpms under engine load (accelerating) I was chocked when I heard that and actually sounds a lot like my real bike at those rpm, Amazing job 👍
edit: while looking at the video again, the wobble appears when the presure in the intake valve increases, due to the high rpms/load ratio, some vehicle's get rid of that with some exaust desing but definetly this is prof that your simulator works beautifully
The smile I had on my face after hearing the acceleration through the gears and that growl at 15:48
Excited for the future where we can test and hear the differences in exhaust layout for different engines.
Imagine being able to find a good sounding exhaust for the S55 by building the engine and exhaust in this simulator and figuring out what makes it work digitally first
Whenever I feel like a competent programmer, I like watching this guy's channel to keep my ego in check👍
the walmart parking lot is hilarious i didnt expect that
But I can hear and almost see the fartcan civic
literally laughed out loud FOUR times with the degenerate engine creations, abducted by aliens, core 2 duo seek help, and 3am Wal-Mart parking lot comments
@@spamcan9208 I keep laughing evertime I watch that part
im pretty sure ive heard this exact car in a vid in a wallmart parking XD.
Jaw dropping results. Looking very much forward for turbos to join the party. Keep being great, Ange!
19:07 the groan tubemobile approaches
The brapmobile
So many ricered fucking cars sound like this, I simply cannot understand how someone would want to ruin their car like that lol
18:50 You taking shots at Honda owners haha.
I legitimately predict this will in some way shape or form find its way into almost every video game with vehicles in the near future
You bet. BeamNG being one of the top patreons of this channel and subsequently of the engine sound project might have given us a subtle hint. If BeamNG doesn't have first dibs on that I would be surprised.
I predict luxury electric cars will get it soon, followed by all electric cars on a long enough timeline, to make up for a lack of engine sounds.
Those companies have huge amounts of cash to throw at something like this and clearly want it given how common and tragically forced fake engine noise is already.
Even current gasoline cars like the BRZ have garbage fake engine noise from the factory
@@komodoensis-rex Smart - that'd bring in some goooooood bank for Ange
Arma reforger rally
Eh. Idk.
Simulating an engine on this level of detail uses way more resources than whats necessary just to properly emulate the sound of an engine.
Maybe car sandbox games or mechanic games, but certainly not every game in the near future.
You need a decent rig just to run this simulation. Just imagine what that would cost if you have multiple cars running the same simulation.
Something i've noticed with car's exhaust is what i call the drainpipe effect. When the exit of the exhaust has a long unmuffled pipe section beforehand, for example vehicle with 2 mufflers but with rear muffler replaced with pipe, there's a certain warbling resonance that can be heard.
Engine Simulator 2 has it, that sound was there around 15:17. Superb work.
my jaw just dropped at around 16:00 in. I'm not even into cars and I was just amazed
I've got a 750cc GSXR, and that resonance sound you mentioned about the exhaust in the new simulation really does happen at certain RPMs. Your work blows me away. Intellect can take a man far, but your work and determination on top of it is what leaves me marveling at this.
It's a night and day difference! Sounds amazing!
11:51 god that sounds so good, you can almost feel the vibrations in the exhaust, very cool
The non processed comparison of OG vs current iteration ES is absolutely wild. What a staggering amount of process. Not to mention, it's still a WIP.
I can't wait to hear the final version
I hope this method becomes the industry standard. Engine sound is honestly one of the most important things to me in a car, and this will get the realism so much closer than it's ever been. Not to mention all the possibilities of simulating really uncommon/non-existent engine types. Like, I'd love to hear a 14 cylinder radial without a prop on it, or that Chrysler six bank 36 cylinder.
For the radial you can look at a sherman A1/2
Ange, seriously dude.
Every single time you come out with something that blows my mind even more. your creation gives me GENUINE goosebumps with how good it sounds.
But not only the quality of the audio, the visuals and intricacies you put into your work are stunning and jaw dropping. PLEASE DONT STOP!
No lie, every time you demonstrate a new type of motor layout and you get on the gas, banging through the gears, it makes me shiver.
Its amazing what youve acomplished.
I had no idea anyone complained about the new version sounding less realistic, and I'm giggling now because OF COURSE it initially sounded less realistic, it was really early in development. It's STILL really early but SO GOOD.
The gritty sound that the new sim makes when rapidly changing speed is what really sells it.
One day soon I hope to hear a Lexus LFA through your application, I think that would be really dope.
17:43 straight up sounds like a tractor to me
The tailpape sounds too wide... Otherwise you are right
sounds like me after taco bell
I'm just waiting for the diesel engine
Brown sound
to me, it just sounds like a supercar, semi, and tractor combined, which is kinda what the exhaust is
@17:50 yes this is the content I like. I am eagerly waiting to see how your simulator runs a hit and miss engine
I don't know if you have any musical inclinations, but it'd be awesome to be able to design your own brass and wind instruments with such an accurate simulation.
Definitly possible. He allready made a trumpet.
Hey man, appreciate your work and the evolution of the engine simulator. I have a suggestion, simulating the engine vibration of the engines too.
Sounds fantastic dude. The deepening with exhaust length sounds really nice. The resonance "wobble" is accurate: There's always an RPM range where an exhaust sounds raspy and terrible. Also, changing cam profiles changes the sound of an engine dramatically. Very cool! Love your work
That's incredible. I can't believe that sound is being generated in real time, all from code. That is really crazy. I wish I could contribute. I have a very sharp ear for details. It would be very cool if the model would take into account how different cam specs would negatively impact the idle and light-load scenarios (rough idle / misfires / partial combustion) to give it a little extra life. And/or, when the engine hits cutoff, how it will often pop or bang as ignition / fuel is re-enabled. It would be cool if these were not necessarily just options, but the result of fuel / ignition tables and scalars. Either way, this is very impressive work. Thanks for sharing it with the community.
god damn if the final product ends up in BeamNG then Ange better get some of that income, this is godly
He already does beam is his biggest supporter
@@Alucard-gt1zf forgot about that lmfao
13:54 a fully generated, purely digital combustion engine noise is amazing.
It's beautiful to me what technological advancements have enabled for hard working and brilliant individuals like yourself to create.
This is insanely cool. The future of sim racing and sound generation lies here.
Next step ideas (as a car guy I guess): simulation of variable intake headers (like the mazda 787B), carburetors (the unperfect flow sound), and turbocompressor sound (recreating the whistle would be amazing).
thank you for your time invested in this, you're a genius
You are a real software developer and I love how you keep improving your project. I'm sick of other YT devs making single programs for each video
15:44 sounds like it's on an engine dyno and immediately makes me think of the Honda and Renault F1 V10 engine test videos! This sounds incredible and I cant wait to hear what you'll achieve at those higher cylinder numbers!
I was in awe when Doom came out with 2.5D graphics. Amazed when 3DFX cards appeared. Poor when I bought a direct drive force feedback wheel in 2011. Blown away by VR. This is the closest I've felt to those experiences for years!
18:50 ah yeah, the straight-piped Honda's arrived
19:00 the fact that i know cars that sound exactly like this is crazy
What a huge update! One thing I can't wait to look forward to with the flow simulation idea exhaust tubing, I can 100% see this program being used by exhaust shops for pipe diameter, exhaust shape, tone and reverb!
how could have people said that his was a worse upgrade, like bruh this new hayabusa engine that uses fluid dynamics to make sound is so much better sounding. in this demo stage of the new engine simulator seems very damn good already omo
I have watched these videos because I liked programming, and at some level acoustics. Seeing you now just straight up have sponsors and seeing people using your stuff in random youtube videos has been so cool! Keep at it, and good luck! I know at this point you would never change (and rightfully shouldn't) but I would like to say that I have been doing my personal audio analysis project in go(lang) and the multithreading / concurrency model made my life so much better and I can consistently keep all cores fully saturated and working with lower overhead than I thought possible. Hopefully you can figure out a good way to handle thread sync as it has the potential to do a lot, especially as core counts keep rising and you become due for a new pc.
Yoo that's totally Tubular
This pun is *driving* me into the ground
I'm impressed at how good the simulation is, running on a 6 core CPU that's half a decade old. I had assumed your rig was much beefier when you were doing complex simulations. Then, you pointed out that it's only single threaded. That's crazy. Sometimes I forget how fast modern CPUs are. It's almost unfathomable to realize how fast these things are.
Also, gurgling snorkele engine says hello. 17:42
This is awesome. I hope you're truly proud of what you've achieved with this. Thanks for sharing with us all, I can't wait for further development!
16:17 the improved response to low intertia really makes it sound like an old racecar while downshifting... incredible
This video makes me realize how the old engine simulator sounded like you were in a cabin of a car or something. Sounds muffled. This new version sounds fucking sick, can't wait to see what comes with it!
Never had any doubts that this would sound far superior. Can’t wait to hear it after a bit more polishing and support for exhaust junctions.
As it is right now it would probably fool me if I didn’t know. Especially if there was some background noise and synced to a video.
The sounds are def a massive improvement!
19:10 I LOLed. Seriously, this is so cool. As a mechanic who loves playing with engine sounds, this absolutely does it for me.
This is absolutely awesome! Well done! There’s so many practical applications for this!
The downshifts… this project is absolutely materializing into more than the expectations set by the earlier progress videos… and those videos set HIGH expectations. I love the comprehensive of this simulation, and the result is SO good
The fact that I'm playing this video in the background and had to look over if a video of me on my busa was being played somewhere really says something about this program
The pull at 15:45 was astonishing, I have a personal love for the downshifts. Only thing it's missing is some of the afterfires. Mine has some fat two brother exhausts so it's nothing obnoxious but it's definitely there.
This might actually work to help me develop an exhaust setup that doesn't sound terrible for a V4 swap I'm working on. Amazing stuff.
Sounds incredible, keep it up Ange!
I am an F1 PU Systems & 1D simulation engineer, and your job is certified as completely incredible. GT-Power at the door of everyone. Congratulation dude 🎉
You know it's gonna be a great night when AngeTheGreat uploads
This is the most mindblowing thing ive seen this year. Can't wait to see what it becomes in the future.
One thing you mentioned were junctions. I think those might help out with the simulation because instead of having to simulate multiple long pipes, there would just be the one pipe leaving the engine. I do know that exhaust systems tend to split and merge multiple times depending on the vehicle so it seems like something that is quite important to the sound.
But regardless, this is a project I am deeply interested in. I have spoken to many people about this irl including my college professors. One of them even wanted to see about having you come in to present ES2D and ES3D.
He did say not supported YET! So it is a feature he is going to implement. But what you have mentionned is true engin sound does matter depending on the exhaust configuration, just search up x pipe vs h pipe that already makes a big difference in sound depending on the car.
Man every single time you make a video, you play it down and make it out as if it's not one of the coolest individual creations ever.
The love you put into it shows through with just how good this sounds, and it's clear just how much work you put in. It sound absoutely insane and it's still a work in progress! I've heard worse audio on other videos of REAL CARS on UA-cam!
If you blindfolded me and had me listen to @15:50 onward and a real exhaust camera (like gopro stuck above exhaust in supercar videos), I'd legitimately have a hard time telling you which one was real.
16:15 I used to have an Acura CL Type S that actually had a similar type of resonance at part throttle at a certain rpm range, I think its all in the exhaust design.
15:23 I love how accurately you get that "pipe sound"
Idk how to describe it but that sound you get when you speak into a pipe
Very cool
This makes me very excited for whats to come. Well done Ange! ~AngleTheGrinder~
Clearly a lot of hard work has been put into this and the results are pretty convincing. Seems that you got a solid grasp on some of the real math happening behind physics of a real combustion engine. This is appoaching a model that could be used in automotive industry for designing better exhaust systems, for example. I see much potential in procedural content generation. In a racing game, it might be impractical to simulate engine sound and dynamics constantly using a model like this one, but one could "pre-render" samples for wild custom engines that the player could build and tweak like a real engineer.
When you showed the exaust noises getting lower when you made the pipes longer, I thought "Of course, this is how a trombone works."
I dont know if it would come across, but I find the idea of the moveable bend of a trombone as part of some engines exaust hilarious. Maybe you can use that in some semi-serious testing. :D
Edit: Also thank you for that "degenerate" part, thats totally my alley. lol
It's wild. Never did I think this was possible to simulate in real-time. Even less so on a machine that isn't an absolute monster. You're insane :D
I can't wait to hear a big V8 with the New simulation
I would love to see a visual representation on where you put the microphone if there is one, would also like if your demo have the microphone moved dynamically. Imagine hearing the intakes too.
I like how the highest tier patrons of every channel have the craziest names
Nice, I'm waiting for junctions, so you can simulate effect of exaust back pressure :)
@AngeTheGreat I do hope this reaches you.
Allow me to contextualize the following: I am a 20+ year master mechanic, who has personally owned, worked on, and operated 13 different cars/trucks, and 13 different motorcycles, and worked on countless others. I am mildly autistic, and thoroughly fascinated with the operation, and the SOUND of each and every component of an engine. With these traits in mind, I trend towards objectivity, and less towards overlaying my preferences and emotions (remember: autism. How to emotion??). I tell you that to tell you this: I'd like to express my observation, that those who seem to criticize and question the thing they have no control over, seem to react as such (YOUR project vs THEIR experience - come on, people. LET HIM COOK.).
That being out of the way, I hope it provided context for my next statements. I have been following your progress on your previous projects, and I am thrilled that the almighty algorithm has favored you - YOU. HAVE. EARNED. IT. The path, and every milestone along the way, has lead you here. And I am ecstatic that it has. I am thrilled to be able to observe the fruits of your labor grow!
I have trust and faith in the way you are tackling this project. I have a DEEP appreciation for everything you've done to create as accurate a simulation as possible. I'm not a programmer, nor am I an engineer. But I'm thoroughly impressed at the now raw, procedural sound you've managed to replicate - I was really tickled at how the old sim had additional audio processing to achieve SOME level of real-world sound, and how the new sim doesn't need it at all to sound just about perfect to real world!
I am particularly interested at the content after the 16:00 minute mark - the 'wobble'. My intuition on the subject, and the reason I've decided to write and share, is thus: I have experienced this resonance, especially on motorcycles, where the experience is very raw, direct, and visceral. I think your simulation is extremely accurate! Resonance is a very big deal in the world of engineering an engine - it is absolutely crucial to scavenging, efficiency and performance. I know you already know that. So on to my intuition. If it's simulated, rather than addressing your code, which I feel the results would suggest is very accurate, I'd point you towards the area and angles of the intake and exhaust ports (particularly the exhaust, but both matter) and see if it does not directly correlate with a change in the behavior of the resonance 'wobble.' If these parameters are simulated, and changes to the design of the ports results in a direct change in the behavior of the resonance, I think we can reasonably conclude that you're 100% on the right track!
And if these parameters are not simulated, then I am talking directly out the wrong cheeks and I would be embarrassed to have written this college paper and consumed your time in vain.
Either way - I love what you're doing. I'd hazard a guess that most people who view, but do not comment, feel similarly, and trust and understand the kitchen is properly staffed - and those who criticize are the exact same people in starbucks saying their drink needs more ginger, and when the employee takes the cup, turns around and turns on the grinder for 8 seconds, then hands the customer the exact same drink, unchanged, who then take a sip and say, "perfect"!
This is really good work, if you listen to F1s with zoomies from the 80s youll hear the same character. This is a huge step
Amazing progress, keep up the good work. Hoping to be able to use this to tune my 996 Carrera exhaust at some point...
This is really unbelievable, totally love the result of your hard work❤
i hope this one day leads to a revoltion in racing games. realistic sound would truly be a game changer.
I paused your video thinking "thats pretty good. its not perfect though" then i got distracted, heard the 4 cylinder revving, thought "howd that video start again?" then realized it wasnt your video, it was a car outside that sounded identical... so good job.
Looks awesome!
Thanks!
Ah now that's more like it!
I was the audio engineer in one of your early videos asking why you felt the need the use convolution for the pipes. Now I hear why haha.
This one sounds very close.
It seems to be missing some inherent resonance properties of metal material but the reflections inside the pipe as a general solid body sound great.
Incredible work, and thanks again for sharing your journey and knowledge 🙏
fwah 13:46 that had me going, that resonance makes it so real
Yeah, sound so real
omg its already sounding infinitely better than the engine sim 1
Wow! Your work is incredible!
This is amazing! The first time you ran the ‘Busa engine in ES3D for the video reminded me of what it might sound like recorded on an old microphone. For the 3am Walmart engine, only thing missing is the “crackle tune” (where it is just random pops and bangs doing nothing, as opposed to the metered and consistent sound of anti-lag 😂😂) Amazing work, Ange. I may not be able to support right now, but I look forward to at least being able to purchase this in the future. Best wishes!