Forgot your password?
typodupeerror
Programming Graphics Software IT Technology

Splash, Splatter, Sploosh, and Bloop! 100

Posted by samzenpus
from the wouold-you-like-some-ottamottapia-with-that dept.
Acoustic Bubble writes "Researchers at Cornell University have developed the first algorithm for synthesizing familiar bubble-based fluid sounds automatically from 3D fluid simulations, e.g, for future virtual environments. The research (entitled 'Harmonic Fluids') will appear at ACM SIGGRAPH 2009 in New Orleans in August 2009. Check out some videos of falling, pouring, splashing and babbling water simulations (computed on a Linux cluster)."
This discussion has been archived. No new comments can be posted.

Splash, Splatter, Sploosh, and Bloop!

Comments Filter:
  • Hmm. (Score:2, Funny)

    by Luke727 (547923)

    Was I the only one who immediately thought of a cumshot upon reading the title?

  • Thanks (Score:5, Funny)

    by Anonymous Coward on Wednesday June 03, 2009 @08:37PM (#28204199)

    now i gotta go pee...

  • Graphical Adventures (Score:5, Interesting)

    by Celeste R (1002377) on Wednesday June 03, 2009 @08:37PM (#28204203)

    This is a simulation physicist's wet dream, and I'm sure it'll be somewhere in a graphical adventure soon.

    My bet is that the FPS genre will like this too.

    • I'll just say: BioShock 2 or 3!

      Although I would love more, to see a Terminator 4 game with BioShock 2/3 graphics and a System Shock 1/2 style evil futuristic system (Skynet) in it.

      • by Brian Gordon (987471) on Wednesday June 03, 2009 @09:59PM (#28204831)
        From TFA, three single drops of water took one and a half hours to simulate. The babbling water simulation took over 12 hours.
        • Give it a few years and it'll be built into Havok and used as lazily as every other feature.

        • From TFA, three single drops of water took one and a half hours to simulate. The babbling water simulation took over 12 hours.

          That's okay, we could prerender it and just use the recording! Sure, you could just record REAL water, but, damn it, if I wanted real water I'd go outside! I don't like reality! It sucks! I expect objects to appear in my hands when I step on them! When I die, I expect to respawn in my base(ment)! I want all injuries to color my vision red for a moment and then do nothing to impair me! I want to survive multiple gunshot wounds to the chest just because I happen to be carrying a few first-aid kits (that I jus

    • wet dream

      pun intended?

    • You're sure? Well I'm sure that it's much easier to just record the sound of water than simulate it in real time to get the same sound. We'll see it eventually (maybe when it's even remotely feasible?) but nobody's in a hurry to start measuring their games in seconds-per-frame.. well, maybe CryTek.
      • by master5o1 (1068594) on Wednesday June 03, 2009 @10:02PM (#28204845) Homepage
        I'm assuming that the simulation can then be used to add distortions. Say, water trickling down a creek, and ducks splashing. The two sounds would be distorted by both the direction they are coming from and the interaction of the two sounds. The idea would be to make the sounds more realistic without having to record every possible case.

        Mind you, I haven't yet read the article (or summary) and I am not a sound engineer of any sort.
      • it depends. in modern games you have for the most part fixed and predictable levels and design.

        Imagine in the future games where the game world (environment) is truly active and/or deformable. a bit of an extreme example, but imagine setting off a large number of high explosives at the top of a waterfall, physically changing the land over which the water flows. with this kind of system, the flow of water would change, and as a result the sound would change.

        or maybe imagine blowing up part of a riverbed, or

        • by moortak (1273582)
          I think early uses would just be to have more realistic sounds when walking through streams and such.
      • by ihavnoid (749312)

        Of course, it is cheaper to just use the record right now. However, it isn't that simple.

        What I find, is that the situation is similar to animated videos. You can draw some cartoonish character, a few cuts manually, and send the whole stuff to some country with massive number of people drawing all the in-between scenes *manually*. Just like 'The Simpsons' or 'South Park' or whatsoever.

        On the other side, you can just generate everything using your computer. Create some 3D model of some character, create

    • I can't wait to play this as yet undeveloped game. The procedurally generated splashing noises should greatly enhance gameplay... uh, wait can't they just make fun games that don't require a cluster worth of processing just to procedurally simulate the sounds of fluids?

    • by dkleinsc (563838)

      This is a simulation physicist's wet dream

      Since this is going to be happening in New Orleans, I'd say it's an extremely wet dream. Like up to your attic wet. Particularly with all those splash and sploosh noises.

  • by Sowelu (713889) on Wednesday June 03, 2009 @08:41PM (#28204235)
    I want a simulation program where I can move around rocks and pools and have a water hose. Used to do this all the time in the backyard as a kid, it would be nice to do it without getting wet or wasting water. Wonder how long until this is realtime? My kids, of course, won't get to play with it. They need to play in the REAL backyard.
  • Can they (Score:1, Interesting)

    simulate real systems yet, or just computer-generated ones? Better simulations of the relationships between fluid and sound would be fascinating if applied to superfluids (Ahhh, the soothing sound of superhot plasma).
  • by Opportunist (166417) on Wednesday June 03, 2009 @08:47PM (#28204283)

    Finally bukakke (and less ... pleasant) hentai anime with realistic sound. Yay.

    What? What did YOU think this is for?

    • Re: (Score:1, Funny)

      by Anonymous Coward

      Clearly this is to simulate realistic enema sounds and dripped blood for hentai anime. Bukakke is for weirdos.

  • by youn (1516637) on Wednesday June 03, 2009 @08:48PM (#28204311) Homepage

    In other news from the overpriced useless ressearch dept... the research has found an unexpected application... generating fart sounds from facial expressions... thus giving speech to farts. a student called in for volunteer testing of the system said, "amazing! it actually sounds like me... I was always embarassed because my farts didnt make any sound... now I know wether it is a pzzzt or a plrrrrt or a puffff.... Thank you"

    Privacy right groups caution against wide use of the system, "We got to preserve the right to privacy while farting... imagine if these devices were everywhere? our privacy would be gone like the wind"

  • Why did they take so much time to develop this? Now, they need to develop a more generic algorithm, for other kinds of materials.
    • Re: (Score:3, Funny)

      by fractoid (1076465)
      Generic algorithms are only really useful if you want to model evorution. :P
    • The "generic algorithim" in physical simulations is called "finite element analysis". All they would need to do is swap some constants in the formula for other liquids. For solids they may have to change the formula but the algorithim would still be the same.

      The same algorithim is used in everthing from climate simulation to casting simulations for engine blocks to simulating the trajectory of space probes. The technique is not new but supercomuters have used it to quietly revolutionise science and engin
  • by Ungrounded Lightning (62228) on Wednesday June 03, 2009 @09:09PM (#28204453) Journal

    The simulation sounded somewhat muffled, like the high frequency components weren't right or weren't of sufficient amplitude.

    Can some of the rest of you listen and tell me if it sounds muffled to you too? (I want to be sure it's not my machine or earphones.)

    Might be the CODEC used with flash rather than the original simulation itself...

    • by Carnildo (712617) on Wednesday June 03, 2009 @09:17PM (#28204503) Homepage Journal

      The effect reminds me strongly of the water sounds in Myst. I think it's an artefact of recording flowing water in isolation: without an environment to reflect the sounds, the frequency mix isn't right.

      • Re: (Score:3, Insightful)

        The effect reminds me strongly of the water sounds in Myst. I think it's an artefact of recording flowing water in isolation: without an environment to reflect the sounds, the frequency mix isn't right.

        The reflections make a huge difference to the sound, and probably a huge difference in processing time too. This simulation + an environment around the source + surround sound (eg binaural processing, for us folks with earphones) would be very cool indeed.

        Can we do audio-tracing on a GPU yet?

      • by Gulthek (12570)

        Unfortunately it just sounds like their simulation is (for all of their tremendous effort!) quite lacking.

    • by phantomfive (622387) on Wednesday June 03, 2009 @09:46PM (#28204737) Journal
      Yeah, I agree. In all honesty the sounds are about as good as the animations. Recognizable, but clearly not real. I think they would have done better if they had just gotten a microphone and sampled some real water.

      On the other hand, maybe someone with a good ear can come along and adjust the algorithms until they really DO sound good. Much like computer visual art isn't all that great unless someone with artistic talent is deploying it.
      • by hannson (1369413)

        They sounded a little weird to me also. Perhaps it is actually "correctly" synthesized but sounds weird because the sounsd from a real life splash (which we're used to) is a bit distorted before it hits our ears (echo, noise and other variables). I'm sure this could be made to sound even better (normal) with a sound equalizer and other lightweight tools.

        • I wondered about this too. I think it could do with some reverb and general acoustic massage, but then I thought that might be the point. They are simulating the "pure" sound, which can then be combined with some other model to make it sound like it's in a cave, cathedral, street, etc. Their paper is about producing that initial input, so that's the version they present, although they might do well to include a processed version as well (just for interest).
      • Re: (Score:2, Insightful)

        by ephraimX (556000)
        Er... You're kind of missing the point, I think, which is that it's really neat to be able to get it that good entirely by synthesis. Without doing any sampling.
    • I think that human hearing is very subjective and modelling sound accurately is more important than making flashy videos that sound perfect to you.
    • It is not your machine or earphones, or the codec that is attenuating hi frequencies. Look carefully at the spectrogram accompanying the image. You can expand the image to hi resolution to examine it more carefully. You will note that the peak amplitudes of the frequencies generated cluster from around 500 Hz to 1500 Hz, with much lower amplitudes outside of that range.
    • That's what I was thinking. If they can create such complex algorithms, why can't they encode a simple video and upload it. The one they put up there is totally screwed.
  • by Anonymous Coward

    It's WOULD. And ONOMATOPOEIA.

    I want a refund!

    • by jonadab (583620)
      Indeed, from onoma (third declension, mat-stem noun meaning "name") and poieo (epsilon-contract thematic verb meaning "make"): onomatopoeia, a word that makes the sound of its own name.

      Spelling is so much easier when you know the etymology.
  • Sounds amazingly good for a first paper on the subject, I'm sure the kinks will get worked out as more researchers ply their talent
  • They sound too "chirpy" and "sharp" to me. It seems like there should be more noise in them. I wonder if this is just because I know they are synthesized. We need to do a blind comparison to see if it's good enough.

    • by McNihil (612243)

      I agree the water seems to have the wrong viscosity (too low) making it too tense of a sound.

      But I could be wrong and its the rendering that has a slight speed up.

  • I've always enjoyed utilizing an onomatopoetic nickname.
  • It's technically impressive and all, but what's the point? For games you can just record the real thing, and what other use is there?
    • if you have the code, you can modify it or experiment further to create impressive textural .. ok more life "fluid" .. soundscapes. Add that to music production and whoosh / wow / ahhhhhh ok ...

    • Why not?

  • Might have been sitting around smoking on a bong and thought hmmmmm

  • Do we really need to know the sounds were calculated using a Linux cluster? Linux is popular enough in science now to leave that away.
     
    And I'm sorry, but only the stone that was thrown in the water sounded convincing. I did like how they could add water to the already full container without it overflowing, though!

  • I've been thinking about this type of thing for a while now - getting to a more basic level of sound reproduction, like we've been doing the last few years with graphics. Compare this to lighting in the Quake era, compared to today. Before, we had pre-rendered lighting in the .bsp, or worse (painting it on the texture). Now's it's simulated at a more basic (read: realistic) level, like a lower level emulator, with real-time lighting. And just as Doom 3's lighting was innovative but not terribly practica
    • You need to know a smattering of HTML. The line break is
      which of coarse just broke this line instead of displaying. What I did was left_angle_bracket followed by br followed by right_angle_bracket. This sentence begins with left_angle_bracket followed by p followed by right_angle_bracket, and ends with left_angle_bracket followed by forward_slash followed by p followed by right_angle_bracket

    • Doom 3 had lighting??

      You line break with html: <br />
    • I've been working on this backwards. I developed this wonderful spectrum analyser [tropicalcoder.com] and now would like to stick it into a graphic environment where various visual events are generated from analysis of the audio. Download it and load in a wav file or mp3 and you will see the display. Individual musical notes are displayed. Then you could control a graphic based on the flow of the music, for example. Of course, it has all been done before, but I don't believe I have seen this level of resolution done in real ti

  • who thought that those are the names of new social networking sites?

  • by ledow (319597) on Thursday June 04, 2009 @07:31AM (#28207473) Homepage

    Looks like PC gaming is likely to be heading more and more towards procedural generation of the universe. Real-time shadows, dynamic lighting and now, dynamic sounds.

    It'll all make it more realistic (but at a high CPU cost!) - being able to not have "splish, splosh, splish, splosh" when wading through water but a full-on sound relative to individual parts - bullet shells, limbs, objects in the water, etc. We won't see it *practically* for years, but gaming is getting closer and closer to that dream of "virtual reality", where you won't be able to tell the difference between a real scene and a computer generated one without touching it.

    I think you can make uses of it in gaming too, extending the basic science to a consumer level - skim stones across water that sound like they're being skimmed (and with proper fluid physics similar to that which we already have, individual sploshes and waves etc. affecting that stone) - or be able to throw a coin into water behind an enemy and see if you can use it to distract him. Maybe even, the bubbles that you breath underwater hitting the surface with their true sounds, thus giving your position away if you were hoping that holding your breath would let that enemy walk past you without hearing you.

    When you play games, you don't notice the "cheats" at first - the static sounds that just play on certain events, the pre-lit textures, the echoing of sounds generated inside a certain fixed area. Even in things like HL2, boxes thrown into water either splosh or don't, splosh based on certain primitive criteria that provide a few levels of believability. But as new technology comes along to make it possible to actually *create* that effect rather than script it, everything suddenly feels much more alive.

    Dynamic sound has to be one of the next "big" areas - hitting a wall with an axe in a game used to give "Doink", then it gave a selection of "Doink, Donk, Doink, Donk" sounds each time. Moving forwards, the only way is to actually determine exact angles, shapes of the wall (proper destructible objects for everything are, sadly, still only a dream) and to generate a simulation of the sound it would produce (how cool would it be that if you strike the axe slightly off, you get a reverberating axe coming back at you, with a horrible sound that tells you not to do it?. Maybe even with the axe breaking on a critical point if you mis-use it too much, e.g. try and chop at a steel wall).

    We already have proper echoing and other effects available and 7.1 surround can take away the whole "Where the hell did that come from?" effect if it's too clinically applied. But having sounds *generated* by the interactions within an environment... wow. Imagine Left4Dead-style atmosphere, but with proper echo effects... you walk towards a corner and from around it, a zombie stumbles into a puddle - suddenly the sound not only tells you there's something near, but the echoes from the corners confuse just as in real life, and the sound is only the tiniest little splish, and it may even be possible to determine the *type* of zombie around the corner by the type of splash it makes - something with a large flat foot would create an enormous popping bubble of a sound, something with stick-like appendages would generate barely a ripple.

    This will have a small but critical effect on gaming and, I imagine, a million other uses. But we're *years* away from seeing it used.

    • To get real time sound effects, to begin with you will need use a better audio driver such as ASIO. DirectX just won't cut it because of the latency.
  • Yes but... (Score:3, Funny)

    by IceCreamGuy (904648) on Thursday June 04, 2009 @09:11AM (#28208309) Homepage

    Check out some videos of falling, pouring, splashing and babbling water simulations. (computed on a Linux cluster).

    Yes but does it run on a beow... oh. Sweet.

We want to create puppets that pull their own strings. - Ann Marion

Working...