WebGL superlative overload
category: code [glöplog]
@gloom: the prod you're referring to is done for entertainment, not for the technical challenge. They have no need to push the envelope since there is no competition, hence nothing to be compared to. Actually, for the masses, they are sort of cutting edge: interactive 3D animation synced with some music, in a browser, wow!
And after all, even in the scene some successful (sometimes award winning) groups aren't pushing anything either, "just" making enjoyable experience out of simple effects carefully crafted.
So I don't really understand what you seem to be so unhappy with. Because they're bragging on simple things? I didn't feel they did so, and I agree with Mr.Doob they are just make their explanations accessible. Because some simple prod turned out successful? Well let's get into the bandwagon. :)
And after all, even in the scene some successful (sometimes award winning) groups aren't pushing anything either, "just" making enjoyable experience out of simple effects carefully crafted.
So I don't really understand what you seem to be so unhappy with. Because they're bragging on simple things? I didn't feel they did so, and I agree with Mr.Doob they are just make their explanations accessible. Because some simple prod turned out successful? Well let's get into the bandwagon. :)
smash: Well put.
Zavie: It was just a rant, with a rant-warning front and center. I wouldn't put too much emphasis on the reasoning behind it. :)
Well, I guess I could do this: if anyone wants to do a WebGL experiment (like "Lights") to try to get some press for themselves, their company or team of developers, I'd be more than happy to help out. Either with contacts or design, syncing or anything really. I'd love to see what would happen if a few experienced coders "went at it" with WebGL to make something more new-media pleasing.
Zavie: It was just a rant, with a rant-warning front and center. I wouldn't put too much emphasis on the reasoning behind it. :)
Well, I guess I could do this: if anyone wants to do a WebGL experiment (like "Lights") to try to get some press for themselves, their company or team of developers, I'd be more than happy to help out. Either with contacts or design, syncing or anything really. I'd love to see what would happen if a few experienced coders "went at it" with WebGL to make something more new-media pleasing.
Quote:
It was just a rant, with a rant-warning front and center.
Don't worry, I saw the BEWARE OF DOG signs. ;)
Quote:
Well, I guess I could do this: if anyone wants to do a WebGL experiment (like "Lights") to try to get some press for themselves, their company or team of developers, I'd be more than happy to help out. Either with contacts or design, syncing or anything really. I'd love to see what would happen if a few experienced coders "went at it" with WebGL to make something more new-media pleasing.
Quote:
It's not that I feel that "the demoscene can do better!!!111" either, because I know perfectly well that this is in almost all cases not even close to being true.
;)
I do feel "the demoscene can do better" : the top-notch demoscene groups could make a killer thing in WebGL (and get PR) indeed.
A WebGL version of "Cortex sous vide" would be at least as much enjoyable as Lights, even for a non-tech audience :D.
BUT I have the impression most of demoscene coders aren't attracted by this new platform (mainly because it isn't exactly new, from what I understood: "it's OpenGL but in worse").
Or maybe they are cowards.
(Also, "you can move the camera = interactive prod" sucks . Maybe a more interesting lead is live data visualisation. In the era of public contents and APIs, I think there is creative opportunity there - kooma tried it a bit)
wullon: while you demonstrate impressive quoting-foo, I don't feel those are contradictory statements. Trace is entirely correct when he says that the approach is completely different between demsoceners and media producers. What I suggested was to actually get a few good coders together and see what would happen if they approached the new media experience with their demoscene skill. It's still true that demosceners aren't "better" at making things than people who do it for a living.
..and no, a WebGL-version of Cortex wouldn't impress anyone within the media landscape because it wouldn't make any sense, like most demos. :) It's disconnected and focuses on wildly different things. Like smash touched on: the difference is in taking a simple concept and focusing on that.
as i was saying to gargaj..
a big difference is, in the creative coding numedia world, people actually purposefully want to make things with abstract generated shapes, spherical harmonics, blobs, lines, flat shading, simple physics, particles, etc etc. - things you can do really easily and well with code which are a pain in the ass to do with traditional 3d packages or film. (until cinema4d mograph came along. but thats another issue.) and they love it. they love the hard, clean, flat coded look, the blobby shapes etc.
whereas we, on the other hand, are all totally bored of that shit cos we've seen it 1000 times since the 90s - so we're trying to push and do things that are quite hard to do in code, and look more like film or games or at least really nice 3d renders. and we dont quite get there, because its really hard to do in code - what we're doing is way more difficult and cleverer but it doesnt look as good as, well, a render.
we're trying to win democompos with coded things that have a lot of variety whereas they are using code for one part of a bigger composition, and exploiting one idea over and over in different ways. and they're prepared to use a mix of media - film, prerendered content - where its what is needed to do the job, whereas we try and do it all realtime.
so the strange irony is, we get passed over by the outside world _because_ we've moved past the kind of things that are easy/possible to make look good to the untrained eye when they're coded.
we're so busy trying to impress each other, we make things that probably don't interest the rest of the world.
a big difference is, in the creative coding numedia world, people actually purposefully want to make things with abstract generated shapes, spherical harmonics, blobs, lines, flat shading, simple physics, particles, etc etc. - things you can do really easily and well with code which are a pain in the ass to do with traditional 3d packages or film. (until cinema4d mograph came along. but thats another issue.) and they love it. they love the hard, clean, flat coded look, the blobby shapes etc.
whereas we, on the other hand, are all totally bored of that shit cos we've seen it 1000 times since the 90s - so we're trying to push and do things that are quite hard to do in code, and look more like film or games or at least really nice 3d renders. and we dont quite get there, because its really hard to do in code - what we're doing is way more difficult and cleverer but it doesnt look as good as, well, a render.
we're trying to win democompos with coded things that have a lot of variety whereas they are using code for one part of a bigger composition, and exploiting one idea over and over in different ways. and they're prepared to use a mix of media - film, prerendered content - where its what is needed to do the job, whereas we try and do it all realtime.
so the strange irony is, we get passed over by the outside world _because_ we've moved past the kind of things that are easy/possible to make look good to the untrained eye when they're coded.
we're so busy trying to impress each other, we make things that probably don't interest the rest of the world.
Quote:
you say you are a broken record and claims to know what the mainstream wants.
I'm a broken record, but only if you have seen this talk http://www.youtube.com/watch?v=WEEhODweXYg
Smash: I applaud you for doing something other than abstract generated shapes with flat shading, but I can't help but feel the demoscene is still primarily exactly that. The recent explosion of 4k prods is almost entirely flat shaded generated shapes and particles.
yet the most famous 4k is elevated, the most famous demos are debris, lifeforce, stargazer.. they're very much not of that asethetic.
trace: there's one part of the talk where you mention that the problem with demos is that they look old compared to modern CG, and yet e.g. Lights is treated as "fresh" and "exciting". don't you think that's a bit paradox?
Quote:
Trace is entirely correct when he says that the approach is completely different between demsoceners and media producers. What I suggested was to actually get a few good coders together and see what would happen if they approached the new media experience with their demoscene skill. It's still true that demosceners aren't "better" at making things than people who do it for a living.
Ok, agreed.
Quote:
a WebGL-version of Cortex wouldn't impress anyone within the media landscape
Sorry, I meant an interactive-WebGL-version of Cortex ;).
Demosceners do make things that interest the rest of the world concept/style-wise.
1/ the Ellie Goulding thing is totally demo-ish. It's roughly an average Traction demo with a bit of "interactivity" (@usernames put in the demo + you chose where the light goes).
2/ Another example, the other way: you are lucy
Don't you think the exact same concept and execution, web-enabled (webGL, with pictures-imported-from-your-Facebook-account) would have his 15mn of (web-)fame?
The fact is the guys rather did it on the motherfucking Amiga and rocked the Kindergarden party :D. And it's probably better, that's also why the demoscene is cool.
I still think demoscene would gain in experimenting more around the web platform, both technology-wise (webGL) and concept-wise (interactivity).
Quote:
Sorry, I meant an interactive-WebGL-version of Cortex ;).
considering how balls-tight the demo is directed, don't you feel that "interactivity" could only be shoehorned in into it, to somewhere where it's either irrelevant or utterly out of place?
Quote:
trace: there's one part of the talk where you mention that the problem with demos is that they look old compared to modern CG, and yet e.g. Lights is treated as "fresh" and "exciting". don't you think that's a bit paradox?
I bet people wouldn't care about "Lights" if it wasn't interactive and didn't pull data from Twitter.
...or if it wasn't for the music by an umg artist and the overwrought making-of.
The thing with "Lights" is that if it was made by a random dude with music by a random non-signed artist (without vocals or the radio-friendly song-construction), nobody would care. IMHO.
In other words, "Rome" (interactive) was used for pitching the "Lights" project instead of "No comply" (non-interactive). The label wouldn't have wanted something like "No comply". And I'm aware of another webgl based musicvideo in production that some weeks ago EMI wanted to cancel until they saw "Lights".
Quote:
trace: there's one part of the talk where you mention that the problem with demos is that they look old compared to modern CG, and yet e.g. Lights is treated as "fresh" and "exciting". don't you think that's a bit paradox?
i guess its what i was saying: demos trying to be like renders and failing, vs creative coding trying to do simple cg and being sufficiently different to renders etc to not be comparable and instead be considered "different".
see here's my beef with interactivity (and i might be approaching this from an unusual angle but bear with me):
i've always enjoyed video games, especially the ones with a narrative, and over the years they adopted more cinematic approaches to certain situations, like, say boss fights. i've been playing crysis recently and the ending sequence of boss fights is really spectacular, but with the added problem that if i don't interact properly (because i dont aim well, i dont take cover well, etc), i get to watch and listen the same (fairly scripted) sequence over and over. what i mean is that if i hit all the notes correctly in such a sequence (i.e. play the game well), it gets an engaging exciting experience; if i die a lot, then the whole script becomes tedious and repetitive, and especially frustrating when i have to listen to "QUICK, SHOOT IT BEFORE [etc.]" over and over for the seventh time. in a way, the game's artistic content gets diluted by the fact that i, the player, become the director, cinematographer, actor, stunt coordinator, etc., and if i dont perform well, the ultimate experience suffers. on the flipside there's the ending sequence of crysis warhead where the final fight against the primary antagonist is a really well done cutscene, which i had to WATCH instead of playing it.
i see the same problem with interactivity in demos as well. my example of RO.ME might have been a polarized one but it illustrates the far end of the problem rather well: either you limit interactivity so that the user doesn't miss much (in which case you might as well just not have any) or you allow a huge degree of freedom. in the latter case, you will need ten times the content to make sure the user doesn't find any blank spots (like me looking at a wall all the time) or encounter the same music/sounds/objects over and over again, and that still makes no guarantee that the user gets an optimal experience.
funnily enough, i just noticed a critical point: i keep writing "user", but really, what i want is a "viewer". "audience".
i've always enjoyed video games, especially the ones with a narrative, and over the years they adopted more cinematic approaches to certain situations, like, say boss fights. i've been playing crysis recently and the ending sequence of boss fights is really spectacular, but with the added problem that if i don't interact properly (because i dont aim well, i dont take cover well, etc), i get to watch and listen the same (fairly scripted) sequence over and over. what i mean is that if i hit all the notes correctly in such a sequence (i.e. play the game well), it gets an engaging exciting experience; if i die a lot, then the whole script becomes tedious and repetitive, and especially frustrating when i have to listen to "QUICK, SHOOT IT BEFORE [etc.]" over and over for the seventh time. in a way, the game's artistic content gets diluted by the fact that i, the player, become the director, cinematographer, actor, stunt coordinator, etc., and if i dont perform well, the ultimate experience suffers. on the flipside there's the ending sequence of crysis warhead where the final fight against the primary antagonist is a really well done cutscene, which i had to WATCH instead of playing it.
i see the same problem with interactivity in demos as well. my example of RO.ME might have been a polarized one but it illustrates the far end of the problem rather well: either you limit interactivity so that the user doesn't miss much (in which case you might as well just not have any) or you allow a huge degree of freedom. in the latter case, you will need ten times the content to make sure the user doesn't find any blank spots (like me looking at a wall all the time) or encounter the same music/sounds/objects over and over again, and that still makes no guarantee that the user gets an optimal experience.
funnily enough, i just noticed a critical point: i keep writing "user", but really, what i want is a "viewer". "audience".
Mr.doob: no, i havent seen your talk. i will only watch it if after you tell me what subject(s) you're talking about i find is interesting.
rudi: I don't think it's worth your time. Move on.
gargaj: Yep, you have a point. Usually is easier to please the viewer with a directed experience. But interactivity offers some benefits that directed experiences can't compete with.
The challenge right now is exploring these and find something that uses them well. And it doesn't even have to be interactivity, it can just be randomness, which is a resource that hasn't been used too often in demos either (because of the directed experience focus).
The challenge right now is exploring these and find something that uses them well. And it doesn't even have to be interactivity, it can just be randomness, which is a resource that hasn't been used too often in demos either (because of the directed experience focus).
In terms of interactive features online, the only one I've seen so far that managed to bring something interesting into the medium was the Intel "Museum of Me". I believe that sort of interactivity, which looks at you and your datasources, instead of letting you actually control anything mid playback, is a better approach.
trace: frankly, my favorite "interactive experience" might be this. as soon as a narrative/beginning/middle/end is involved, the user gets in the way.
bringing interactivity to a piece of code surely has its own rules, ups and downs.
but if its done clever, i dont think the viewer/user/whatever misses out much experience, this is a problem which relates to a very specific situation (allowing the user to choose his point of view/interest on a scripted/hard coded animation).
there are plenty of ways to avoid this, they are just barely used because the producers of said games/productions seem to value the (illusionary) freedom of their protagonist more than the story they want to bring across.
but if its done clever, i dont think the viewer/user/whatever misses out much experience, this is a problem which relates to a very specific situation (allowing the user to choose his point of view/interest on a scripted/hard coded animation).
there are plenty of ways to avoid this, they are just barely used because the producers of said games/productions seem to value the (illusionary) freedom of their protagonist more than the story they want to bring across.
