The LayerOne conference is over, and that means this last weekend saw one of the biggest demoscene parties in the USA. Who won? A European team. We should have seen this coming.
There were two categories for the LayerOne demo compo, the first using only the LayerOne Demoscene Board. It’s a board with a PIC24F microcontroller, VGA out, and a 1/8″ mono audio out. That’s it; everything that comes out of this board is hand coded on the PIC. A few months ago, [JKing] wrote a demo to demonstrate what this demoboard can do. According to him, it’s the only reason Hackaday sold a single Demoboard in the Hackaday store:
First place for the Demoscene Board competition went to a remote entry – a team called COINE. The video and initial reactions of everyone in the room:
No one in the idea had any idea how this was possible. The hardware should not be able to do that. The resolution and number of colors are too high. It was, by far, the most impressive demo at LayerOne. That doesn’t mean the other submissions to the Demoscene board competition were overlooked. [jamisnemo]’s entry was well received, even though he ran out of time writing it:
The second category for the LayerOne demo competition was the ‘Secret’ Board. There were only 10 or 12 of these boards ever made , but there were still some impressive entries. The board itself is built around an ATMega88 – 8k of Flash, 1K of RAM, and 512 Bytes of EEPROM. If using an ATMega88 as a demo platform sounds familiar, you’d be right. [lft] built the Craft demo way back in 2008 around this chip. The Secret Board is designed to run this demo, and serve as a platform for a demo that implemented a framebuffer on the ‘Mega88:
In all, an excellent competition. It was well received by all attendees, and next year’s compo is sure to be even bigger. If anyone has any idea on how the big European capture these demos to video, please leave a note in the comments. No one at LayerOne could figure it out.
That COINE demo brings the demoscene coding back to it’s roots at 90’s: best demos where ones that contained something totally new and “impossible” to do. Of course all those tricks were “cheating” and used all kind of precalculation ever imagined :) I’m sure COINE’s method is really clever and really obvious when you’re then told how it’s done. Until then.. I cannot come up with any idea how to do that. And I’ve been writing assembler for demos on 80*86..
It always blew my mind, that in the 1990’s, they could produce not just the crazy effects, like hundreds of rotating cubes turning themselves inside out while music played and shiny, 3-D text danced across the screen, but the fact that they could fit it all on a single, 1mb floppy disk, which also had to fit a boot loader. What happened to all those talented hackers? Probably writing crappy accounting software for the man now.
More like founding video game studios and benchmark companies. Idiot.
This demo looks “just” like whirling mist to me.
Could you explain (in layman’s terms) what is so special about it?
The effect is called “shaded bob” which is usually really simple effect if you have proper video memory. It’s basically modifying the pixel values (darkening / lightening and changing color). However in this case you don’t have the video memory needed for making this effect “traditional” ways.
The board in question has 96K RAM. Which is unusual for a PIC-type chip. You’d almost HAVE to have a framebuffer for that. The board’s spec on the HAD store mentions “graphic accelerators”, have to look up what that is.
> Who won? A European team. We should have seen this coming.
Do the math: assuming that the probability that a child born becomes a demo nerd is uniform over the whole population of industrialized nations, and the chance to win that contest is uniform among these, then the ratio of the total population sizes between the US and EU is indicative of the chance that a EU team rather than an US team wins. The EU has over half a billion citizens.
Overall, why do you sound surprised at all? If this is supposed to be a fair international contest, you’d expect the winner to come from anywhere.
“the probability that a child born becomes a demo nerd is uniform over the whole population of industrialized nations”
Which it isn’t.
“the chance to win that contest is uniform among these”
Which it isn’t.
There are far fewer active sceners in the US than in the EU, or other ‘industrialized nations’ such as Japan.
As a result, the few US sceners that are there, operate in a ‘vacuum’ more or less, and it is much harder for them to reach the levels of quality that EU sceners have.
It’s like baseball: the ‘world series’ is pretty much just the US, because nowhere else in the world do people play baseball at the same professional level as in the US.
The talent may be born anywhere over the world, but the facilities aren’t there.
If I’m not mistaken, that was sarcasm.
To me, it looks like a bunch of parametric equations being graphed: in the same scene, take one equation and modify it with small factors to displace a few extra lines with different colors being rendered.
Entries are nice, though all but the first one are really too short.
In general the fact that the banner of demoscene is used for commercial promotion of a certain product feels forced to me. Certainly my understanding of the scene radically differs from whoever organized this.
The first place demo and the “reactions” in the room makes it clear that nobody in the room or near this project has any relation with the demoscene. Nobody heard of shaderbobs, it’s more than lame. We coded back this effect on amiga (7 mhz) when we were 10-12 years old. It’s just a PR marketing. And the “video” recording quality??? Huh.
Yeah, I think if this were a marketing stunt you could be on better video.
This is a computer security conference called LayerOne. It’s not mainstream demoscene for the elite competition. I really liked this challenge and might give it a whirl for next year. The hardware is meant to provide a standard platform that limits the system resources available.
You can also tell it’s not marketing because you can’t buy these anymore. Although [Arko] was telling me he has an idea for a new design so maybe there will be a different hardware set for next year.
The plan is to have a version 2 of the board. Possibly just adding more SRAM to allow for a larger frame buffer (more color, res, etc) and maybe even changing the mcu to an stm32.. but nothing has been officially decided yet. We also welcome folks to bring in any hardware they have (amiga, c64, emulators, toaster, you name it). If enough folks submit the same hardware, we’ll make a category for it. As for this PR marketing conspiracy theory.. I can’t help but feel honored that someone would even think that. We barely broke even and used the profits from sales to buy crap tons of ice cream and root beer which we gave away for free at the demo party :)
To address the other points: We will have better video recording next year. The audience maybe new to the demoscene, but they are having fun. I will always encourage everyone to participate. Make demos, show off your skills, help each other out, form groups, wow the audience, and have fun.
I’ll admit it, I’ve never been to a demo party before. Seeing how I wouldn’t be flying out to one anytime soon (hope to in the near future), I figured I’d just start something similar locally.. even if it doesn’t follow the traditional guidelines of a demo party.
Kudos for the effort anyway and best luck to your conspiracy ;) I don’t think there’s a set of rules defining what a demo party should be, though there is always a lot of fighting about what a demo is. And it’s not uncommon for the entries from over the pond to look different, which is not a bad thing at all. Who knows, maybe you’re about to start some new wave of demo making.
Go for the STM32, it’s the best
and anyways, we need to finish our trilogy someday, would be nice to have other people compete on similar hardware
i was in the room (well technically just outside), and had involvement, been in the european demoscene since pre amiga days, now more people know about it, it’s all good. most demo scene people i know went on to use those skills in a very marketable way. and yes we know what shaderbobs are.
Capturing the demos nicely doesn’t seem like a big problem…
http://www.amazon.com/IOGEAR-2-Port-Splitter-Booster-GVS72/dp/B002TR6KZ6 <– 2 port vga splitter, split the audio as well, wire to capture card : http://www.amazon.com/VGA2USB-Video-Grabber-Capture-Device/dp/B000UO94DU
As for the shaded blob effect, I suspect the answer lies in the relatively small number of blobs. These function as 'pixels', and the actual video pixel effect is done programmatically rather than using rasters. use a lookup table to find the blobs that overlay a given video pixel, then mash their opacity/color values together. The color lookup table would likely be ignored, and the graphics cores pipelined or run in sequence (two rendering a frame while the third actually displays it)
+1, I came here to post this. A little pricey, but when you look at the competition it starts to look more and more reasonable, and supposed to be compatible with v4l as well. I use them at work for screencaps off equipment displays, quality is fairly good in my experience.
(I have had it lock up win7 somehow, but that’s not surprising with the amount of iffy custom control software and USB-to-RSxxx dongles I have hanging off that test machine.)
I looked at that, but I’m not convinced it will do the weird VGA resolutions this board puts out.
I think the current plan for next year is to build an FPGA-based ‘video capture of everything’ board that outputs 1080p @30fps over HDMI. That can be easily captured to a computer with my gear. Then we need to do the livestream/audience reaction.
If anyone knows what the euro parties use for capture, send me an email. Their capture equipment alone must be an incredible build.
For that price, and given amazon’s return policy, it’s worth a try. Given that the TV was able to play these “weird resolutions” I’ll bet they can be managed at least as well in the capture widget. I mean, I like to build stuff, but I’m not about to reinvent the wheel when something this cheap hasn’t even been tried ;)
I got ten made up by dirtypcbs. Working well! If anyone wants a pcb in the UK let me know :)
Awesome!
What is so special about the shade bobs? The PIC24FJ256 seems to have a 256 entry color look up table and 16 bit RGB output. Really, the challenge here is none… this is what everybody did in mode13 back on dads PC in the early 90ies.
My two cents: Why was the ‘secret’ board so secret, and now that its been ‘demoed’ is there any plans to release its schematics/software
The schematics for the atmega88 board are found on lft’s site http://www.linusakesson.net/scene/craft/
It was pretty easy to guess from the description on the layerone site that would be the platform, so not really that secret. lft’s uC demo work is fairly well known I think.
Ugh… everyone… They’re called ‘shadebobs’, not ‘shaderbobs’. As said above, this effect was done way back in the Amiga days, and isn’t all that hard to do. No, it doesn’t have anything to do with ‘shaders’. Just a simple form of additive blending.
shaderbobs are the new hotness ;)
http://www.pouet.net/prod.php?which=981 was always my favourite amiga demo too..