Demospotting

Today I spent some time watching demoscene productions in order to relax and I felt like posting some of my thoughts here. I first watched:

- Squish by AND (64k intro, 1st at Assembly 2002). A masterpiece, very impressive. It's a one-man production by AND a.k.a. Dmitry Andreev, a Russian student at the mathematical faculty of Voronezh university who is just a bit more than 1 year older than me. I interviewed him for Hugi #28. Squish is technically sophisticated, it features a lot of not-so-trivial effects, well combined, and what especially impressed me: the colours have been chosen well, it's anything but coders-colors. So Dmitry did all the coding, 3d modelling and also music-composing. I remember people commented on his intro in a positive manner, but some critized the music, which is why he had the music made by someone else in his sequel intro "zoom3". But I also enjoyed the music, it's well done, too. Now the question is: How much effort has been made for such an intro, and what has been the motivation? We can read in the info file that he worked 1 year (August 2001 to August 2002), 3 to 4 hours a day. That makes about 1400 hours of total work-time. Enthusiasm and endurance is required for such a thing. But I guess it's pleasant, you have to do a lot of literature-study in order to find out how the effects have to be implemented, and thus you learn a lot. AND is/was also working as a game developer so he even made money with his knowledge gained in demo development. In other words, it was not just a waste of time. - I read that AND previously made other demos for ZX Spectrum and PC, and that one year of work per production has been usual. Well, it's important to consider this. I think one of the most difficult things about software development in general is to estimate the amount of time that will be needed to invest in order to bring a project to an end successfully. - I decided I wanted to see another top quality production, and so I started:

- Iconoclast by Andromeda Software Development (demo, 1st at Assembly 2005): It's not the first time that I watched this one. But this time, I switched off the music since it is so great that it would distract me from the action on screen, and the action on screen was what I wanted to pay attention to this time. I was very impressed about the effects and the ways they were arranged. It's at such a high level that most amateur coders probably won't ever achieve. I think one may guess from watching the demo that Navis has a degree in computer graphics. - Now after seeing the very high end of demos, I wanted to check out more average productions in order to get an idea of what skills you need to have to create an average demoscene production. The first one I chose was:

- Composition No. 2 by eL (demo at Syndeecate 2001): Well, quite a contrast! This one features such simple code even me could have done it, and that means something. (I mean, I'd have to learn how to set a pixel and how to do 2D rotation, which would be a matter of a couple of minutes, and I'd have to read the minifmod manual in order to find out how to synchronize the thing with the music. The rest would been straightforward.) Not bad though. But obviously a totally different category of demo than Iconoclast. Comparing these two demos would be like ... I guess you've already thought of a suitable analogy. - Then I decided to watch some productions from Sector Omega, as they're going to be featured in Hugi #32 since I watched their releases some months ago and was quite impressed back then:

- Decadentia by Sector Omega and Stravaganza (demo, 1st at ifparty2004): Technically, it's far more sophisticated than the previous demo, but it's far simpler than Iconoclast. I wouldn't be able to code it without learning about 3D and shading. Anyway, it looks somewhat oldskoolish due to simple models and flat shading. In these days when more sophisticated techniques are widely used, this adds a special feeling to it. 

- Mirages by Sector Omega (demo, 2nd at bcnparty 2003): A story demo with a (simple) political message. What's more interesting for me is the technical component, though. There is pixel gfx. I wouldn't be able to draw anything like that. The effects are rather simple though (I've already forgotten the details, but I remember a tunnel appearing to me the most sophisticated thing). I guess it would be a matter of a few days to learn the basics required for making a demo similar to this one.

- The Message by Sector Omega (demo, 3rd at bcnparty 2005): This is the one that gave me the idea to interview Sector Omega. It contains very professional looking graphics. It's 3D and clean. But it stops executing somewhere in the middle on my PC, unfortunately. Anyway, it's not to be compared to Iconoclast, either. But it's a good one.

Hugi 32 is going to contain charts again, including coders charts. I recall writing some article a couple of years ago in which I doubted the voters' competence to rate coders' skills. But I was probably too naive; of course you can judge coders' skills from their productions. If the effects are very sophisticated, this shows the coder responsible for a production is probably far more skilled than the average coder. Of course charts are not objective, and you cannot rate democoders like people who participate in a size-optimizing compo where the outcome can be measured very easily. But you can reward people who've showed they've made an effort.

As soon as I have more time, I'll get my hands on SDL and learn how to implement modern effects in order to get an even better understanding of demos.