From looking at the trailer, some of the scenes kind of seem 'worse' from an art perspective. Not sure how to put it, kind of like when a cartoon includes some live-action art / 3D model or just cheap CGI in general, just feels 'off'. Still, great job by the devs.
There's not enough information in the original maps to truly exploit all the features of a raytracing and global illumination engine. None of the textures have normal maps for example, so all of those beautiful bricks and such look flat as a board. Lighting was only defined ambiently at the entire sector level (think shapes on the map) so there's no information on light source locations, direction, spread, etc. So you're really just left with some fancy dynamic lighting and shadows--which IMHO still looks pretty cool (love the shadows and green glow as a BFG ball whizzes by) but yeah it doesn't look anything like a game designed with these features in mind from the start.
Yeah, a bit like sticking the newest, biggest, fanciest engine you can find in the 30 year old car in your garage without any doing further modifications as part of it. Sure, the engine itself is impressive... but the car as a whole is now so unbalanced and hacked together it is worse off than before you messed with it.
Well, I certainly applaud anyone wanting to raytrace doom, but take it from this old code monkey, I've spent my entire adult life in dm-1, and a program like this one can do more harm than good.
If you only develop one part of your renderer (and that's all a single exercise like raytracing is going to do for you), you're setting yourself up for artefacts down the road. I've seen it a hundred times.
It's like putting a powerful engine in a stock Toyota Tercel. What will you accomplish? You'll blow out the drive train, the clutch, the transmission, etc., because those factory parts aren't designed to handle the power of an engine much more powerful than the factory installed engine.
Raytracing basically only renders the shadows and to some extent, transparency. What you really want to do is redevelop your entire graphical pipeline, all the major visual assets (models, textures, uv maps, particles, rigging and animation) at the same time, over the course of a release. And don't forget your level design!
I'm proud of you guys wanting to do this. Three cheers! Falling in love with raytracing, rendering, etc., is one of the greatest things you can do for yourself. And you WILL fall in love with it if you can just force yourself to stick with it a year or two and experience the amazing progress you'll make.
But do it right, okay?
My advice, find a good project, with qualified developers who will assign issues to you (especially in the beginning, until you get the hang of it yourself) and guide you in your quest for sick graphics. Thirty to 45 hours a week, eleven months a year, is all you'll ever need to do (I refuse to believe anyone is so busy that he or she cannot make time for that, especially considering how important it is).
And don't worry about being embarrassed or not having dependencies installed the first time you clone the repo. You have to start somewhere and almost every one of us were there ourselves at one time. So no one will say anything to you and very, very slowly after lots of trial and error you will finally just look it up on youtube.
I prefer not to tell people what they can and can't do with their spare time. This hack was probably a blast to write. Let people make fun things.
Ironically, your advice is essentially "make an indie game", but it's way easier to get noticed by writing impressive hacks that excite people than by developing yet another indie title that's overwhelmingly likely to just get buried on Steam. The vast majority of indie titles will never attract articles on Ars Technica. So this is poor advice even if your goal is to optimize for maximum audience.
This is a very off take and not what I was talking about at all. It can be _fun_ to do ridiculous things without purpose (and a great stress reliever) which is reason enough for a project to exist but the article paints it as being a _visual upgrade_ and an improvement to the game. It's not, for the reasons stated, but that has absolutely nothing to do with the project or development needing to be done in a way that it would be.
Separate the purpose of the project from its presentation in the article to reapply these thoughts towards the presentation in the article and I think we'd be on the same page though.
