neozeed

Members
  • Content count

    11
  • Joined

  • Last visited

About neozeed

  • Rank
    Warming Up
  1. I finally got a VPN to work in China with YouTube, so if anyone wants to see it, here is dosdoom 0.2 on DOSBox. Chainsaw glitches and all. I'll add the changes to my horrible linuxdoom thing I keep dragging around over at https://sourceforge.net/p/crossdjgppv1/DooM/ci/master/tree/ Although Allegro doesn't like a straight GO32 environment, or GCC 1.27 so I don't see audio without a lot of work...
  2. I got this to compile somewhat easily with djgpp 3.45 from Windows 10 to MS-DOS. The allegro support at this point is only for sound effects, and it doesn't seem to query what the card is capable of, or how it initializes.. so it sounds distorted. In i_sound.c I made the following changes to raw2SAMPLE SAMPLE *raw2SAMPLE(unsigned char *rawdata, int len) { SAMPLE *spl; spl=malloc(sizeof(SAMPLE)); spl->bits = 8; spl->freq = 5512; //11025; spl->len = len/2; This uses a lower frequency and sets the samples to half the length. Now at least they sound normal(ish) and don't repeat. Thanks for digging up this old relic it was fun to play with!
  3. I killed the page as I was unclear about the level source John had released. I just never realized how well the freedoom artwork fits the iD levels until I had mapped them. I redid the graphics one last time with no dithering against the imagemagik palette and it looks a lot more on target with just loading the iD levels as an iwad. I'll update the videos later as VPN access to YouTube has been hell in China this weekend. I just thought it was interesting that even using the"wrong" it horrible quality settings still gives something that is immediately recognizable, and that by showing that assets from Freedoom map so nicely on vanilla (even version 1.1) that people's complaints about Freedoom are largely in their heads, as it's obviously different looking, but when on familiar levels things are imediatly recognizable, and very playable. I'll clarify that I was altering sound, music, palette, and the graphics so that it's clear. What I also want to try is to take a bunch of pictures of flooring and wall tiles, and use high resolution, high color depth, and have imagemagik do all the hard work to show other people that they too can try to do stuff in the doom engines. I doubt any of it is of any use to Freedoom, but I just thought it was super cool.
  4. I tried something else crazy, using Imagemagick I stitched together all the art into one big file, then let it dither it down to 256 colors, used that as a map, and re-transformed the entire art set. And it looks way better now. It's an incredible amount of work to go through, but I guess the only ones who would care either want to keep higher resolution/depth assets or want to avoid copying the magical palette from iD, and don't want to spend all that much time on it. I just thought it was an interesting exercise.
  5. I'd just liked the idea of building raw levels as I've never done it before. Just like even re-processing all the graphics, audio and music. If anything, it let's me source assets from a much higher resolution and depth and vanilaerizing them. I'd never actually built an iwad before, and it's been an interesting experiment. Sure things can be effortless, but I almost like that this is crazy top heavy. And it was cool Romero released the level source but I'd never done anything with them. Also I should be able to take other level sets and make them standalone iwads, which would be a lot more cooler than just patch wads. I was more so drawn to Romero's dump as it's reconizable data, and it needed an additional step to make into levels. I'm still impressed as an outsider I was able to crawl through the build process and get something. And a chance to mess with imagemagik. I know that you guys have been burned with it but it's still an incredible tool. Just sucks that it's not bit by bit reproducible, which I can see as a pain, but for one off fun it's just that. It takes my laptop, an i7 about 20 minutes to re-process all the assets, which is still amazing to me. It'd take a lifetime in the 90s. Although for my messing around with 1.1 I have to downsample the music and audio for sure as 1.1 doesn't know what a midi is. Or audio at 22Khz... But at least, programmatically I can make a few changes and not downsample anything and try vanilla 1.9 .... I mostly like 1.1 because it runs under OS/2, and it has that left/right network view thing.
  6. Interesting, I always got crazy rainbowing when I was using the default palette stuff so I went through learning how to re-adabt the graphics. It was still a fun experience. I'll have to try it again, as needless to say it takes a while to go through all those images. Although I was thinking about having imagemagik stitch all the images together as on 24 or 32bit deep PNG image, and letting it decide what 255 colors best work for everything. It'd be an interesting experiment. I'm still I'm pressed I got anything to actually work on 1.1 thought. And with Romero's original level dump it just shows how immediatly recognizable it is. When people cry about freedoom, I think seeing it over familiar levels just shows how great the assets really are.
  7. It's the DooM 1.1 engine, it only used 11khz 8bit audio. And yes the images have been downsampled. When I tried to use the freedoom graphics I got rainbow artifacting, so that is why I ran them through a dither / remap through this 'vga palette' . These are doom 1 levels from the 2015 Romero dump. Oh yeah it is E2M2. That is what I get for doing this @3AM. Why? Because I wanted to see if I could get something 'recognizable' but different using the 1.1 engine. And learn the freedoom build system, and tweek stuff here and there for my own twisted enjoyment. After Romero had released all the dwd files, I thought it'd be interesting to try to setup a build system to build them all and run them.
  8. I just found out about the deutex update to handle PNG's. So I dumped all the old stuff I could and re-downsampled everything just as I had done before. The only thing that hit me was the text generation, it needs python modules I don't have and pip doesn't work. So I filled in the text blank with some older versions. I wonder if it's just easier to have imagemagik just use a TTF, and just 'print' all the pictures to a PNG? I'm still using the midi3mus for ultimate vanilla compatibility. I know the 1.666 onwards can play midi, but I wanted to go all the way vanilla. I also had to chop up the status bar, and overlay the 1.1 bar so it'd match the same dimensions. I think it looks good. And here we go again, on the registered v1.1 executable, this time using general midi I think it still looks great! I need to look at the sprites to see if there really is some fringe pixels, or just an error in my methodology. Anyway, I know people always complain about freedoom, and by using the assets over Romero's 2015 level dump it really shows that freedom has really captured the DooM environment!
  9. And FYI it works surprisingly well. I downsampled everything I could, and it's even playable on the registered DooM v1.1 executable. I shoved it on sourceforge as zeeDoom. Sorry if this is all off topic, herrasy or just plain evil. I made a lame run through e1m1, although I should have upped the difficulty as there just wasn't enough action in this video. I'll maybe make a new one later on. Again apologies to all. And Happy New Years!
  10. Not that I'm going to cry, but yeah it's the 1.2 BINARY..... If you want to re-live the NeXT excitement, there is Previous, a pretty amazing NeXT emulator! It can do the 030 and 040 cubes, some of the NeXTStation hardware etc. It's pretty awesome.
  11. I guess it's too late to ask for something like the NeXT source for DOOM... I did a half assed quake port a long while back, but I've always wanted to see how DOOM did it's magic on NeXT, as my Quake stuff 'runs' but you really need a P3 to get some performance out of it.