Operation - Arctic Wolf
ETERNAL DOOM III
Wolf 3D - The Original Missions
Wolf 3D - The Nocturnal Missions
Wolf 3D - Spear of Destiny
PHOBOS - ANOMALY REBORN
Star Wars - Darkest Hour and Dawn
Doom Raider - Crypt of the Vile (full version) "Warning: this is bloated with unused textures"
The gzDoom engine supports Hi-Res display but the PWAD Textures just do not have enough information so the onscreen image is pixelated and blurry with washed out colors.
The downloads for Doom, Doom II, and Heretic are what you would expect for High Resolution graphics and Models, a modern visual with clean lines and detail.
For "Operation - Arctic Wolf" through "Doom Raider - Crypt of the Vile (full version)" the textures, flats, graphics, and sprites are not Modern High Resolution. They are not pixelated and blurry. They are Old-School PNG Images with rich color and atmosphere. If you see one of your favorite - try it out.
True. I was experimenting with these PWADs to see if I could see something clesre for gameplay. They look more colorful to my eye. I uploaded them because I had the web space and others may be cutious.
The Models for Doom and Heretic are from the pack "ds-models-r14.tar.gz" which seems to have disappeared from the web. The files for Doom, Doom II, and Heretic are really the only ones worth the download if one what High Resolution.
The other Games are just fun to play and are easier on the eyes than blurry pixelation.
Didn't you create "ds-models-r14.tar.gz"? On the Skulltag Forum
Awful filtered textures are awful. This is not "hi-res" and it's certainly not "enhanced" -- from all the smudging and other processing you've done they're actually quite "blurrier" and less detailed than you say the actual graphics are, and certainly far harder on the eyes.
Just_Jim said: Any constructive advice to eliminate pixelation and blur apart from drawing all the graphics from scratch?
Yes: don't bother. If you do a search for "high res Doom sprites" or similar, you'll find at least 3-4 projects in the past 5-6 years that aimed to do just that. Some were doomed (pardon the pun) from the start, as all their authors did was just apply a resize filter to the sprites, which is nothing that any OpenGL port can't do, and WITHOUT a memory penalty, so that kind of "enhancement" is out right off the bat.
A few others tried to redraw a few sprites, and that's where they got left off: a few. I don't think there's someone with enough talent/speed to redraw a few hundreds of animation frames in a reasonable amount of time, not even if it's a full-time job for an entire year.
Not only that, but in practice, there's no difference between a high-resolution model and the best hand-drawn sprite we've ever seen. Take a look at the very latest such project, here (it's a hand redraw) and tell me honestly: is it worth anybody's time?
To answer your question, you can tinker with Doom Ascension high quality model, which easily beat any sprite-based approach for several reasons. Oh and of course any "solution" is going to be highly port-specific.
I had planned, eventually ... , to create a program that could take a 3d description of a sprite (with movable appendages), and from that generate the frames needed.
With depth cues (perspective, color, intensity), shadows, skin stretch, ... .
The generated format would be selectable, but could include
Got to be easier than week long sessions that at best, generate 8 frames a night.
That's actually an industry-standard procedure for generating sprites from 3D models, e.g. those used in older RTS games such as Warlords Battlecry, Age Of Empires, some TBS like Age of Wonders etc. and I think that many 3D modeling programs even include such export modes or can be batch-programmed to do it.
I have come to admit to myself that I no longer see well enough to draw textures. I have worked on several games in the past and a number of packs have my contributions. So, the comments here serve as a valuable reality check - I knew it was true before I Batch Processed this stuff for Doom PWADs. I just don't see things the way I did in the past.
So, shaders in gzdoom.pk3 can intensify color, may-be add grainyness and modulate sharpening? Is it a matter of loading a pwad - say ETERNAL, then clicking through options until it looks good to my eye onscreen?
Well, I searched fpr information about applying filters to flats and textures(patches) without success. Walls, Floors, and ceilings still look washed out and without detail (from the wads). I will leave what I Uploaded on line and maybe advertise it as "Doom for the Visually Impaired". If you know of any Dim-Sighted-Doomers, let them know its there. I highly recommend ETERNAL. If you know of filters that apply to flats and patches, let me know. Thanks,
No filter can add real detail or reconstruct detail that got lost to downsampling, and certainly not a filter with a real-time applicability constraint, unless you make very restrictive assumptions about the quantity and type of detail present original data. Claiming otherwise is just pseudoscience, right there with free energy and perpetual motion.
Sure, you can interpolate or make various assumptions about the missing information (that's how e.g. spline-based rescalers work) but claiming that you can literally reconstruct what the original artist had in mind based on a decimated set of pixels....nope, that's just delusional.
Just because in Hollywood movies then can infinitely zoom in a bunch of pixel and Magically Reveal Hidden Detail, it doesn't make it real.
That scene, in all of its 80s sci-fi naivete, is actually much more credible than the modern NCIS/CSI crap: if I recall correctly, the device Deckard is using is actually a sort of standalone scanner + TV combo (it was fed an actual printed photo, IIRC).
Anyone familiar with traditional photo prints knows that the equivalent "megapixel" resolution of a standard photo developed from a 35mm negative is in the order of 20-30 MP (a goal that only recently has been reached by digital cameras, and then only by cheating) and it was no unusual to ask the photographers back in the day to "zoom in" e.g. somebody's face and then make a full-size portrait from it.
The most interesting aspect of the scene, IMO, is that the device portrayed would be -at least partially- feasible even with 80s technology (if it was not an actual device, camouflaged for a more futuristic look). The only thing they lacked is, I think, scanner sensors in the Kdpi range, but that could be compensated by using lens. After all, there were "video microscopes" back then. Of course, they hadn't a fancy voice control interface ;-)