Help
Duke4.net
DNF #1
Duke 3D #1
I also remade about 1/2 of Pigsty before I had to move on to other things. I believe there is a link to that here somewhere. I'm glad to hear you liked it.
[RELEASE] Hollywood Holocaust Rethinked
- Polymancer
-
#212 Posted 02 May 2021 - 04:03 AM
I did a walkthrough myself - complete with Art Director commentary!
As to NPCs - in the original game, there was an alien invasion on, and all the strip clubs, bars and even supermarkets, burger joints and banks are all open for business, so direct your complaints to 3D Realms.
Unfortunately, I can't consider this project as a success. The brief (revised after IceColdDuke bailed on us) was "How close can you get to Unreal if you push Polymer with modern modelling methods and slavish attention to detail?" The answer was unfortunately, "not very close at all."
I think that Mark hit the nail on the head with the ideal Polymer project with his 'Graveyard' mod. A small area, few lights and confined spaces work best for Polymer. The minute you try to expand the situation, things fall apart quickly. We couldn't have enough lights to keep the normal maps illuminated. If you did, we ran into performance issues, so half the models look flat and lifeless with no specular highlights or normal map details. I even at one point wondered about reverting to Polymost and baking all the details into the model skins it got so bad. The look would have been something akin to Wrath, Aeon of Ruin - primitive, but cohesive and looking good for what it was aiming for. I still think that there is an argument for approaching any extensions to the HRP in this fashion. It depends on what RAZE comes up with on the render front.
Hank and others are right in that, looking back, we should have used Unreal, or Unity, or some other, more modern engine. EDuke32 is great as a Duke Nukem 3D port, but shoehorning a modern renderer into an old 2.5D engine comes with a lot of compromises. When you factor in that the 'modern renderer' was never optimised and had a ocean liner full of problems even before we started, it's a wonder we got finished at all. A large amount of time was spent trying to pin down and work around problems with Polymer, problems with alpha transparency and issues with normals on MD3 models. Programmers seem to like to bitch about Unreal and FBX models, but shit usually just works, which is priceless. If I were to do it again - I wouldn't. Too many compromises were made on quality and problems with the engine took up too much time. At a certain point we realised that what we wanted was basically unobtainable. The final results show this.
If I were to approach something like this again, I'd get a small team together (more than two people!) and use either Unreal or Unity to make the game. If it was just me, then I'd use Serious Duke 3D and add my own map and models to that as it has all the game setup already established.
As to NPCs - in the original game, there was an alien invasion on, and all the strip clubs, bars and even supermarkets, burger joints and banks are all open for business, so direct your complaints to 3D Realms.
Unfortunately, I can't consider this project as a success. The brief (revised after IceColdDuke bailed on us) was "How close can you get to Unreal if you push Polymer with modern modelling methods and slavish attention to detail?" The answer was unfortunately, "not very close at all."
I think that Mark hit the nail on the head with the ideal Polymer project with his 'Graveyard' mod. A small area, few lights and confined spaces work best for Polymer. The minute you try to expand the situation, things fall apart quickly. We couldn't have enough lights to keep the normal maps illuminated. If you did, we ran into performance issues, so half the models look flat and lifeless with no specular highlights or normal map details. I even at one point wondered about reverting to Polymost and baking all the details into the model skins it got so bad. The look would have been something akin to Wrath, Aeon of Ruin - primitive, but cohesive and looking good for what it was aiming for. I still think that there is an argument for approaching any extensions to the HRP in this fashion. It depends on what RAZE comes up with on the render front.
Hank and others are right in that, looking back, we should have used Unreal, or Unity, or some other, more modern engine. EDuke32 is great as a Duke Nukem 3D port, but shoehorning a modern renderer into an old 2.5D engine comes with a lot of compromises. When you factor in that the 'modern renderer' was never optimised and had a ocean liner full of problems even before we started, it's a wonder we got finished at all. A large amount of time was spent trying to pin down and work around problems with Polymer, problems with alpha transparency and issues with normals on MD3 models. Programmers seem to like to bitch about Unreal and FBX models, but shit usually just works, which is priceless. If I were to do it again - I wouldn't. Too many compromises were made on quality and problems with the engine took up too much time. At a certain point we realised that what we wanted was basically unobtainable. The final results show this.
If I were to approach something like this again, I'd get a small team together (more than two people!) and use either Unreal or Unity to make the game. If it was just me, then I'd use Serious Duke 3D and add my own map and models to that as it has all the game setup already established.
- Let's go Brandon!
-
#214 Posted 05 May 2021 - 01:37 PM
Glad to see you understand what I've been saying for years, Tea Monster.
#215 Posted 05 May 2021 - 01:50 PM
Shame to see the HRP wasted though, given the license not allowing it being used with other engines
