Resident Evil: Behind the Mask [+Hazardous Battle]

[-Rusty-]

Coming from a great texture artist such as you it's quite the honor.

lol no problem.

It's a modified version of the GS original, yup. I replaced parts of the uv map and his pants are completely new.

How'd you manage to get it? (if I can even ask that) Your UV map or textures need to be adjusted more though and fix the smoothing.

Also, now that I saw that video of the mansion I'm extremely interested in what else you've got planned for the future.

[Gemini]

How'd you manage to get it? (if I can even ask that)

The format is kinda documented in Sony's references. I wrote a wrapper for it and pushed some manual conversion on bits, then exported everything to a format Cinema 4D could load natively and that doesn't work like ass (i.e. collada). I managed to do the same for Vincent's model in the rendered scene, but it currently looks bad. Hypnos is also on his way and I hopefully with a complete animation set.

Your UV map or textures need to be adjusted more though and fix the smoothing.

You're right, I didn't even realize the horror on this pants and the misaligned belt. <.< I guess I'm a little too used to the low details of the PSX resolution. I would rather redo that mess of a texture+uv map entirely, it looks extremely horrid on his jacket.

[-Rusty-]

Well I can't wait to see what else you've got coming to us in the future.

Also you got a pic of Vincent? I want to see how bad the model is.

[Gemini]

Here it goes:

It's the same model used in the very first renders we had. He's already defined in that pose, rather than applying an alteration of the bind matrices with animation data. The original Gun Survivor is full of junk like this, apparently. Even Hypnos in the tube is done the same way.

[Kegluneq]

The format is kinda documented in Sony's references. I wrote a wrapper for it and pushed some manual conversion on bits, then exported everything to a format Cinema 4D could load natively and that doesn't work like ass (i.e. collada). I managed to do the same for Vincent's model in the rendered scene, but it currently looks bad. Hypnos is also on his way and I hopefully with a complete animation set.

In the rendered scene where Vincent confronts Ark it doesn't look that bad. Is that what you're using? The first cutscene, with the silly pose, looks a lot worse.

[-Rusty-]

Here it goes:

It's the same model used in the very first renders we had. He's already defined in that pose, rather than applying an alteration of the bind matrices with animation data. The original Gun Survivor is full of junk like this, apparently. Even Hypnos in the tube is done the same way.

Lazy developers, but I guess it's not that uncommon in other games back in the day to have them like that. I wish they would have at least reskinned the RE2 zombies though.

The model doesn't look that bad though. I was expecting missing faces or some corrupt uv mapping.

[MeganGrass]

*picture*

Ah, I remember that one.

It was in RSD format on the PC version disc, iirc.

[Gemini]

In the rendered scene where Vincent confronts Ark it doesn't look that bad. Is that what you're using? The first cutscene, with the silly pose, looks a lot worse.

It's a different model in that case. Gun Survivor has a bad habit at copying stuff all over and over. Each cutscene has a separate model of every actor only to provide different animation data, probably because of how HMD works and packs data together. Not that you can't combine them for extra sets, but go tell TOSE how to do things properly - they'd look down and wonder about strawberry pie.

I wish they would have at least reskinned the RE2 zombies though.

There is one reskinned zombie, but it's not even used and it looks like poop.

The model doesn't look that bad though. I was expecting missing faces or some corrupt uv mapping.

The TMD version of that model should have an issue like the one you described, IIRC. Probably some conversion hiccup, go figure.

[Guest]

Those lousy developers, if they had no intention of making a quality game, why bother making in it in the first place ?

[Gemini]

Because "TOSE", and easy profit. TOSE has been known for being one of the most horrid developers of the last 25 years. Not only they get extremely low budgets to work on a project, but they also had some of the laziest developers ever. They did get a little better with the DS (think of Super Princess Peach or Chrono Trigger DS, both developed by them), but when they developed Gun Survivor they were still programming in cave speak.

[Guest]

As per what I said on twitter just yesterday, here's a little demo with Ark-ley man:

This is some excellent work Gemini. I've always wondered, how do you manage to implement all the flags in this kind of development? I spent a lot of time messing around with the BIO1 debug beta and became pretty familiar with the flag menu, being able to see what was going on behind the scenes as each action was taken and how the game was keeping track of where the player was at. But the one thing I never knew was how it was actually put together. Does it involve a large number of locations in memory and some incredibly complicated script? It's simple enough to populate with some items and throw in a few enemies, but when it comes to actually coding the scenario and making sure that everything works as it should, it seems there's a lot more to it than you'd first expect. Have you got a plan on how this is going to be done?

[Gemini]

The plan is: build a database of enumerated entries, give them a scope field, let the database parser generate the rest of the stuff for me to be used for all the scripting data. That's exactly what I'm working on at the moment and I'm slightly tempted to do this in Excel for simplicity's sake, unless I get bored of Microsoft's output and create a flag manager some other way more suitable to my needs.

I'll probably start working on this as soon as I'm done with setting up the trigger environment. I've just finished right now writing the generic check for trigger boxes and 4 point areas.

[Guest]

The plan is: build a database of enumerated entries, give them a scope field, let the database parser generate the rest of the stuff for me to be used for all the scripting data. That's exactly what I'm working on at the moment and I'm slightly tempted to do this in Excel for simplicity's sake, unless I get bored of Microsoft's output and create a flag manager some other way more suitable to my needs.

I'll probably start working on this as soon as I'm done with setting up the trigger environment. I've just finished right now writing the generic check for trigger boxes and 4 point areas.

Sounds complicated! So is it like a matter of coding all the status functions individually and then using a piece of software to generate the tables in memory for you? Is this where you have to decide what are enemies, what are keys, how many scenario events you have, etc? Is there a limit to this or can you modify this as you go along? I only have experience with BIO1's flag menu so some of this might not be relevant to you at all, it just seems like there's a lot of work involved in regards to making sure everything works in game and judging by BIO1's menu there's a ton of stuff to consider.

[Gemini]

Enumerations have the capability of growing as much as you need and they don't require you to keep track of the current count or amount. With a database I can integrate them into just anything, being it a compiled script, C/C++ source code, or some kind of structure that may require flags. As long as the DB is compiled and the required output created, there's really no maintaining to be done, other than filling in entries of course.

[Guest]

I get you. I'll look forward to seeing the results!