WXX-Rebirth : a modern engine for Wipeout 3 Special Edition !

[aybe]

UPDATE 7/7/2017

Due to numerous reasons (mainly the sheer amount of work), the project in its current form is suspended,

As of today, WXX-Rebirth is becoming OpenWipeout, a library that will provide access to Wipeout 1, 2, 3 SE, 64 (and maybe Fusion) assets, more than 90% of the games assets have been reverse engineered, beside simulation algorithms. For the latter, as times passes the outcome dwindles, because it requires expertise in ASM, IDA, hacking and an indefinite amount of free time (which I don't have) ... but see below.

I have joined the BallisticNG team, in an endeavour to bring all these games to it, much of the WXX-Rebirth project work will be used, improved and enhanced so nothing was lost in the process, beside some deception (that's life !).

From now on, stay tuned on BNG channel comms @ Discord where I am now, or its dedicated thread here, to hear about the advancement.



OLD README

Hi !!! After many hours, days and weeks of coding, I'm proud to announce the very first demo for the project. :P


Here is the README file along a few screenshots:



================================================== ====
WXX-Rebirth demo version 0.0.1 README (5th April 2016)
================================================== ====

Hello and welcome to the WXX-Rebirth project !

WXX-Rebirth goal is to offer a modern engine for Wipeout 3 Special Edition :)


================================================== ==
DISCLAIMER/WARNING/WARRANTY/COPYRIGHT/LICENCE/PIRACY
================================================== ==

WXX-Rebirth DOES NOT contain copyrighted content.

WXX-Rebirth is ONLY an engine that aims to recreate a Wipeout experience, no more.

YOU NEED a LEGITIMATE copy of Wipeout 3 Special Edition, it won't run otherwise.

WXX-Rebirth comes with ABSOLUTELY no warranty.

WXX-Rebirth is NOT affiliated with Wipeout brand, SONY and current copyright holder.

WXX-Rebirth is a FREE project under the terms of a GPL licence (code available soon).


============
REQUIREMENTS
============

- Windows 7 or later
- a graphics card that supports DirectX 9 or later
- a LEGITIMATE copy of Wipeout 3 Special Edition


=====
USAGE
=====

- grab your Wipeout 3 Special Edition CD-ROM
- copy WIPEOUT3 folder from it next to the demo executable.


==========
DEMO NOTES
==========

As this is the very first release, there are still many many things missings.
You can access the 22 tracks and 8 ships that are in the original game.
Ship handling and collisions are still very primitive, as a result you will
probably have a hard time keeping ship on track : press Backspace to reset it.

See the options for changing track/ship and tweaking a few things.

Etc etc etc ... but expect a lot of improvements over the next weeks :)


10198 10199 10200


And if that wasn't clear enough, I'll say it one more time:


WXX-Rebirth does not contain copyrighted content
WXX-Rebirth is only a game engine aiming to provide a Wipeout experience
You must possess a legitimate copy of Wipeout 3 Special Edition, it won't run otherwise




Download link:

https://github.com/aybe/WXX-Rebirth/releases/tag/game


That said, now I'm waiting for:


your feedback
your suggestions
your impressions
etc etc etc ...



There isn't much for now you can do in it, as you'll see, but it's a start !

Expect many improvements over the next weeks :D

[Hybrid Divide]

*blinks* I was always under the impression that WXX-Rebirth was meant to reproduce WO1 and XL/2097?

Very confused... Excited, but confused.

[aybe]

Well ... that was the intent before the 'reboot' :p

Basically,


bigsnake got in touch with me as they managed with Xpand to extract the ships from it, minus some weirdnesses they encountered in the format
I then took the opportunity to take a look at Wipeout 3 files
now we're at the point that practically everything has been deciphered and is well understood, even those mysteriously missing polygons on track edges; skybox and a few (non-essential) others are unknown though
since things were somewhat different in the formats I coded all from scratch, what turned out to be a good thing as I ended up in ~1 month to something much more consistent
there are still a few models that are WO1/WOXL-like in the game EXE and now they're using the new pipeline since I've integrated them


TL;DR : I will probably bake WO1/WOXL (relatively easily) at some point thanks to the new code base, the other way around would have been trickier; and attacking WO3 was a rewarding challenge in its own, too :g

Here's an excerpt of what's coming up in next release: track perks

10201

And hopefully with full scenery, better handling and better physics !

[skythian]

This is awesome!

[aybe]

Thanks !!!

I am going to push an update soon, hopefully by this week-end, with many improvements :D

There are still a few things I'd like to fix, but physics have greatly improved since last version !

I'll also try to bake in the full scenery in the mean time, it'll feel more appealing for testers.

[giscar]

Impressive. I can just imagine how this will look, this has so much potential... Thanks a lot!

I'll go get my WO3 CD and try it right now!

[aybe]

Thanks !

Things have stalled quite a bit since I've been having issues with ship hovering not being strong enough ... However I think I'm holding the right end as I'm getting pretty convincing results such as in the original game : https://www.youtube.com/watch?v=yTJvU4jofmU

Sampa Run track in fact is the most problematic as when going too fast the ship would either hit the floor or the ceiling, currently it's mostly behaving like in the original game but still needs tweaking. So hopefully I will soon be done with that and post a much more realistic update :D (and finally continue in implementing the rest of the game)

[80T]

I was about to blow up in hype when I've arrived to this page and saw the screenshots. It was really Wip3out 3 SE! Not XL/2097, or 1, or the newer ones. Seems like nobody cares much for WO3 overall. There are a few other projects like BNG and another branch which I forgot it's name, that are trying to remake WoXL/2097 and Pure (?) but now finally I've stumbled upon a real WO3 project (and on top of all things it's the SE version no less) Dream comes true! Or not... ? When I've tried to download the Demo version linked here, I ended up with an 404 error :(( I've tried searching over GitHub and even Google to no avail... The project seemed to have vanished. :(( I guess I'm too late to the party ? Did this project got abandoned or got closed/nuked due to copyrights issues or something like that ? If that's so, then it's a sad day... :((

[aybe]

No it wasn't, I'm actively working on it, it's just that I went to a private repository in GitHub hence the 404. I will soon setup some download links that won't vanish over time... It takes an incredible amount of time to get to something 'decent', I improved many things recently but there are still a billion things I have to do.

Here are the builds so far:
https://1drv.ms/f/s!AqAtR29m3cqGsUsZ9Z5LYX0fRxm5

These are pretty old now and do no reflect latest changes, but please be patient for the next one :)

[80T]

This is AWESOME! An absolute relief to find out this is really going to be a thing :D I hope for the best! Just tested v0.0.2. Oh man! The hype is real :)
Is this planned to become something widely public available like BNG that was Greenlit some time ago in the future maybe ? So people could play REAL Wip3Out on PC ?
And better yet, Online MP ? That would be for AG Racing like what Counter-Strike 1.6 is for First Person Shooters :D
Imo, WO3 is the pinnacle of AG racing from all times as far as graphics (for the time), art style and more important, pods handling goes, closely followed by XL/2097 :)
This has just took the 1st place on my hype list which was until now, the nearly released AMD RX 480 :D
I've never had much hope in coming across an actual WO3 project, but I've never "stopped listening to the cues" on the Internet.
Turns out these last 2 years or so were full of surprises when it comes to remakes, remasters, old games ports and projects like this one :) I'm allover this :)) Hurray!

Oh and about:


your feedback - Not much to give feedback on (yet) :D There's not enough "physics" and visual/audio cues implemented to judge on (pitch, roll, yaw, animated control elements, engine sounds, Analog Controller Input Support, AirBrakes not yet implemented, etc... )
your suggestions - For the future: (not sure on development steps... ) Custom Folder/s for custom Music Tracks (ex. CustomTracks01 > Track01.format, Track02.format, etc... ) Custom Folder/s for user created custom pods sets (maybe in Valve's GoldSrc .SMD format... ?)
your impressions - Wip3Out 3 on PC ! ! ! ! ! ! Yey! Finally! After soooo many years :D
etc etc etc ... - Can the physics be "reverse engineered"... ? like from the actual game code "translated" to "PC like" code, so it would be 100% same handling, but running in 60 or variable FPS... ? I think the handling would be the most important part... like 90% important.


Other suggestions: Probably sometimes in the future when the game would be nearly finished.
- Advanced settings for the Analog Directional Inputs (Analog Sticks) such as Dead-Zones X/Y, anti Dead-Zones X/Y, Linearity X/Y, Center Offset X/Y (for poor off-centered Analog Sticks on some Gamepads/Controllers due to poor build or simply wear and tear over time)
- Analog AirBrakes maybe... ? After the physics are 100% in place... ? Since we're now having Analog Triggers on the most commonly used Controllers on PC.
- Maybe even Support for XBOne's Controller Impulse Triggers... ? It could rumble with the intensity that would reflect the "stress" on the pod's control surface elements...

[aybe]

Thanks for the feedback :D

Definitely public one day and open source, but it won't be on Steam obviously for copyright stuff and co.

When I am done with the base of the game definitely I will look for multiplayer mode.

Virtually it could run anywhere since it is Unity and I didn't rely on any particular library, the thing is such project will never end on any store so beside PC it would mean a high jacked hardware... Actually I did start with the first two episodes but the rewrote the code for episode 3 but it also handles the former ones so this also a thing I'd like, to have WO and WOXL.

Now for the rest, I have a much better hovering than what you've seen, looking for roll and pitch and airbrakes. Will be analog by default but will also support digital. For the sound obviously I'll make it happen but the videos too and this was a rather tough thing as I had to write a library for extracting STR streams from XA CDs ...

I deciphered pretty much everything except animations and some mysterious things, for the rest that won't happen probably, what you should know is that it is a really optimized game on the PSX therefore the code is too, i.e. I won't get much more than that as I'm not an ASM guru on the PSX. I did manage to get these missing polygons on track borders as they were deeply buried somewhere else through a 30+ hours debugging session so for the visuals and sounds e have all we need, there is too little info for the animations, i know where these are but I can't get anything out of them, maybe some quaternion but since it's mostly zeros I'm screwed ... But these are simple and I'll do them by hand.

So you have some EXE that has surely been optimized out using a piece of software because it really made no sense for it to be done this way by humans: there's practically no chance i figure much more than what I've already did.

You were saying that control is important but let me tell you that camera is, too. I'm trying to get a similar camera but it's not easy ... The next version will open new settings and maybe people will figure out better settings than I thought of.

Thing is as I keep refining, things are progressing more slowly, but surely ! We're not there yet but will one day, and I'll ask bigsnake and Xpand from BnG to see what can be improved.

I'll try to wrap up a demo of the latest soon within a few days ��

[80T]

Looking forward for the next Beta :) And yes, Camera is also a very important factor :)
I like the idea of streaming the music tracks and all sound effects from the CD itself.
It will make for a more genuine experience overall, as close to the real thing as possible.

[aybe]

Of course, Wipeout wouldn't be Wipeout without the music :g

[80T]

Yeah, I know WO games always shined at this chapter. I'll give 2097/XL the 1st place for music tracks ;) It was also the game that introduced me to the series.
Man, do I remember the thrill hovering at insane speeds listening to Prodigy back in th 90's... Then came WO3 with it's clean look and that overall polished feel...

Anyways, is there any kind of "roadmap" that we could track for this project ? Or any plans to set one up if there isn't one yet... :D ?

[aybe]

I started playing with WO shortly after it's been in stores I guess around Sept. '95, then WOXL followed and is too my favorite esp. the music ! The Prodigy music was simply insane, it never gets old :guitar

For WO3 actually I haven't played until it around 2006 IIRC when I bought a new PSX and I was simply amazed by its look. However now I can hardly cope with Rapier class, let alone Phantom class ... it's just too fast :blarg

Not sure if you have access to this thread http://www.wipeoutzone.com/forum/showthread.php?8897-Extracting-WipEout-3D-models-(large-images-allowed) (I think one needs at least N posts or something), the whole history of the reverse engineering of WO3 I did is in there, and it started in April. I did start the previous project with WO/WOXL exactly a year before, so this one is about 3 months old.

For a timeline I would love to say tomorrow :D but I really have no idea but ASAP !

Here's a rough list of top-priority things:

- hovering/jumps: mostly done except that tracks like SampaRun are problematic and this is likely to need a refined approach; it is really hard at the same time to try play WO3 properly and try understand the inner logic of it

- finish track collisions -> shouldn't be very long

- levels/scenery -> low-level stuff figured out, high-level stuff needs manual work (checkpoints, booster/pads: not found in game data and therefore I have to run through each track and write down section/face) -> not complex but very repetitive and long

- animations -> shouldn't be very long as there is a dozen by track, instantiation by code has been figured out (in fact with Unity when you step out a bit from their intended path you are on your own, in my case 95% of stuff is done by code rather the mouse, a way (officially) not supported by Unity) of course I do profit of Unity facilities by it's not like when one imports models but OTOH there are advantages with code: my scene is not littered with a billion things placed manually

Now when I have these things which I consider essential to qualify the project as a honest 'preview', the cake topping should be addressed:

- audio/video : will be quick, all low level stuff is written: VAG, WAV, looping, extraction etc ... in fact it was not a matter of simply getting PCM or frames, Unity should be able to consume them, i.e. videos must be OGV and so on

- the installer : mostly done, basically one will simply insert his WO3SE CD and it will take care of the rest, extracting videos manually is not the faint-hearted and I don't actually expect an end-user to do this.

- AI/multi-player : will have to be done, no previous experience about it but there is significant info out there and I might even get help from BnG crew

- interface/navigation : not particularly complex but there are a couple of things I need prepare before

- shadow/lighting issues : in fact they have 'baked' them in the textures and therefore applying Unity standard shader simply looks worse, this is likely to happen when topping the topping ,i.e. when the game is done and could be modded or whatever (we're not there yet)


To sum it up / TL;DR:


Pretty much all the low-level stuff is done, high-level remains for the game internals but as a well as for the Unity side.

Actually as time passes on, I spend a more and more significant time on the research part and not that much as one would think on coding, i.e. how can this be done efficiently, would it work and is this really what I need ? On the Internet 99% of stuff is crap when you do a search about game programming/Unity/books, beside people here that helped out for physics and stuff I rely on my own approach but this takes time (but in the end it is reliable).

Another thing is testing, whenever I make changes I have to test and there are 22 tracks, I do not run through all but a few like Sampa Run and Prototype 3 as these have tight tunnels and unless I get an approximative WO experience I'm not happy.

So progress in this project is like a roller-coaster !


To make you salivate a bit :D

10305

10306

10307

10308

10309

10310

10311

10312

10313

The red triangles you see are the missing ones I've told, was about to drop off completely if I couldn't get these but an intensive week with an hex editor and a calculator allowed me to figure it out entirely :D

So many things are ready but I'm simply not pushing them because they need to be optimized, i.e. the scenery loading takes almost 10 seconds on its own because nothing is merged for debugging purposes, if I have say to run the project 100 times a day it equals to ~20 minutes lost in waiting. Add to this the Alt-Tab I have to do everytime I go back to Unity and wait for it to pick the changes (a few more seconds), you quickly end up in lots of time lost by waiting.

But don't worry, I think I'll soon reach the summit of the coaster and gain a huge speed boost downhill (until next uphill :D)

[80T]

I started playing with WO shortly after it's been in stores I guess around Sept. '95, then WOXL followed and is too my favorite esp. the music ! The Prodigy music was simply insane, it never gets old :guitar

For WO3 actually I haven't played until it around 2006 IIRC when I bought a new PSX and I was simply amazed by its look. However now I can hardly cope with Rapier class, let alone Phantom class ... it's just too fast :blarg

I've never returned to WO XL/2097 after I've touched WO3. For some reason WO3 simply fit my bill... I liked the lighter look and the cleaner style and the sounds and all that... the funny billboards and everything make it look more idk... "happy" or something like that... less "rugged" and "brutal". Idk, it felt more like a "sport" than a crazy "futuristic AG fight"... And that's coming from a Survival Horror fan :)) Same was with Gran Turismo. When GT2 came out it felt more "open" and "relaxed" although more complex overall... I didn't like GT2 very much at first, probably simply because of change... But man, did I loved it after I've got accustomed to it... still do. Last time I've finished GT2 was actually last year around Christmas. And if I'm still on the PSX chapter... I've also finished Parasite Eve 2 just about 2 or 3 months ago... something like that. Dang fine game!


Not sure if you have access to this thread http://www.wipeoutzone.com/forum/showthread.php?8897-Extracting-WipEout-3D-models-(large-images-allowed) (I think one needs at least N posts or something), the whole history of the reverse engineering of WO3 I did is in there, and it started in April. I did start the previous project with WO/WOXL exactly a year before, so this one is about 3 months old.

Nope, got no access to that thread just yet :g Maybe later, Idk what's the post count requirement for that.


For a timeline I would love to say tomorrow :D but I really have no idea but ASAP !

The timeline doesn't matter. Only the end result ;) Be it when it's done :)


Here's a rough list of top-priority things:

- hovering/jumps: mostly done except that tracks like SampaRun are problematic and this is likely to need a refined approach; it is really hard at the same time to try play WO3 properly and try understand the inner logic of it

- finish track collisions -> shouldn't be very long

- levels/scenery -> low-level stuff figured out, high-level stuff needs manual work (checkpoints, booster/pads: not found in game data and therefore I have to run through each track and write down section/face) -> not complex but very repetitive and long

- animations -> shouldn't be very long as there is a dozen by track, instantiation by code has been figured out (in fact with Unity when you step out a bit from their intended path you are on your own, in my case 95% of stuff is done by code rather the mouse, a way (officially) not supported by Unity) of course I do profit of Unity facilities by it's not like when one imports models but OTOH there are advantages with code: my scene is not littered with a billion things placed manually

This is looking so good and promising :nod This sounds like dedication and attention to details, not the random run of the mill "quickest way possible to make it run until the dead-line hits" :+


Now when I have these things which I consider essential to qualify the project as a honest 'preview', the cake topping should be addressed:

- audio/video : will be quick, all low level stuff is written: VAG, WAV, looping, extraction etc ... in fact it was not a matter of simply getting PCM or frames, Unity should be able to consume them, i.e. videos must be OGV and so on

- the installer : mostly done, basically one will simply insert his WO3SE CD and it will take care of the rest, extracting videos manually is not the faint-hearted and I don't actually expect an end-user to do this.

- AI/multi-player : will have to be done, no previous experience about it but there is significant info out there and I might even get help from BnG crew

- interface/navigation : not particularly complex but there are a couple of things I need prepare before

- shadow/lighting issues : in fact they have 'baked' them in the textures and therefore applying Unity standard shader simply looks worse, this is likely to happen when topping the topping ,i.e. when the game is done and could be modded or whatever (we're not there yet)

1. Good idea with the Installer :+

2. I'd imagine that the AI it's the same deal as the handling, in the sense that it cannot be "reverse engineered"... ? If I recall right, BNG devs also made their own AI.

3. Are you willing to keep the "origianal-ish" UI style ? Most noticeably, the in-game HUD.

4. About the pre-baked textures: I totally understand what you're saying and for starters it would be perfectly fine to use no shading (maybe only shadows for the pods... if possible ?) But for the future, when the game it's already playable I'd suggest something like "TexMod" but only "integrated in the game". Much like what Counter-Strike 1.6 is doing since the 2013 Update - It introduced an "Addon Folder" actually named "cstrike_addon" that's "mirroring" (sub-folders structure and all) the actual game folder, (named "cstrike" that's containing the standard game assets) and replacing every asset in game with the content found in the "Addon Folder". So if i drop a proper named replacement custom model in the "Addon Folder" in the right sub-folder, when I load the game, that custom model will be loaded instead of the original default one. So with WXX-Rebirth maybe we could have a folder named "HQ_Textures" and a reference index text or html file with texture names and their "description", so people could slowly populate that folder over time with "standard" ("not baked") textures, or why not, even higher resolution ones. Maybe even better, (easier for the end user) have a console in the game to introduce commands and have a command that will dump all textures as they're loaded by the game and create the folder automatically. Or if an in-game console would not fit the purpose of this project, maybe a launch parameter to the shortcut would put the game in "dumping mode" and dump all loaded textures in the folder. From there, people could edit the textures while already seeing what they're editing and have the names already in place. So, with all said and done and having all textures in high quality and "not baked" we could have another launch parameter so when the game runs, it will load any texture found in the "HQ_Textures" Folder. Also with all those new textures, we could now use some shading :g Maybe we could even go as far as using multiple textures like bumpmaps and stuff like that... ? Ex. texture_01.dds > texture_01_bump.dds ... ?


To sum it up / TL;DR:


Pretty much all the low-level stuff is done, high-level remains for the game internals but as a well as for the Unity side.

Actually as time passes on, I spend a more and more significant time on the research part and not that much as one would think on coding, i.e. how can this be done efficiently, would it work and is this really what I need ? On the Internet 99% of stuff is crap when you do a search about game programming/Unity/books, beside people here that helped out for physics and stuff I rely on my own approach but this takes time (but in the end it is reliable).

Another thing is testing, whenever I make changes I have to test and there are 22 tracks, I do not run through all but a few like Sampa Run and Prototype 3 as these have tight tunnels and unless I get an approximative WO experience I'm not happy.

So progress in this project is like a roller-coaster !

Reliable and slow it's much better than fast and ugly :))


To make you salivate a bit :D

10305

10306

10307

10308

10309

10310

10311

10312

10313

:hyper:hyper:hyper:hyper:hyper Can't wait!... to :rock on those tracks :g


The red triangles you see are the missing ones I've told, was about to drop off completely if I couldn't get these but an intensive week with an hex editor and a calculator allowed me to figure it out entirely :D

I thought those were some selected faces in a 3d editor or something. Glad you figured them out :+ It's a pain to be so close to something and not having it all just because... random reasons. That week's work well worth it ;)


So many things are ready but I'm simply not pushing them because they need to be optimized, i.e. the scenery loading takes almost 10 seconds on its own because nothing is merged for debugging purposes, if I have say to run the project 100 times a day it equals to ~20 minutes lost in waiting. Add to this the Alt-Tab I have to do everytime I go back to Unity and wait for it to pick the changes (a few more seconds), you quickly end up in lots of time lost by waiting.

But don't worry, I think I'll soon reach the summit of the coaster and gain a huge speed boost downhill (until next uphill :D)

Until next uphill... :g

[aybe]

WO3 is the pinnacle in terms of Wipeout franchise on PSX, before :brickwall (IMO) with subsequent releases ... I've never played GT because at that time I did not have the PSX anymore (sold it I can't remember for what reason, was 15 then :paperbag). I was of those who had and played Bio Hazard in japanese centuries before it's been released as PAL, was astonishing ... now I've played PE2 in the same period as with WO3 so roughly ten years later and had to say it was very impressive though I've never finished it.

Well I seek quality about this project, we won't be able to get a replica of WO3 at least in terms of sim, but everything will be tweak-able so as I said before, one might come up with better settings than the ones I've picked and eventually we'll get to something closer. I have no deadline and therefore no real obligation but that could have boosted the thing quite a bit...

The problem with the original is that it is PSX, ASM, and optimized. First you need to know these well (I don't), then since it's ASM there's a ton of inlined stuff which makes things even more confusing and finally compiler optimizations further change the original shape of the code ... that was probably necessary since the game runs fast and is pretty for a PSX game.

The texture modding should be fairly easy to accomplish since it's really just about different sizes for Unity, a clean integration in the code might be the hardest part; now the shading is in the polygons themselves, i.e. if these are not flattened against their 'parent' shade, HQ textures will be shaded too. This is likely to take a lot of time and I guess will need a dedicated editor. Not sure I'm very clear in explaining :D basically there are up to 3 color slots so 2 should normally be free but they won't be because this is where are the extra shades for Gouraud-shaded ones and it is what gives these nice colors. I could somehow treat all polygons as non Gouraud but I'm pretty sure it won't look nice ... unless there's a huge work with the lighting sources.

So on bottom: raw color, global tint, local tint

10314 10315 10316

Idem:

10317 10318 10319

Idem:

10320103211032210323

In short to get real lighting, it will be very long because one has to normalize these shades, then re-create appropriate lighting and re-coloring -> a job for a graphic artist that will need a tool to facilitate this.

22 tracks * ~10000 polygons is what to expect and these screens are when I was finding it, now it's deeply buried in the game, would have to defeat it and compare but there are scattered throughout levels. They used a 3D editor and then it was converted, doing the opposite will not be easy :cowboy

[80T]

I think I know what you're saying. (I guess... ) In order to use textures with the nowadays game engine standards in mind, thus, lighting, shadows, occlusion, etc. there should be one diffuse texture (the color map > 3 color channels - R/G/B), one normal texture (the normal map > 2 "color" channels R/B) and optionally bump maps (one "color" channel - could be mixed in the normal map... as "G" channel ? I think Alan Wake was doing something like this... ), occlusion maps, specular map, etc... Not only that, but many (or all I think... ?) textures should be "averaged" by texture artists in order to "convert" them from multiple baked instances into something closely resembling the diffuse map (original intended color map). The one slight "advantage" here I suppose would be that many textures are "duplicates" (tiled most of the time, although in the PSX game itself it's just giving the appearance of tiled textures, since every texture it's being it's own instance being pre-baked multiple times for multiple use instances... ) So many textures are just the same "base texture" but with pre-baked shadows, or lighting, or maybe even reflections... Yep, seems a lot of work, but I think it's "doable" with the help of a few decent texture artists. Many times there will be tens of textures that could only be done once (then, either duplicated and "rename" to the original textures names, either through the game engine one texture could be mapped to a number of predefined polygons... by indexing the corresponding replaceable texture names originally assigned to those poly's ?)

Anyways, this would be a distant future thing to consider :) Meanwhile, PSX pre-baked GFX should be totally fine ;)

[aybe]

Yes and if it's for sub-par modding I really don't see the point.

I'm not really a fan of modding actually, HD textures could be interesting but even that sometimes does not look well due to poor environment. I do like nearest-neighbor rendering actually ! And a CRT shader could come in handy :D (new engine but old look'n'feel)

I will certainly put these later https://github.com/aybe/RetroArch-shaders esp. the CMYK one, that paper ink color is just great (you really have to play a game you like with it to feel the experience, it's really subtle)

These are great too:

https://github.com/keijiro/KinoGlitch

https://github.com/keijiro/Retro3D

[80T]

CRT Royale FTW! If you plan on integrating multi-pass shaders like that one, then Bob's your uncle! I COMPLETELY APROVE!
:rock:rock:rock:rock:rock:rock:rock:rock:rock:rock :rock:rock:rock:rock:rock:rock:rock:rock:rock:rock :rock:rock:rock:rock:rock:rock:rock:rock
Better than any shading :clap:+ Oh, I can't even imagine some CRT-Royale combined with some NTSC & S-Video/Composite & VHS passes :dizzy

EDIT: Yep, those "Analog Glitch" and "Digital Glitch" look awesome (especially along with some CRT shader ofc :g)
The "Analog Glitch" it's currently implemented in BNG as a "visual damage indicator" if I'm not mistaken :)
Oh and that "Retro3D" also works very well with this game. Also a shader I've seen in BNG and it's looking fine :+

[aybe]

You don't seem to have been impressed that much by the CMYK shader :evil, believe me, go grab Retroarch (there are many top CRT shaders as well in the latest build, such as Royale and so on) and play some of your favorite game, preferably with 15+ bit colors so SNES or anything beyond ... it adds quite some drama to the scene ... come back and let me know then. :)

Since I posted this, a couple of color-accurate shaders for handhelds such as GameBoy Color have appeared on Retroarch, basically it's the same principle as one of the shader in the Unity standard assets but pushed to the extreme -> instead of a 256x16 pixels CLUT, it is a 24-bit CLUT so 4096*4096 and that along a nice color profile (from Photoshop) gives really subtle results. I deliberately chose the AGFA SWOP profile because it's a bit out-dated and it gives to greens, blues and magentas a whole different experience.

Now for the shaders, I will probably pack in 2 or 3, then people will be able to extend as needed, would like to write my own CRT shader though :D (btw the pics already use a custom shader since Unity did not offer one that is PSX-friendly :banzai)

also, if you know any racing game with nice camera angles, let me know so I can take a look at it and get some inspiration for my camera :p thanks

EDIT: Bob's your uncle <- :clap:clap:clap

[80T]

I kinda see what you did there with the CMYK shader :) (btw, I had to swap the texture extension from .jpg the .png in the .cg to actually make it work. Maybe a typo... ?) But I think it's not only a subjective, but also a of video chain setup matter.
I have an nVidia GPU and a cheap TN panel. Maybe if you have an AMD card and a better display, the colors look better on your end. Regardless, when it comes to PSX games, I'm all about accurate levels (which seem to be a pain to achieve) and strong, "vibrant", yet "natural" CRT like colors... (yep, the slight cyan tint included :g) If only I could find a proper gamma/tint, ("gamut compression" I think would be the best term... ?) shader for once... to closely resemble CRT colors contrast and brightness/"radiance" or however I could call it... I am rarely looking for enhancing "overall tones" in the detriment of other "shades"... Usually I aim for a clean, neutral look without tinting effects or over sharpening or HDR or stuff like that. I like everything to be "enhanced on it's own", every element to have it's own "visual power" in a defined range and not being overpowered by other visual elements... idk how to explain it better... For ex. I never use dynamic contrast or stuff like that to "enhance" the final image. At most, I'd be using slightly cooler color temp than standard.

[aybe]

There's something wrong with the commit, there shouldn't be a jpg ... The one I wrote was png and somebody pushed a commit and I accepted without really looking well... It's absolutely vital the clut does not get compressed for it to work well ... Will take a look.

I use f.lux and that can be seen in shots. In fact it should be disabled...

I perfectly understand your view, it's just that sometimes I want to drift a bit from neutral look and these Crt shaders while being cool are quite fuzzy in the end and it gets boring.

Btw have you seen epsxe 2 ? They finally added shaders and there are some crt ones !

If you have a shader you like and is not gamma enabled it's very easy to get it, raise the final color to the power of 2 and Bob's your uncle :D

[80T]

Yeah, I was wondering why there even were those .jpeg duplicates there in the first place :)
I know about f.lux, I think there was a Windows version of it or something like f.lux on Windows but I never actually used it.
In fact I've tried it few years ago just to see the concept in action, but never felt the need to use it on a daily basis.
Yes, i know various tone mapping and other effects just works for some games depending on theme and I've played such games.
For the majority of times though, for my PSX games, I just got used to their "CRT look" (too bad that I don't have a CRT any more...)
I had to get rid of it since it was dang big and I was realistically using it maybe once or twice a year at best... I kinda regret tho...
I know about ePSXe 2 but wasn't aware of the changes. I've mainly used RetroArch since I've found about the PSX libRetro Core.
I recall there was Pete's OGL2 GPU plugin that was supporting shaders on ePSXe since 1.6.0 times... ? And edgbla's soft GPU plugin.
The "issue" with those for me is that they only support one pass and you can only do so much in just a single pass...
Since I've discovered RetroArch, nothing else satisfies me any more :)) The multi-pass shader support was the ultimate feature for me.
There's also ReShade that you could use for pretty much any game but I can only see so much use for a "final pass" solution sadly...
Yeah, you could use that for some final tone mapping or a CRT post-process effect in OLD (PSX like old) games, but other than that,
I don't see much use since it's messing with the UI, the fonts and all that. So unless there's prior access to the game's "shading chain",
it's usage it's really limited imo. Also there's no CRT-Royale in ReShade yet :))
Yeah... raise to the power of two and... yeah... mhmm... sure. Should I put a comma before or afte the power... :?:
Sory, nicht sprechen GLSL, nein! Niet C++, niet Python, niet ASM! Nada! Je ne sais pas d'utiliser le code :dizzy :naughty

[aybe]

I really don't know why there is a JPG in the commit ... I guess I've changed things after a commit, because here I don't have it. Anyway, should be the 13Mb PNG, by the way make sure you select 'Don't care' and for scaling and 'Nearest' so you retain the old blocky style rendering.

F.lux is good at night only, obviously :) And yes CRT is the way to go but it's cumbersome, actually I've bought the PSIO thing to play PSX games from SD card but it needs to be soldered and haven't done it yet ... on the other hand, a mouse click on PC brings the whole access to games through Retroarch ... but the latest version is quite weird.

here's gamma for retroarch: (right of the pic)

10325

cgp



shaders = "1"
shader0 = "gamma.cg"
filter_linear0 = "false"
wrap_mode0 = "clamp_to_border"
float_framebuffer0 = "false"
scale_type_x0 = "source"
scale_x0 = "1.000000"
scale_type_y0 = "source"
scale_y0 = "1.000000"


cg



void main_vertex
(
float4 position : POSITION,
out float4 oPosition : POSITION,
uniform float4x4 modelViewProj,
float2 tex : TEXCOORD,
out float2 oTex : TEXCOORD
)
{
oPosition = mul(modelViewProj, position);
oTex = tex;
}

float4 main_fragment (
float2 tex : TEXCOORD,
uniform sampler2D s0 : TEXUNIT0
) : COLOR
{
if (tex.x > 0.5)
{
float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);
}
return tex2D(s0, tex);
}


If you're happy with the results and want to apply full screen, simply do replace



if (tex.x > 0.5)
{
float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);
}
return tex2D(s0, tex);


by



float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);

[80T]

I really don't know why there is a JPG in the commit ... I guess I've changed things after a commit, because here I don't have it. Anyway, should be the 13Mb PNG, by the way make sure you select 'Don't care' and for scaling and 'Nearest' so you retain the old blocky style rendering.

Yep, there is a 13Mb PNG there. And about the "blocky style", I don't really care about that. I'd use 2x internal res, but it's a pain with the shaders and with the smoothing of 2D backgrounds. The CRT-Royale goes Interlaced for ex. and other issues...


F.lux is good at night only, obviously :) And yes CRT is the way to go but it's cumbersome, actually I've bought the PSIO thing to play PSX games from SD card but it needs to be soldered and haven't done it yet ... on the other hand, a mouse click on PC brings the whole access to games through Retroarch ... but the latest version is quite weird.

Dang! You have a PSIO! (and you didn't yet installed it :dizzy) I envy u! If I've had one of those on my hands, I wouldn't stand one second without installing it. But then again... I'd never play a PS1 game on an LCD TV... :blarg


here's gamma for retroarch: (right of the pic)

10325

cgp



shaders = "1"
shader0 = "gamma.cg"
filter_linear0 = "false"
wrap_mode0 = "clamp_to_border"
float_framebuffer0 = "false"
scale_type_x0 = "source"
scale_x0 = "1.000000"
scale_type_y0 = "source"
scale_y0 = "1.000000"


cg



void main_vertex
(
float4 position : POSITION,
out float4 oPosition : POSITION,
uniform float4x4 modelViewProj,
float2 tex : TEXCOORD,
out float2 oTex : TEXCOORD
)
{
oPosition = mul(modelViewProj, position);
oTex = tex;
}

float4 main_fragment (
float2 tex : TEXCOORD,
uniform sampler2D s0 : TEXUNIT0
) : COLOR
{
if (tex.x > 0.5)
{
float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);
}
return tex2D(s0, tex);
}


If you're happy with the results and want to apply full screen, simply do replace



if (tex.x > 0.5)
{
float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);
}
return tex2D(s0, tex);


by



float4 color = tex2D(s0, tex);
float alpha = color.a;
float gamma = 2.2;
return float4(pow(color.x, gamma), pow(color.y, gamma), pow(color.z, gamma), alpha);


For some reason the original .cg with the "if" statement only works for the BIOS Boot sequence, then it's gone... I can see a darker grey stripe/shade area on the right side while the SONY Logo it's displayed, but that's kinda it...
Once it's booted into the game, it's completely gone. After replacing the "if" statement paragraph, it's working, but it gives me a very dark image. Like I'd lower the brightness to something like 25%... Meh, nevermind...
Even if it had worked, without "implementing it into a shader" I wouldn't have much use for it. I just can't stand the "non-filtered" blocky/pixelated look of 224/240p of PS1 games. I'd use a gaussian blur at the very least.
I'd take a blurry mess over the blocktacular sharp ginormous pixels every time in the absence of anything better :nod That's why I usually use the CRT Shader for low res PSX games with emulators (RetroArch)
It "mixes"/"fuses" those huge pixels into something looking more like an actual image :) Granted, a very low detailed image, but at least my eyes aren't bleeding any more from the sharp edges poking into my retina.
I think an overlayed/blended "fixed resolution mask" (same resolution as display's native one) on top of the upscaled native PSX game's pixels would work wonders. Currently, most CRT shaders are tied to game's vertical resolution.
That's why the scanlines won't ever align/space perfectly equidistant and diffusion, bloom effects will always be distorted, since PSX's resolutions are anamorphic and 1080p for ex, never divides perfectly by 224 or 240, etc...
But I digress...

Some edits I've done to CRT-Royale's mask texture while I was playing PE2 :g
Yeah, the levels are allover the place and the contrast and colors and all... :mad:
But the "pixels effect" it's almost decent... Not what I was aiming exactly...
The smoothing and blending I think it's the best part. Round (almost) fonts etc...

https://s32.postimg.org/6i0nmz435/Retro_Arch_0608_031235.png (https://s32.postimg.org/rro9xtkdv/Retro_Arch_0608_031235.png)https://s32.postimg.org/4oxqyniwh/Retro_Arch_0608_031210.png (https://s32.postimg.org/6ti3zqkj7/Retro_Arch_0608_031210.png)https://s32.postimg.org/cwzojnasx/Retro_Arch_0608_031247.png (https://s32.postimg.org/lf94nzhbn/Retro_Arch_0608_031247.png)https://s32.postimg.org/3qhdwd5kh/Retro_Arch_0608_031309.png (https://s32.postimg.org/ru85kno1f/Retro_Arch_0608_031309.png)

Zoom to 100% (1:1) to avoid downscaling distortion.

[aybe]

These Retroarch shaders are strange in regards to how they behave, I've put 0.5 which should be screen center but it's not ...

Now what you're experiencing is normal, a better shader would account for a PC screen gamma and adjust the output accordingly because it's different for a PC screen ... but this is the formula to go from linear to 'gamma' ramp space.

Remember that you can set or not linear/nearest in each of the passes in Retroarch, so take the one you prefer, shader logic does not interfere at all with this.

etc ... right, we're drifting a bit from the original topic :g

- - - Updated - - -

These pics are nice, mask is too, but see there's a problem in fact with their implementation: the color smearing should occur between masks just like in a real TV and since it's not, when you use a mask the overall loses brightness while on a real TV we could say that brightness literally consume your eyes, especially in darkness :)

[80T]

The best part about real CRT TV's it's that it's "interlacing"/"interleaving" different colors from different "pixels" since the scanlines are thicker with brigtness and overlap, but they overlap without blurring...
I wonder if a well thought fine grained "explosion" shader would give such effect... something like how CMYK pixels are looking layed out on a photo... There is actually an "explosion" shader in RA, but it's "rough" and too blocky...

Some more images in the waiting for WO3 :g Overbrighten to Heaven and back lol...

http://postimg.org/gallery/2lzdq7fzs/

[aybe]

Pretty cool shader !

I think the only way to be satisfied would be to write the thing directly, there are numerous examples out there, but it has to be done :mr-t

[80T]

If you're talking about the CRT-Royale shader, (since that's what I've used in all those screens, just with a modified mask texture) then yes, it's quite decent, by far the best "CRT look" approach from all I've seen so far.
But it's far from being a very good representation. There's lots of room for improvement and I think even efficiency (it "only" takes about 14 shader passes iirc). And all the light diffusion effects are susceptible to distortion.

[aybe]

Back to the main topic now !

Here's a video of what I've been working on lately : animated billboards


https://www.youtube.com/watch?v=tMMBvQByRWk

Did capture a WO3 footage, identified and wrote the transitions all by hand :pirate

OTOH it took quite a bit to get to this as, as I said before I cannot build anims using mouse as that would mean I've included copyrighted content, what would open doors to any copyright holder.

So it's all done by code but since it's a grey area it took time to get to this, I ended up writing a mini framework to generate all things I needed and now that I'm hooked to Unity animation system I should be able to leverage its features. Gave a try to naive approach like blitting texture atlas directly but it eats CPU and did not address other concerns ...

Decided to pause the work on the camera, while it works I'm not satisfied with it, so I'm focusing on other parts and maybe I'll get something better after letting it sleep for some time !

[80T]

Can't wait to see those billboards "live" :g And yep, the camera needs to be tackled with on a good day ;)

Regarding the hundreds of duplicate textures and maybe other stuff... may this be of any help... ?
http://phoboslab.org/log/2015/04/reverse-engineering-wipeout-psx

[aybe]

This was the first article I read when I decided do that project, he did a great job and the web-app is appealing ... so I know that article.

Been able to hook events to my animation system in clean way, should be able to get sounds up easily but now I need to address outer layers to not go berserk. Actually I've run a code analysis yesterday, and while I do progress slowly I have a high maintainability rate which means that I can add things relatively easily :p

[80T]

Nice! :g

OFF: I'm wondering where everybody's at...

[aybe]

You mean people are not actively participating in this thread ?? I guess they expect the 'goods' and I too ... Whenever I have something decent I'll ask the admins whether they can add a section in the forum next to AG racing projects and it will likely gain some visibility.

[80T]

Weekly Check-In! ;)

[aybe]

Here's a really quick report of what I've been onto:
- finished game provider data
- namespace and files sorted (much clearer now)
- started looking at other parts, i.e. sound
- the coder-friendly animation framework is now on par with features provided by Unity

Just as last time, taking a break from the racing part and focusing on other parts -> the interface :)

[aybe]

Sound is up !


https://www.youtube.com/watch?v=ouQ4hZqfgdk

The path to get to this was long and with lots of traps, but we're getting to it !

Basically, in the unconventional way Unity is used, most of the stuff that happens smoothly using the mouse simply doesn't in code :brickwall Like, why does a sound is effectively loaded but is simply silence ... at this point most of the time spent is about figuring out why it doesn't work, little really in the coding ... but over time I'll get done with this dirty low-level work and boost productivity somewhat :)

[80T]

Nice! Idk much about coding, but... maybe the sounds need some manual channel assigning and other params like pitch, volume, and stuff like that when you're manually trying to implement them... ? Just a thought...

EDIT: Just found this at random...

https://www.youtube.com/watch?v=ttr_VItKoXk

That CRT neon like glow effect on the highlights (billboards, windows boost pads, etc... ) Dang!

[aybe]

I will clearly attempt a CRT shader one day :D

In Unity you have audio mixers (https://docs.unity3d.com/Manual/AudioMixer.html) and you effectively assign sounds to channels ... while this is easy with mouse, the problem in code is finding exactly what's needed to achieve one of the different behavior you can set with mouse ...

Now add to this that you need to figure the order of achieving things, and since Unity is a black box (we know nothing about how things are done internally even though the classes they use are readily available) it's what's taking most of my time : figure out basic path, then specific path;

Like for the animations, I (relatively) quickly got them up but they were blending incorrectly, I needed rough transitions ... one has to figure some tangents to have the correct behavior etc etc ... that's what taking most of the time -> research.

And for the animation and so on, I need to write extra methods but instead of quickly coding them, I take the time to write something clean so I won't have to come back another time at the topic :)

I have the situation (mostly) in hands, but as a one-man effort it just takes more time :pirate

[80T]

Weekly Check-In :g

[aybe]

Haven't really had the time this week to work on it ... :g

[TroJan_Virus]

I suppose this essentially makes my Wip3out 4K remake redundant... :(

Looks great though!

If you need any help with graphics let me know, although considering this uses base game textures you probably don't haha

[aybe]

Thanks,

If you wish to achieve a personal project I think you should go for it :D I already knew about BnG but that hasn't stopped me from working on this project, WO3SE also did exist so there were numerous 'excuses' to not work on this but I didn't ... So if you're dreaming about something I think you should make it happen :D

Regarding WXX I really want to make it open in GitHub, but not unless there's an 'incentive package' for people to get hooked on ... I know I'm repeating myself but I'm getting to it :cowboy these last 2 weeks I haven't had time to work on it and I guess probably a few others I won't be able to (summer) but I'm definitely committed to it. By package I mean shaping up the overall thing, menus/UI, scenery, animation, sounds etc ... all what makes a game in the end. Now when this is done I'll ask some of the experts about looking at better simulation and IA.

The main things I'm spending my time on recently is figuring out how to achieve things by code in Unity, since it's a black box and there's pretty much no documentation when using it that way, it's a matter of trial and error. But sooner or later I'll be finished with that aspect, then as I've already worked on many other aspects of the game I will get a productivity boost.

I see all that happening as 2 releases, the first one would show a 'familiar' WO3 world with full scenery implemented (i.e. racing part) that should give a smile to people, the second one would be about a game-like release (menus, etc ...).

For the graphics I'd definitely love to see creations of some people, while not exceptionally complex to achieve, the difficulty lies in the many places where there are textures and how to hook up (easily) according data structures of the game. The other difficulty would be to provide a tool to modders to visualize/highlight their work in the scene (some sort of editor). It is definitely doable and should not take ages (not a priority, though :))

[80T]

Nice job there TroJan!


https://www.youtube.com/watch?v=dvwiEyBiWYg

How 'bout "joining forces" for an ultimate WO3 Dev. Team :g ?

[aybe]

Very impressive :clap

Honestly I didn't even think about that and I'm clearly open about collaborations ! The roll out I've talked about is not 'fixed', merely some stuff I've thought about ...

So I'm totally open, let me know what were your initial thoughts about all that and we could go from here ... These animations are great for introducing teams !

I can provide unpacked game content if necessary (sounds, graphics), and just pause what I've been onto to help shaping up supporting code. It's just that these days I really don't have much time to work on it but soon I will have plenty of time again :)

[80T]

I suppose you were replying to TroJan, since that's his work in that video from his YT channel.

[aybe]

Today I had some time to work on the project :p

So the first step in getting the whole thing up !


https://www.youtube.com/watch?v=3MIknGqjgHg

The rest should follow soon : fountains, sewers, lights, ad rotators ...

- - - Updated - - -


I suppose you were replying to TroJan, since that's his work in that video from his YT channel.

yes or anyone actually :D

[80T]

Yeheyyy! Progress is made :g Nice!

[aybe]

Crane animation !


https://www.youtube.com/watch?v=TxgEUKayYP0

Sorry for the quality, I use the XBOX DVR in Win10 but still at max quality it's :turd

[80T]

.................. in Win10 .................

Wait, what !?!


https://www.youtube.com/watch?v=umDr0mPuyQc

[aybe]

I like living on the bleeding edge :rock

I really don't know why all that recent fuss about it, I use it since day 1 and I'm really happy with it (after the initial period of nasty bugs with video corruption); it's fast and works well. Note: I don't use that funny start menu but Classic Shell instead, also I don't use Cortana nor all the UWP apps. (I mainly do work with it, i.e. I don't have time to waste on funny features). The other thing I don't get is that on nearly 10+ Win7 machines I support on a domain, absolutely none of them auto-magically installed Win10 even though the wizard was opened at times, technically Win10 rocks but 'reputationaly' it doesn't !

[80T]

Yeah... I just don't want to deal with the "extra setup" after a fresh install... or deal with the "identity crisis UI" or mess with the registry settings, or dealing with "Modern Apps" leftovers since you can't completely "eradicate" them, or lack of control over my updates... privacy settings... and so on... Too much pita to go by... I mean what's new after all... DX12... ? Meh, I can live without it for few more years. Instead of implementing Tabs to Windows File Explorer, they give me Cortana and 2 types of Control Panel... Yeah... :)) Nope, seriously, I don't see the point in Windows 10. It's not like my PC would crawl on it's knees or something and Windows 10 would be this be all end all save my PC from dying kind of deal... If anything, I see Microsoft going towards Apple like policy and slowly taking away more and more out of my personal user experience and most importantly the control I have over my own PC. I wouldn't be surprised if down the road Windows would turn into a monthly subscription based OS asking me to "pay my rent" in order to Log in... :)))) Well at least pay a fee to get rid of Log In Screen Adds... I can definitely see that becoming a thing in the nearest future if we're just embracing "everything new" from Microsoft, or any other big company for that mater... Just telling myself that, "hey, I'm gonna do that, disable that, tweak that, not use that... and so on" won't take us anywhere good in the future imo. Btw, did you installed the "Anniversary Update" ? I've heard that you can't disable Cortana any more...

[aybe]

No I didn't install it, had to customize settings once at install then I forgot about them ...

hint 1 : I don't fiddle with the OS
hint 2 : I use a computer how it's initially meant to be used, a device for work and eventually gaming -> no webcam, no mike etc ...

Mmhh we certainly won't agree but I find it a definitive improvement over the previous releases :bomb

Eventually it will become the next Win7 in a few years ...

[80T]

Ok, just please, please don't make this a DX12 exclusive game :lol

[aybe]

Here's a quick update about what's going on,

The scene editor is in good shape ! :P

10408

I took the decision that levels should have their own scenes, much like any Unity game:


easy to design
easy to extend
less code, more mouse
etc etc etc ...


This has many implications but I think it's the best because endlessly gathering numbers and numbers for models and polygons then writing some code was really tedious and in the end very long.

Obviously this is a singular editor specific to Wipeout (which one is not anyway ?):

it allows to make quick selections such as parents of some polygons (there are 22 tracks nearing each ~10K polygons)
relieves from the error-prone processes of selecting them manually or selecting incorrect groups, some visual cues too
creates part, objects and groups for making human-friendly entities (see bottom right)
resulting in simpler and more efficient hierarchy (i.e. a billboard will be a mesh with two sub-meshes/material)

The correctness and creation of entities works well, saving as well since it's done by Unity :p

What's next:

next big step is to re-write behaviors and try take advantage of animation window in Unity instead of using code (e.g. billboard editor using mouse) :bomb
ease the process of assigning behaviors to entities (using mouse)
build a complete scene with it
write a test scene button
finalize the whole thing for export (remember, there shouldn't be any copyrighted Wipeout material in the project)
build scenes for all tracks
also it will support Wipeout and Wipeout XL later in the process, I'll just have to generalize the 'game content class'

As usual it's Unity-related stuff that took most of the time, but submarine crew is hard at work and soon we'll surface ! :naughty

:pirate :pirate :pirate

- - - Updated - - -


Ok, just please, please don't make this a DX12 exclusive game :lol

I will do a Linux build myself but not for Mac since I don't have any.

[80T]

Hurray! Progress it's being made! I don't understand 99% of what you're saying, but progress is progress nonetheless :g Let the Wip3Out submarine surface :sonar
I'll take it that there will also be a Windows build ?

[aybe]

Of course there will be as this is my main machine :)

Trivia: so basically I'm working on a tool to build levels using mouse and Unity interface instead of code, and it's generic enough that it should work on WO/WOXL.

But these last days I'm struggling on a problem I thought relatively easy to solve but is tough to crack : performance issues for the editor :blarg:blarg:blarg.

I first made it simple by gathering all objects I need when a user performs an action but it starts slowing as soon as a few dozen objects are added. The feature I've added to automatically select and rename a newly object stops working because of that ... (this is a must because I am going to have do that hundreds of times per level so everything I can use to ease the process I'm sold)

And since Unity call some parts many times per seconds it is choking...

Roughly I maintain a list of ~10000 polygons (entire scenery) but searching the scene for these is slow, I needed a mechanism to listen on objects being added/removed -> ok, low-level Unity crap solved. But now there are other lists as well, referenced and non-referenced objects, and queries between those simply slow everything down to the point that it freezes.

It is really important that the editor helps me as much as possible for very repetitive tasks such as those, I also would like to do A/B comparisons with resulting levels and I need the same concepts...

I guess I will tackle it over the next days, hopefully.

So these days, quite some work on it but hasn't really paid off :evil

And hot temperatures these days don't help much :bomb:bomb:bomb

[80T]

Itsy bitsy bump.

[aybe]

I was busy the whole month fixing house's kitchen/bathroom/painting etc ... :mr-t

Just started working again on it tonight, need a refresh ... hmm I guess I have to fix this perf. issue so I can finish the scene editor and so on. :cowboy

In short, I'm back !!!

[TheXTR09]

Ayy, nice to see the project being back on track. Good luck at it mate!

[aybe]

Thanks !!

After a few days of accustoming I'm slowly getting back on track. I finally fixed all these performance problems in the scene editor, it was much more subtle and longer than I expected but I think we're there now.

Basically as I would add a few objects things were being exponentially longer (stats/validation) because Unity goes berserk calling methods many many times per seconds resulting in freezes ...

I still feel this was the right choice: visual cues and validation for building levels, it would not have made much sense to manually find a missing needle in the haystack of polygons that are each level.

I will probably be finished mid-week with this chapter and be able to advance in the project :)

10429

[aybe]

A quick ping to let you know I'm still working hard on the project, it's definitely not dead :pirate, it's just that it takes way longer than what I thought and had other things to do.

In the meantime I've been looking at Wipeout 64 and it's pretty much like WOXL, and I will add it to the project.

You can see what's been discovered if you have access to this thread : http://www.wipeoutzone.com/forum/showthread.php?8897-Extracting-WipEout-3D-models-(large-images-allowed)

Cheers !

[80T]

Happy NEW Wip3Out YEAR ! ! ! ! !
:hyper:hyper:hyper:hyper:hyper:hyper:hyper:hyper:h yper:hyper

[aybe]

Hi bro, happy new year, I know it's almost 3 months late but it's never too late :) !

(I still work on the project, it's just that there's so much to do and days are not long enough)

[80T]

There's no rush, just checkin' in to see what's the weather like :))

[aybe]

Here's what's the weather like in here :




new, unified asset system, supports Wipeout, Wipeout XL and now Wipeout 64 :)
(WO3 is out for now but it'll come back)
abstraction of game data, no more version-specific code
level editor ... basically I will create the levels myself by dragging and dropping track, scenery, add behaviors etc
(i.e. the rescue ship must be split into different meshes e.g. body, headlights which need their own material)
and while this is a simple concept, implementing it is not that simple, i.e. I'm getting to it but it takes time
this is the most time consuming as Unity is full of bugs, traps and shitty documentation (basic stuff is ok but I need more)
and remember, project shall not contain copyrighted stuff, easier said than done because Unity is a moron (getting to it)
lots of code written ...
some news, see below


Reverse engineering of WOXL problematic tracks such as Gare d'Europa where it was all checkerboard:

(before)
10618

(now)
10617

Wipeout 64 support:

10622

Retro-rendering effect:

(no effect)
10619

(dithered 8-bit MS-DOS style)
10620

(dithered 16-bit PSX style)
10621

also, regarding our old discussion, there might be a possibility to have real and accurate shadows one day,
stripping out the ones baked in by the designers, but only when I'm an really advanced state in the level editor

also I think I can even pull out smoother transitions in the track (see Gare d'Europa for instance (https://archive.org/details/PSX_Longplay_243_Wipeout_2097)),
we didn't pay attention back then but now it's pretty obvious, i.e. get a truly 24-bit rendering

etc etc etc ... I'm spending all my time on the asset abstraction and level editor,
goal is to finish these, pack the levels, create level behaviors, finally work on the game itself.

so, no game yet, but we're getting closer to it each day !
:P

[80T]

Dang, that's some complex weather you have there, hope you're ok :)) Anyways I'm not planning on going anywhere any time soon so, be it 10 years or whatever, I'll still be around ;)

[aybe]

How do you like the retro-effect ?

Or do you only swear by Royale CRT :P
(actually I've started working on my own but it's too early to show anything good)

[Cipher]

It's great ;)

[aybe]

Thanks :)

[80T]

How do you like the retro-effect ?

Or do you only swear by Royale CRT :P
(actually I've started working on my own but it's too early to show anything good)

The 16bit PSX style looks quite good ;) Also you can always show some sneak peek of your own CRT wip :D
Nope, I don't swear by CRT Royale :)) there's much left to be desired with it too. There's always room for improvement.
Then again, I'm only rockin' a 1080p TN... so, there's that... There's only so much one can ask from a shader.

[aybe]

sorry I was kind of 'disconnected', 'no Wipeout eat/sleep/think/code'-punishment for almost an entire month, hope I'll get back on track soon :)

[aybe]

Project updated, please take a look at the 1st post in the thread !

[80T]

Cheers man for all your hard work! I don't have any more time for gaming anyways since I've switched countries... I hope BallisticNG will get better with your help ;)

[aybe]

Thanks !!! Where did you relocate ?

I hope, too

:)

[80T]

Ireland.

[aybe]

cool :):):)