By replacing the aging under-the-hood engine of the Shenmue I & II games with Unreal Engine 4, the project team aims to take advantage of modern rendering features to really show off the world of Shenmue as never before. Not only that, but the project will offer players two ways to enjoy the games: Dragon Edition, which will have gameplay that is entirely faithful to the originals and Phoenix Edition, which will be a "full UE4 reimagining" that modernizes the game with new features.
Can't wait to find out more? Segalacious has come to the rescue in his latest video below with tantalizing details and information as he breaks down exactly what the Dragon and Phoenix Collection is about, along with a Q&A session with the members of the modding team working on this ambitious task.
Shenmue 1&2 In Unreal 4 | Dragon and Phoenix Collection Modder Q&A, by Segalacious
Transcription:
The original Shenmue games are well known for being technically and visually impressive. Chasing the full potential of what Shenmue could look like using modern technology has long been an obsession for fans. Kid Nocon famously captivated the internet with his fan-made Shenmue HD remaster concept in 2014 that contributed to him being hired to work on Shenmue III. Videogame co-development studio d3t would assist Sega in creating an enhanced re-release of Shenmue I and II that was ported to the Xbox One, PlayStation 4, and Windows that both impressed and frustrated fans. As good as this port was it still missed the opportunity for greatness by neglecting to give the Shenmue games a thorough HD treatment in terms of cinematic cut scenes, sound quality, and other various blemishes. Soon after this release Digital Foundry broke a story about Sega's abandoned attempt at a ground up remake of the Shenmue games and fans were tantalized with the question “What if?” What if Shenmue I and II could be cleaned up to their fullest potential? Even if a true remake isn't financially possible, just how well could the original assets clean up on a modern engine like Unreal Engine 4?
Enter Shenmue Dragon and Phoenix Collection. An ambitious undertaking in which a small group of devoted and talented Shenmue fans are, as we speak, engaged in the goal of completing an engine replacement mod for the 2018 PC Shenmue ports that replaces the original engine with Unreal Engine 4. In this video we will be breaking down what exactly this project is, showing some images of what has been accomplished so far, and sharing exclusive information from a private online Q&A session between Segalacious and a member of the Dragon & Phoenix Collection team. We’ve even got a small selection of previously unreleased images provided by the modders themselves specifically for this video, one of which you've already seen in the thumbnail. Friendmues, you won't want to miss out on this one. Let's get into it!
So what is specifically meant by the name Dragon and Phoenix Collection? Well, the goal is to provide for two ways to play the original Shenmue titles. Dragon Edition is intended to be a true one-to-one port of Shenmue I and II to Unreal Engine 4, preserving as much of the original experience as possible, even the aspects of the game that are considered to be dated. Not much will be offered in terms of new features as this version is focused on polishing the original games to their utmost potential. On the other hand, Phoenix Edition will be a full Unreal Engine 4 reimagining of the original games. The modding team stresses that all the core assets and gameplay present in Dragon Edition will be present in Phoenix Edition, along with new features focused on modernizing the games. This could include anything from importing a stamina system akin to Shenmue III’s complete with consumable items (but implemented in a less intrusive way) all the way to the addition of new cassette tapes at the Tomato store.
With no definitive release date on the horizon, I wanted to see if I could get more information directly from the team behind the Dragon and Phoenix collection. Fortunately I was able to get in touch with Dewey who helped to provide screen shots and video of the project.
When I asked how the team first came to the conclusion that this project was viable, he responded:
Dewey: When the re-releases came out in 2018, I think all of us wanted to mod the game to various degrees. As LemonHaze and others got more and more of the game reverse engineered, it became apparent that the original engine was just a bit of a mess. That's not to say anything bad about it, it's just a product of its time; point is, it became obvious that replacing the engine was the best course of action if we wanted to actually do anything with these games. From there, it just became a matter of understanding and implementing the various game systems in Unreal Engine 4, which isn't quite the herculean task it seems, since the original engine is fairly modular in a sense anyway, so each aspect of the game can be implemented and added as it’s ready, while we simultaneously work on improvements both visually and in terms of gameplay... [For example]... the 30fps lock... I think it's reasonable to say that a majority of people who played the re-releases would've preferred to see it at 60fps...This isn't QUITE as impossible as everyone has been led to believe, but the steps you'd have to take to run the game at 60fps in the original engine with no major issues are just a bit “hacky”.
Q: What exciting discoveries can you share that have been made during this project?
A: There are a fair amount of things we’ve found, too numerous to list all of them here, but we have found things such as an alternative beta layout for the You Arcade which is different from both the final game and What's Shenmue. If memory serves, you can see this layout in some early screenshots. We’ve also found a few interesting textures, like a set of UI textures that may date back all the way to the Saturn era of development, given the resolution of the texture sheet, among other clues. On top of this, we’ve also found things such as some cut fighting moves… sort of. We’re only human, so we don’t know all the moves by heart; but we can see there are more fighting move animations than there are moves, something of a game of hide-and-seek at this point for those.
Note from Switch: we hope to share some of the intriguing discoveries made by the project team (with their blessing!) as they have been working on the project, in future posts here on the blog.
Q: What are your biggest challenges with completing this project?
A: Honestly, just getting it done. There's only a handful of us, so it leaves a lot of work to be done, but we've got a pretty good workflow. This is more of a passion project than anything else for most (if not all) of us, so it's really a matter of figuring out which tasks are suited to which person, and then getting the task done.
Q: How can the community help?
A: The community can help in a number of ways. If you've got some sort of experience that you think would be useful to us, we'd be glad to hear from you; you can find us on the Shenmue Reverse Engineering Discord, just reach out and we'll see what we can do. Beyond that, just the general spread of awareness is helpful, as have been some of the suggestions we've already received. We've largely kept ourselves motivated just by seeing the improvement with every little change we've made; and seeing the immediately positive response from the Shenmue community to our announcement and first screenshots has really reinvigorated our resolve to make sure this is the best product it can be. At the end of the day though, we're all just fans of Shenmue, and we all have our different takes on the series and each of the games...
Q: What specific type of activity comprises the workload of the engine conversion? What specific skill are you looking for in terms of bringing in additional hands?
A: We could use people from all skill paths, to be honest. We could use a few artists, to help with things like creating or editing normal maps, or some mesh editing for places where things may not be optimal in a modern engine (for example, last night I found a spot in Dobuita where you can fall through the world; there was an invisible wall there in the original game that doesn't exist here… at least not yet). We could also use programmers, just to take some of the workload off LemonHaze, since he's doing the majority of the coding for the project. I wouldn’t say he’s our only programmer, but he’s the only one who really has a complete understanding of the original engine.
Q: Can you share what prospect of this project excites you the most?
A: Personally, I'm excited to see some aspects of the game fully realized. For example, cutscenes; I can't say for certain that this WILL happen, but when we get around to working more on Phoenix Edition, we'll have the opportunity to upgrade some of the audio/visual aspects. So things like an improved soundtrack, or improved cameras. These are what I'm looking forward to. Seeing cutscenes that actually look like a movie, as opposed to just moving the game camera around. I think of it like the difference in cutscenes from Metal Gear Solid 1 to Metal Gear Solid 4. In MGS1, cutscenes basically just moved the game camera around; in MGS4, however, cutscene cameras functioned more like movie cameras, you might see the camera "lens" actually change focus in a scene. With Unreal Engine 4, we can do that, whereas with the original engine you just... can't. This is just one of those small upgrades that can really add a lot to a scene, and has me really excited to see just what we can do with it.
Q: Are you able to port all assets of Shenmue to Unreal Engine 4? Even the Virtua Fighter combat engine?
A: More or less, yeah. We might have to make some artistic changes for some things like FREE battle, just as a consequence of being a totally new engine; but as it stands we've got animation and data for the entire moveset (which is where the cut moves mentioned before live; they’re just thrown in here with everything else). So physics and movement might wind up a bit different, but all the moves should be there, along with the various visual effects that come from training/leveling up moves. At least that's the current intent, and we see little reason why we shouldn't be able to achieve that.
*At this point LemonHaze chimed in:
"Both games mostly use the same format and the ease with which we can recreate a lot of the original mechanics is basically thanks to switching engines. Interaction in and of itself is something that used to be 'optimized', this is why you see heavy optimization in Shenmue, and VF, and Space Harrier / Hang On. They all rely on tight timings due to hardware constraints and common design practices at the time, and partially because Yu Suzuki is a mechanical engineer first and foremost, this is evident in most, if not all, of his games, where timing and fluidity is a big part of the game itself. Not to mention, back then that's just how games were made; so most of the original code for things like interaction, whether it's UI stuff like Yes/No/Buy/Sell/Cancel or something gameplay related like an automatic door.. all of those implementations in the original engine are archaic in comparison to modern game development; it was great back in the day, but today it's just inefficient. What we can do in one function, the original game has a series of functions to handle, because it had to be optimized enough to run on the original hardware."
Friendmues, with such exciting prospects on the horizon, it has truly never been better to be a Shenmue fan. If you would like to help out LemonHaze and his team comprised of Dewey, DaviDokuro, and Nicolas_D, among others, please see the link the description below to get in touch with them. Once again, let's support this dream until it becomes a reality. I've been your humble friendmue Segalacious urging you to keep fighting the good fight, the virtua fight! Until next time.
-Segalacious
Related Links
- Forum thread on Shenmue Dojo
- Invitation to join the Shenmue Reverse Engineers Discord.
About the Author
Segalacious is the co-founder of the Virtua Bros YouTube channel. He is best known for his Virtua Fighter Lore video series, Shenmue Video essays, Top Ten lists, and Game Creator documentary retrospectives. Follow him on twitter @Virtua_Bros and check out his video “Why Shenmue was Always Good”.
Check out his insightful guest posts for the blog here:
Fantastic job on this transcription artical switch!
ReplyDeleteThank you for creating the video! It's a pleasure to share your finely-crafted content.
Delete