Soul Calibur VI: General discussion

Does anyone know if SC6 will be at Comic Con?

Screen Shot 2018-07-17 at 7.33.36 PM.png
 
from BandaiNamcoBR:
We had a chance to chat with the producer @achilles_Okubo about his work on #SOURLCALIBURVI and he revealed a lot of cool things, like the influence of the first Dreamcast game on this new episode. Check it on our site!
We had the opportunity to talk to the producer of SOULCALIBUR VI, Motohiro Okubo, who told us about the production, the inspirations and the news of the newest game in the series.

BANDAI NAMCO: What is it like to work for the first time in one of the most famous fighting games of all time?

MOTOHIRO OKUBO: After SoulCalibur V, getting approval to develop a sequel was quite challenging, so the future of the series was uncertain. But Project Soul members, who had been working on the series for a long time, were confident that "SOULCALIBUR VI" would see the light of day. Seeing their commitment and knowing that they would put their hearts and souls to create something spectacular, I pieced together their ideas and concepts and led the effort to get approval and start development. As the team entrusted to me is extremely talented, I have confidence that we will achieve our goals. As we approach the end of the development cycle, much remains to be done in a very short time. But I really believe that when we launch the game, the passion and effort of the team will reach the players. Thinking about it motivates me a lot.

BANDAI NAMCO: In your opinion, what makes SOULCALIBUR such a success since the first game?

MOTOHIRO OKUBO: I believe there is more than one factor responsible for the success of SOULCALIBUR. First of all the fact that it's a one-on-one weapons-based fighting game. It is a game with a wide variety of very interesting weapons that give a sense of satisfaction when used. So you have the fluid motion system, characterized by 8 Way Run controls and 3D scenarios that give players freedom of movement. Finally, there is the wonderfully unique world of SOULCALIBUR, based on a fantastic historical version of the sixteenth century. I believe this is one of the central pillars of the SOULCALIBUR franchise.
During the planning and development of this new delivery, we gave much importance to the aspects mentioned above. We strive hard to deliver the best possible version of each of the central aspects of the game, but mainly to make the gameplay very satisfying and fun. We really hope that both long-time fans and newbies will have a lot of fun playing SOULCALIBUR VI.

BANDAI NAMCO: The first game, released for Dreamcast, comes to mind when we see the clothes and scenery of the new release. Is the 1998 classic the biggest inspiration for SOULCALIBUR VI?

MOTOHIRO OKUBO: Thanks for the great question! In fact, when we started developing the new title, the codename we used for the project was "LUX", that's Latin for "light." The team chose this codename because we wanted to bring back the direction of bright and colorful art from the first SOULCALIBUR to the Sega Dreamcast. The higher the brightness of the game, the more it requires the processing power of the hardware, and that was a big challenge for us. When the players say that the game is very beautiful, we were very happy because our hard work was worth it. We hope more players enjoy our work and enjoy the visuals of the game.

BANDAI NAMCO: Since you also produced TEKKEN 7, will we see some elements inspired by the game in SOULCALIBUR VI?

MOTOHIRO OKUBO: Yes, during the creation of SOULCALIBUR VI, we definitely use many of the lessons and experience gained during the development of TEKKEN 7. For example, using Unreal Engine 4 as the engine of the game and technical knowledge of the development of the game on Steam , which is still not a common practice in our company. But more importantly, since I had the experience of producing TEKKEN 7, which is a melee game, I put a lot of emphasis on wanting to make sure that the weapons-based fighting style of SOULCALIBUR VI passed a different feeling, instead of just being a fighting game with guns as a later addition.

BANDAI NAMCO: How did the characters chosen to appear on SOULCALIBUR VI selected?

MOTOHIRO OKUBO: When we looked at the potential characters, their popularity was a good indicator, but the most important factor was that they fit the world of SOULCALIBUR. Since the series has always been of sword fights games in historical fantasy, we needed to make sure that the characters fit this scenario. Each weapon that the characters use is an integral part of the action, so we are always thinking of balancing the various types of weapons to have a more enjoyable experience.

BANDAI NAMCO: Speaking of the characters, Geralt from The Witcher is the guest character of the time. How was the process to choose from among the many good options available?

MOTOHIRO OKUBO: The guest characters have always been very important in SOULCALIBUR, so it is one of the aspects of the game that attracts the attention of the fans. So I think it's very important to surprise the fans and choose a guest character they'll be excited to play with, but it's equally important that the guest character fits into the world of SOULCALIBUR. When we discussed the candidates, Geralt's name came up and we believe he would be perfect. We immediately asked CD PROJEKT RED if they would lend us the character and they gave us tremendous support for it. So things went very well for Geralt. We are very grateful to CD PROJEKT RED, they are really incredible.

BANDAI NAMCO: Customizing characters is also an important feature of the series. How did you develop the SOULCALIBUR VI system and what are your biggest news?

MOTOHIRO OKUBO: Character creation quickly became one of the core aspects of the series, so it's very important to us. I am happy to say that the creation of characters will be included in SOULCALIBUR VI, but I can not talk much about it at the moment. We'll give more details with the official announcement later, keep an eye.

BANDAI NAMCO: What do you think a SOULCALIBUR VI e-sports competitor needs to be a champion in the game?

MOTOHIRO OKUBO: That's a difficult question. I think I can give some tips to those who are thinking of competing. In the planning and production of SOULCALIBUR VI, our priority was to make sure that he was true to the series, that he looked and felt like a SOULCALIBUR game. There are many aspects that players need to understand to compete, such as the range of each weapon, the impasses of three options, the use of 8 Way Run (especially since players can lose by leaving the ring), etc ... I I believe that the understanding of the central systems of the game, as well as its novelties, will be the difference between victory and defeat. Also, when I talked to a famous player during E3, he told me that this new title gave players several defensive options which makes battles more fun.


BANDAI NAMCO: What new battle systems can surprise longtime fans when they play SOULCALIBUR VI for the first time?

MOTOHIRO OKUBO: There are several additions to the battle system in this new title. I'd like to talk about "Reversal Edge". With the new system, players can hold and release a button to counterattack an opponent's attack when the button is pressed with a vertical strike. When started properly, a slow-motion effect is triggered and the player has 3 options: vertical, horizontal or evasion with a pass to the side. The player and his opponent choose their action from these options and we have a deadlock. We believe that this system makes battles more exciting by giving players the opportunity to feel like a "legendary swordsman" they embodied. During the implementation of this system we knew that if done in the wrong way, it would end the balance of the game, so we ensure that Reversal Edge also leaves the player open to risk. So until the player fully understands this new system, I believe they will use Reversal Edge in moderation. Try to use the Reversal Edge when they can not end the attack sequences of the opponent.

BANDAI NAMCO: Could you leave a message for SOULCALIBUR fans in Latin America?

MOTOHIRO OKUBO: Finally we were able to reveal the release date, which will be on October 19th. While we kept our promise to fans to launch the game in 2018, it ended up getting a bit later than expected due to our commitment to deliver a new series title that we could really be proud of. And although I'm working on several titles that require me to travel around the world, I still have not had the wonderful opportunity to visit Latin America. I'm really looking forward to being able to visit and meet all of our amazing fans, so please continue to send your support and enthusiasm!
The interview doesn't have revealing news, but shows that they have not forgotten the Latin market, especially in my country.
 
Yo thanks a lot! Is it possible to pm you questions? Just to save post space?

Which engine was SC5 made with? I read somewhere SC6 is made with unreal engine 4 if I'm not mistaken.

Sure go ahead. I'm not sure how many I'll be able to answer though.

Yeah SC6 is made with UE4. No idea what the other games were made with. I'm assuming it was some inhouse engine that they used.
 
I don't remember this one. When did it happen?
SC1 is the only game where the animation looked correct. SC2 and forward, you had the exact same oddity (her hand looks extra bad in SC3, but that might be an emulation bug). It's pretty minor though. I wouldn't be surprised if Project Soul was never aware of this as it's only apparent if you're able to look at the animation in slow motion.

sc1.jpgsc2.jpgsc3.jpgsc4.jpgsc5.jpgsc6.jpg

I don't want to turn this into a technical nitpicky argument (especially since it's off-topic), but I think you got some stuff wrong in your reply to GhostTear in the previous page. Different games/engines will store data differently, but the basics for models, textures, rigs, and animations haven't really changed in decades (new types of information have been added slowly, but the basics haven't changed since the late 90's). I don't see why they would be prevented to use old assets in UE4. And also, if UE4 doesn't function the way they want, they could always just change it. There are many developers who will make major changes to the engine to fit their needs.

For the heck of it, I tried importing a random SC4 model into UE4 as test. It was surprisingly fast to import SC4 Ivy as a static mesh: (excuse the quality as I just used the first Ivy model I found on the internet which has been converted from SC4 and I suspect some data has been stripped or incorrectly converted for this model. The outfit is grey because SC4 mixes in colours based on CAS data in real-time)
IversUE4.jpg
 
SC1 is the only game where the animation looked correct. SC2 and forward, you had the exact same oddity (her hand looks extra bad in SC3, but that might be an emulation bug). It's pretty minor though. I wouldn't be surprised if Project Soul was never aware of this as it's only apparent if you're able to look at the animation in slow motion.

View attachment 49839View attachment 49840View attachment 49841View attachment 49842View attachment 49843View attachment 49844

I don't want to turn this into a technical nitpicky argument (especially since it's off-topic), but I think you got some stuff wrong in your reply to GhostTear in the previous page. Different games/engines will store data differently, but the basics for models, textures, rigs, and animations haven't really changed in decades (new types of information have been added slowly, but the basics haven't changed since the late 90's). I don't see why they would be prevented to use old assets in UE4. And also, if UE4 doesn't function the way they want, they could always just change it. There are many developers who will make major changes to the engine to fit their needs.

For the heck of it, I tried importing a random SC4 model into UE4 as test. It was surprisingly fast to import SC4 Ivy as a static mesh: (excuse the quality as I just used the first Ivy model I found on the internet which has been converted from SC4 and I suspect some data has been stripped or incorrectly converted for this model. The outfit is grey because SC4 mixes in colours based on CAS data in real-time)
View attachment 49845

Weird I never actually saw that with her BB when I played the game (or not noticeably). I'll prob load up SCII again since its the only one I have on hand atm and check it out. They're are a couple of reasons I can think of for that happening but unless I run a couple of tests with the source file I can't say for certain. I have seen these problems though, and sometimes they're also a major pain in the ass sometimes to fix. Though there are times when they are a lot simpler as well. All depends on where the translation is happening. Oh, it might be from the correction that's being done with the hand switch. I have a theory as to why that's happening but I can't really give a good answer unless I actually took a hard look at it.

As I mentioned in that post, I don't know how Project Soul works as they're are a number of workflows that can be used in game development and that what I was providing was simply an educated guess based on experience. Also I don't think you read my entire post and understood it.

When we're talking about the way models are stored, we're not talking about the core information when you export it as an FBX and OBj file that you can just drag and drop into unreal. If you notice, UE4 convert the model into a uasset file which uses a completly different format for storing data because the engine can read that data much much faster than a native object or fbx file. That is what I was talking about when mentioned models being stored different in assets. A custom inhouse engine in order to optimize assets would take in a special type of customized file (nothing really special with it) that just customizes what information is reserved in what way. In that regard, because it only has information that the engine needs in a format that the engine expects it to be used it can increase the speed at which is process the information. Since I've never looked at the source files for SC I don't know how they were stored and as I mentioned earlier it all depends on if they had them, and if the software they are using is still compatible. I have already seen a number of files back from Maya 08 crash every time on load in a version of maya 15. In any case, I was outlining potential problems since I mentioned I haven't worked in PS so I don't know what type of pipeline they use. I never said they couldn't use old assets, but it would be more work to make it work as it doesn't align with the direction that they want to go in, their not high rez (I can't tell how many polys Ivy is from the image), and the moment they create a new model, they need to redo the texture. The texture is based on the UV Coordinates and the Texture seams, both which are created after the model which is why you can't straight up reuse the old texture. And if they're going to remake all the clothing to fit the concepts, they need to adjust the bodies as well to fit with the age, concept and outfits as well. Might as well remake the stuff if you're going to try and upgrade the polycounts on them as well and give more definition and form.

Trust me. Modifying UE4 as well just for models may not have been worth it any case. Its like a modified version of C++ with Macros everywhere. Something I think the programmers have a lot more issues in which they should spend their time.

You are correct for base rigs not changing much for motion capture purposes. Yes the basics rarely changed, but when you want to start doing more advanced things you have to replace or modify the initial setup to accommodate for that or make them easier to work with. The way models are constructed to accommodate making deformations easier to accomplish and in some cases possible. I personally have not seen information on how they are doing their secondary action elements on the characters. If they are using UE4's built in physics simulation or they've created a simulated joint chain in Maya and exported the animation. Once again though it all comes down to how they have been working and how they continue to implement things. Basically, if they want to change what they are doing, they have to remake a lot of the stuff, and in many cases, it is just straight up easier to remake the assets because they will be more consistent and if a problem comes up, they will have a much better idea of how to fix it. If you've ever had to debug someone else's model, animation, rig, code, you would understand that sometimes, that's a headache you really don't want to deal with. As for the way the information is stored, once again that is all based on the teams preference. You can easily put roughness, spec and diffuse into one map if you want it to and have specific channels read in order to figure out what belongs to what. A lot of methods that are used are proprietary, and not available to be seen by people outside the company. Take a look at trying to get the actual rigs for a League of Legends Character or an Epic Games character like from Paragon. Not a rip and rerig done by a community member but a legit rig and you'll see a difference in the techniques that are used.

Btw importing Ivy as a static mesh isn't really going to help. Characters are skeletal meshes with an entire skeleton bound and weighted. The axis of which could be flipped depending on export settings. If its missing data as well, you're not seeing an accurate use case. It's easy and fast to import and load because its a single low poly static mesh that the engine is optimized to load and handle. That's like putting an rc car inside a lambo and then watching it take off. Yes the vehicle is going fast, but its not really representative of what either of us are talking about.

At the end of the day, they don't want to use old assets because they lack quality and it will be more of a pain to use. Sure they can still reuse and cherry pick things. they can still reuse old animations with a newly built base rig since all they have to do is designate which bones belong where. But it's all about maintaining a consistent level of quality that is consistent with the vision and direction of the game.

Tl;dr Nothing stops them from reusing old stuff except themselves and the rage of millions of people who want improvement.
 
Last edited:
Weird I never actually saw that with her BB when I played the game (or not noticeably).
The animation is very fast, so I bet the developers themselves never noticed bad key frames. Nobody did until recently. Thing is, they've added slo-mo effects into SC6, which exposed hidden problems, so once exposed, people went back to older games with a microscope to compare.
Here is the video that started it all:
Skip to 39:46.
And here is the old comparison post.
I have a theory as to why that's happening but I can't really give a good answer unless I actually took a hard look at it.
I think it's just key frames. It looks like in SoulCalibur games weapon meshes are not put in a child-parent relations to characters' hand bones, but are animated separately, so transitions between key frames can be wonky if key frames are not dense enough.
I'd rather get the game in a timely fashion than have Sophitia's sword in-hand for a single frame.
To my understanding fixing that one Sophitia's sword animation in question should only take 5 minutes maximum from Project Soul.
Now the most difficult visuals problem SC6 currently has that I would love Project Soul to fix is bad shoulder/armptit rigging problem that affects all characters and is quite frankly very common in game development across many games and studios:
Shoulder.jpg
 
The animation is very fast, so I bet the developers themselves never noticed bad key frames. Nobody did until recently. Thing is, they've added slo-mo effects into SC6, which exposed hidden problems, so once exposed, people went back to older games with a microscope to compare.
Here is the video that started it all:
Skip to 39:46.
And here is the old comparison post.

I think it's just key frames. It looks like in SoulCalibur games weapon meshes are not put in a child-parent relations to characters' hand bones, but are animated separately, so transitions between key frames can be wonky if key frames are not dense enough.

To my understanding fixing that one Sophitia's sword animation in question should only take 5 minutes maximum from Project Soul.
Now the most difficult visuals problem SC6 currently has that I would love Project Soul to fix is bad shoulder/armptit rigging problem that affects all characters and is quite frankly very common in game development across many games and studios:
View attachment 49847

Yeah those shoulder problems are ridiculous to try and solve. I've seen a few fixes that looks fantastic but they require a bunch of alterations to the underlying rig. Particularly with the shoulder bones and the which joints are weighted. Honestly I don't expect them to fix it since it would involve altering every animation and rig for all the characters that are done. They could probably attempt to alter only the problem anim's but I don't want to think about what problems that might bring up.

As for the parenting, SC6 might be a bit different because weapons are going to pretty much always be parented at the base level for the character for sure. It's quite easy to do quick reparenting to different hands in UE4 in order to maintain consistency and something I'd probably expect them to move towards since it offers some pretty good results. Expecially because of CAS where the hand positions and positions of where the sword is will be variable. It would be much easier for them and what I expect them to do, is to animate the sword as part of the child of the character, and since Sophie's primarily right handed they might have not bothered doing the socket on the left hand and manually move the weapon, countering the movement on the right arm. Or maybe its just a trailing frame between the end of the attack back to her idle, and the interpolation is what causes it to drift off her hand. Or they could do something else. Who knows. All of this is speculation in any case. If there's one thing I've learned. All the simple things takes a lot more time, than you think and all the complex stuff ends up being a lot more simpler than you think but they all cause the same amount of headache depending on how prepared you were to handle it. I will say this, if it is a keyframe in their animation program, they would know about it, because I don't know a single animator that doesn't check their work 30 times at normal speed and at slow mo and frame by frame before showing it to their head and then submitting it. So unless after Soul Calibur, they straight up just took the animation and did some quick ends and then threw it in, never checking it for any other game, there has got to be something else happening. I also don't want to believe they were straight up that incompetent to have that same issue for 5 games.
 
Yeah those shoulder problems are ridiculous to try and solve. I've seen a few fixes that looks fantastic but they require a bunch of alterations to the underlying rig. Particularly with the shoulder bones and the which joints are weighted. Honestly I don't expect them to fix it since it would involve altering every animation and rig for all the characters that are done. They could probably attempt to alter only the problem anim's but I don't want to think about what problems that might bring up.
It took me months to figure out on a model I was messing with learning modeling/rigging. From that point I can't help but always look with extra attention at shoulders and armpits in games =).
I think the problem can be fixed without altering any animations, just by adding extra bones, touching up weight painting accordingly and making those new bones to rotate outwards automatically depending on how hight an arm gets lifted. That would be the cheapest way, but it would still take a significant effort, so they probably won't do it.
 
Back
Top Bottom