Otakon 2022 Interview: Mikami Masafumi and Kiral Poon (Gugenka, Inc.)

This interview was conducted at Otakon 2022 in Washington DC. Gugenka, Inc. is a company that describes itself as “Sales of digital goods and technical research on Japanese animation using advanced contents such as VR/AR/MR.” Mikami Masafumi is the CEO, and Kiral Poon is the CTO. 

How did you get into the business of 3D, VR, AR—these sorts of subjects?

Poon: Do you mean academic side, or…?

Anything, really. If there’s an academic element, then definitely talk about it. If it’s sort of an amateur hobby, or…just what inspired you to take this up…

Mikami: We were originally making AR stuff for movie promotions, and we slowly moved that content into VR. That became for VR anime promotion, where we promote anime with our VR abilities.

Poon: One of the examples would be the Sailor Moon VR that we built. That is also, like, a movie for Sailor Moon Crystal promotion, and then we did the VR game for the event. And in the end now it’s on VIVEPORT for free, but it was originally promotional content.

I first became familiar with Gugenka through the HoloModels app.

Poon: Really?!

Yeah. I found out about it from a reader, actually.

Poon: When did you hear about it?

A couple of years ago, I think?

So what was the origin of the decision to make virtual AR models? Was the plan from the start to work with established properties?

Mikami: The way it started off was, you know how we have anime figures in real life, right? What if we could make digital versions of anime figures?

In physical figures, it’s probably not a real surprise to see on your figure rack a Dragon Ball character and a Re:ZERO characters side by side. But then in VR contents, it was an uphill battle to get the understanding to have two characters of different IPs sitting on the same app.

Poon: I’d like to actually convey my two cents on that. Japanese companies are really, really, really restrictive on the IP stuff. The more I work with them, the more I think that’s so crazy about it. Masafumi back in the day, he tried very hard. One problem is, how can you line up two different IPs together without any problem with the IP company. Cause they usually want to have their own world, right? When it’s in the real world, it’s just figures, so they don’t care about it. But in digital, they usually have control with their applications. So Masafumi was actually the producer. He is the one who actually produced it and persuaded those companies to do it together. The reason he can do that is because this company has a long history with these companies and these movies. He already knew these people, and that’s why he did it. It is not easy at all, in my opinion.

I’ve heard similar things about the game Super Smash Bros. When it comes to the director, Sakurai, it’s only because he has such a strong reputation in his industry that he’s able to convince all these different…

Poon: Yeah, exactly. In Japan, when you work, it’s not only about ability, but also about how people trust you. So that’s the case.

I also want to point out why this app exists, which is that there’s limited space around the world. We usually line up a lot of figures—I buy my own figures too—they take up a lot of space. In AR applications, you stick them into the space, and you can display them anytime. It’s just like an RPG item box, where you just bring it up anytime—exactly like that. That’s really convenient, and you can do more than with actual figures. You can move the pose and change the face. With real figures, you have to change the parts, and there’ll be lines on the face you don’t want. But with digital figures, you don’t have that problem. So that’s one of the main reason I can tell you why we created them.

You actually anticipated my next question!

[laughter]

So moving on, Gugenka has collaborated with virtual stars of all kinds. You have Vocaloids like Hatsune Miku. You have Virtual Youtubers like Kizuna A.I., Tokino Sora—and I remember that when I got the Gugenka HoloModels app, there was Shinonome Megu

Poon: You did a good job. People don’t even know that one.

So what do you think about the fact that people are embracing virtual characters, and have you noticed any changes in the degree to which people are embracing them?

Mikami: Right now, everyone has their own different characteristics, and everyone would probably be able to express themselves differently in different bodies. I believe that our app would enable everyone to express themselves in their own body of their choice.

Around the world, the COVID-19 pandemic has brought the concept of virtual spaces to the forefront, as friends, families, and businesses try to gather online through various chat programs but also in creating virtual space. How has the pandemic affected your approach to virtual spaces? Has it cast a new light on them?

Poon: So with virtual space, during the pandemic, we have actually been trying to increase our opportunities with all these restrictive companies. In my point of view, usually Japanese companies tend not to do remote work and exhibitions lots online. But because of the pandemic, they start thinking about it, and they approach us. Of course, sometimes we approach those companies—“Hey, do you want to do it?” Usually, it’s the case that they will ask us about doing online exhibitions. Before, there were, like, no exhibitions, but because of the pandemic, we have those events hosted by Gugenka.

For example, the closest one would be the MF Bunko J one that we just did with Kadokawa—the light novel one. It’s a really popular real-world event in Japan. And there’s live concerts, like the Sanrio Virtual Fes. We tried to make it realistic, and we have a very good reputation on that event with VRChat. And of course, we also connect with all these partners from Japan—VRCast and other platforms. And I connect with all others across the world and have partnerships with them, like VRChat. After that, people started realizing what we’re trying to do to make people closer in distance, and display things in HoloModels, and also create your character through MakeAvatar.

You’ve worked on many events, and I know it’s hard to pick a favorite, I guess, but I’ll ask anyway: Do have any events you’re surprised that you managed to work on?

Poon: I guess I’ll let Masafumi answer first, and then I’ll answer.

Makami: I would say Kitty-chan’s Sanrio Virtual Fes is probably my favorite. I can take pride that I participated in it. The reason is that the people who are enjoying the real stuff Sanrio produces, as well as the artists and VTubers—they all come together in one event, so I thought it was a good event.

Poon: I’ll add two cents about the Sanrio virtual event because there’s not many people who know about the event in the US. After I did the panel, people were all surprised. This is a real diverse event that includes actual artists like AKB48—famous, real artists like HoneyWorks. There were also virtual artists like Kizuna A.I. and other VTuber characters. And then there’s also the system that we created so that we can bring this all in, and with synchronization with the timing we can make it feel like a live event. 

What that means is, what if you join late but your friends have been in for five minutes? In a usual game, well, you start from zero, and you play it, right? But in real life, you’ll be seeing things five minutes late. What you’re seeing with friends is the same thing, and when you wave, and when you sing, you’re doing the same thing. This is what makes a huge difference with the Sanrio Virtual Fes event. We created a synchronization system to make sure everyone had the same experience, same timing to all these artists. And it’s a huge event because there’s more than a hundred songs. I didn’t sleep for two to three days just to check the songs, and it was crazy. I wish one day we could join us for the second one. So that’s the Sanrio Virtual Fes.

That actually makes me think: Sometimes, due to the pandemic and the increasing use of online spaces, people want to get together for a karaoke session, right? But due to the differences in synchronization, it can cause problems for people who want to sing together. Is there a possibility of creating a space like that for regular users?

Poon: So there’s a lot of problems in terms of people’s preferences. During the event, there’s a separate instance we created to adapt for each person. For example, if your computer is weak, but you still want to enjoy the event, there’s something called shitei [appointed] instance, which means you are pointed to that instance, and we make sure you can have the best experience in that instance. But because it’s a social VR, if you want to see the group, there’s something called jiyuu [free] instance. But if it crashes, or you just want to enjoy the experience with your weak computer, you can still go to the appointed instance. 

But also, furthermore, for the paid content, there’s outfits using the MakeAvatar app. It’s like when you dress up as Mickey—or Marvel characters because Disney owns Marvel—and you go to Disneyland. So it’s the same thing as when you go to Sanrio Puroland; you dress as Mochipoly, as they call it. It’s really cute, and you can buy the hat from Kitty, or Keroppi. 

There are two reasons for that. First, is because of the story. It’s best delivered that way. It feels like part of them, and it’s real good. When you look at pictures, it’s really nice. Second, is the performance. We want weak computers to also get in easily. So we think about all that and plan the event that way.

I know a lot of people around my age became anime fans in part or in whole due to Slayers

Poon: Ooh!

So I was pleasantly surprised to see both a HoloModel of Lina Inverse—I saw the author, Kanzaki, constantly promoting it—

Poon: Man, I’m surprised you know Slayers, and that we did that?

And you created the Slayers 3D Live event. Were there any fun, creative challenges involved with working with this property?

Poon: Actually, Masafumi was talking about Sanrio, but my favorite event was Slayers. The reason for that is the equal system that I was trying to build. I’m actually the director of the concert for the live event in Tokorozawa Sakura Town. 

So we actually have the live event, and then the sequences. We have HoloModels of Slayers, and usually it would be a waste that you can’t use them on other games. It’s really expensive to build a really high-quality CG model. We have the models, we have the live event in Sakura Town in real life, so it feels like Lina Inverse is there, and she does the song. Then, we host this event online in NeoChat so everyone can join. And then in the end, it’s sold on blu-ray. So this whole flow is the equal system that we try to tell those IP companies that we can balance promotion and also maybe earning a little money for the company. That’s why it was my favorite event, because all the systems of the Sanrio event are based on the Slayers event. 

I’m surprised. Even though I didn’t do a lot of promotion for it, I’m surprised that you and—actually, one more person at the panel, he said he knew it, and he actually joined the event. I’m actually touched because I want North America—like myself from Canada, and there’s also the US—I want them to join us. But there’s limited resources, and there’s a problem that we didn’t have enough English translation at the time. So in the future, I’m trying to push everyone to do more English support. Good question. I want to talk more about Slayers!

Kiral, one of your interests, as written in the Otakon guidebook, is creating technology to help make life easier for animators. There’s an ongoing issue about animators being overworked and underpaid?  Do you think your work can help deal with this issue?

Poon: Do you know Dragon Ball Super: Super Hero? That is one of the good examples of using good 3DCG in movies that doesn’t feel weird. Because sometimes it feels weird in 3DCG. There’s one good example, that you can have an efficient way of moving, and then you can have more stories to deliver to people. 

So that is great, and it’s efficient, but to your question, can it solve the overtime and overwork problem? In some ways, but it comes down to the culture in Japan. Let’s say you need less people for the same project, right? Well, you just hire less people for the same work—there’s a chance like that. But it depends on the company. They could be, like, “Oh, there’s less work now, so now everyone doesn’t have to be doing overtime, and they can go home, right? So there’s two approaches, and it comes down to culture, I think, partially. But the reason I want to focus on technology and I keep improving that, like with Slayers, is to try io make it more toon-like with cel shaders—to make it more efficient. But can that solve the core problem is the question.

Are there any Gugenka products or services that get less attention that you think people should know more about?

Mikami: In regards to what I want people to know more about, it’s basically everything: HoloModels, MakeAvatar, virtual events. But I kind of believe there are two major points that I think we need to focus on to have a wider number of people know about us and these events. 

One is that there is a time difference between Japan and the US. While we were able to get past the problem of distance because we could bring people together from vastly different parts of the world, we can’t really get past the problem of time at the moment. So, I believe what we want to do in the future is maybe have different events for different time zones to address the time issue.

The other thing I want to focus on is basically localization of the English versions for these apps that we have. We need to be able to disseminate this information about the apps and events, and it’s not as easy as Google or word-for-word translation because sometimes they miss the mark or have different words we would like to use. Also, the thing about media is basically that over here in the US, it’s not like we have a strong understanding of where the media pipelines are. So I believe what would be best is if we could get to know the media people here, and how to best communicate with our fans so that we could disseminate the information to everyone interested in an effective manner.

Poon: I actually have the same feeling. When we do the panels, people actually come to me and say, “I didn’t know Gugenka before—MakeAvatar, HoloModels—but now I’m going be a core fan and support it.” So I’m touched, and also really happy that people enjoy our service. 

Like, with MakeAvatar—I don’t think you even know this much yet—but this allows us to use different parameters. The tech part is really hard—I tried so hard on this tech from January—and finally, you can use different kinds of morphs to different kinds of faces. The idea is, usually anime characters are a little bit more the shounen young type, but I want elderly people to also be cool. Even if you’re old, you should be really handsome because you do your own thing. You could be chubby, you could be muscular—allowing the diversity is human, and it should be possible in your own character. That’s why we have all these features.

Of course, we have the SD character, that is, the small one. The main reason we built this one is because we support Quest, cell phone, and browser, so we want the workload to be lower. But we have new, high-quality MakeAvatar that allows you to change clothing and stuff. And the business model would be to buy the cosplay—the costume—and the IP company gets paid, and the user is happy. So you can export to VRChat, VirtualCast, to see online on Vroid Hub, or browser games. 

We think in the future maybe we can do more promotion in the US and more people could take advantage of this. You could become a VTuber easily. I could host a panel and just teach you to use the app, and you could be a VTuber for free. And we have the face-tracking support in this, which means it can detect your muscles and move. In normal characters, it only supports simple expressions like A-I-U-E-O and smlle, angry, crying—that’s it. But we support really detailed motions on this.

Thank you!

Real vs. Perfect: The Two Opposing Idol Values

1983’s Creamy Mami was the first idol anime, and it made an idol out of Mami’s voice actor as well. Watching her videos from back then, a 15-year-old Ohta Takako comes across as awkward and unaccustomed to the spotlight, even in “Love Sarigenaku” above, her most “grown-up” song. Compared to many of the slickly produced pop hits of later years, Ohta can come across as almost unprofessional, but that’s exactly where her appeal lies. When it comes to Japanese idols, there are two general directions: “unrefined and real” or “polished and perfect.”

When comparing the Japanese idol juggernaut AKB48 to the K-Pop sensation Girls’ Generation (who have been enormously popular in Japan), the latter visually comes across as a much more “professional-looking” group. While calling them idolsTheir dance and choreography are on point, and their music videos make them look like a million bucks. But while the girls of AKB48 have a kind of awkwardness about them, and many aren’t the greatest singers, there’s a sense of them “trying their best,” and this is exactly what the fans want. In other words, perfection isn’t necessarily desired. It can be, but that strain of inexperience and perseverance is just as strong.

These dual forces can be seen in idol anime in spades. In Love Live! School Idol Project, the main characters are the ragtag group μ’s (pronounced “Muse”), and the defending champions are the practically-professional A-RISE, who come from the richest high school in Akihabara. In Aikatsu!, Hoshimiya Ichigo is shown as having some kind of natural spark of genuineness that contrasts her from the seemingly unassailable Kanzaki Mizuki. And in Macross Frontier, the main love triangle features, as seen above, the humble waitress Ranka Lee (right) vs. the sultry Sheryl Nome (left). In every case, what causes the “small fry” to ascend isn’t that they transform into polished and perfect idols, but that even as they improve, that unrefined and authentic quality shines through. Perhaps it says something that the main heroines of these shows tend to lean that way as well.

And yet, as touched on briefly in the beginning, voice actors who play idols in anime actually end up being idols themselves. When the girls of Love Live! hold live concerts their flaws come out, but that’s part of the appeal of seeing them in person. When watching the characters in the anime or in music videos, that imperfection doesn’t come across in the performances so much as in the dialogue and supporting materials. A similar phenomenon exists all the way back with Creamy Mami. She comes across as much more “polished” than Ohta Takako does, yet they share the same voice.

An interesting case of the strange interaction with the 2D vs. 3D and real vs. perfect contrasts are those that toe the line, like Hatsune Miku or virtual youtubers. With Miku, her limitations—the fact that her voice sounds robotic—is considered part of her appeal. With virtual youtubers, the fact that there’s a person performing behind the character is much more obvious, and the idea that they start to break down or break character is what lends a sense of “realness.”

In this regard, California-born Japanese idol Sally Amaki is especially interesting. A member of 22/7, an “anime-style characters” idol group in the vein of Love Live!, she plays the bilingual character Fujima Sakura while bringing along her own fans as Sally. Not only does she perform the virtual youtuber role as Sakura, but her native English fluency brings an interesting dynamic that highlights a sense of “realness,” especially for English-speaking fans. Not only is there often a contrast between Sally’s “cute, practiced idol” voice and her Californian mannerisms when switching between Japanese and English, but she’ll mention something that only someone growing up in the US would know off the cuff. This lets American fans connect with her sense of authenticity in ways that they might not have been able to in the past.

In the end, “real vs. perfect” is not a true dichotomy by any means, and every idol/idol group approaches that divide in different ways. Whether you’re an idol fan or not, which do you prefer?

We Don’t Need No Stinkin’ Holograms (At Least Not All the Time): Voca Nico Night

This post was sponsored by Johnny Trovato. If you’re interested in submitting topics for the blog, or just like my writing and want to be a patron of Ogiue Maniax, check out my Patreon.

As part of my Patreon rewards, I was recently asked to write about “Voca Nico Night,” a series of concerts celebrating Vocaloid music broadcasted over the Japanese streaming site Nico Nico Douga. This topic is somewhat outside of my normal purview, as I will admit to not being especially knowledgeable about Vocaloids or the scene surrounding them (I enjoy the music and recognize the characters but would not call myself a devoted Vocaloid fan), but as I watched recordings of Voca Nico Night I found that the performances seen at Voca Nico Night can be viewed as a reminder that, at the end of the day, talented human creators are an integral part of the success of Vocaloids and the music that is created using them.

The story of how virtual idol Hatsune Miku became an international sensation is the kind of tale that can cause some to proclaim the green-haired, leek-twirling, robot-voiced “singer” as the sign of a glorious future, and others to describe her as the death of music. For years the music company Yamaha had put out a voice synthesis program, but with some improvements to the software and the use of a cute girl in chic, futuristic clothing, things changed. Suddenly it was embraced by composers, musicians, music enthusiasts, artists, and writers both amateur and professional, and you can find countless examples of Vocaloid music, art, and more on sites such as Youtube, Pixiv, and Nico Nico Douga.

All of this has culminated in Vocaloid live concerts, where holographic projections of the Vocaloids sing and dance as live crowds cheer them on. This presents a fascinating contrast in the “life” of the Vocaloid. At the same time that Miku and the others become more “real”, their artificiality becomes even more pronounced. After all, even if she were to be brought into reality, Miku still has the appearance of an anime character, and that stylization and abstraction cannot be divorced from her even they’re re-drawn and individualized by various creators.

This can be viewed as a strength, as Miku’s self, constructed from the effort and desires of millions, makes no illusions about the synthesized elements of her music (as opposed to the auto-tuning that regularly occurs in popular music these days), but for those who believe that music comes from the soul, that way of thinking is certainly difficult to swallow. While I’m more of the former opinion, the latter I think is something worth keeping in mind.

This is where I think Voca Nico Night really shines, because without the expensive holograms, without the explicit desire to see Hatsune Miku or Megurine Luka or the Kagamine Twins, what’s left are the DJs, the musicians, the performers at center-stage, engaging with their audience both online and in the flesh. You can still hear the distinctly robotic tones of the Vocaloids, but their images are not taking all of the attention. Instead, the Vocaloids’ role as tools for producing music comes to the forefront, and while talking about the “functionality” of an actual musician would be crass and unfair, the origins of the Vocaloids means that they can be as “useful” or as “soulful” as one needs them to be. Voca Nico Night exists somewhat opposite the live concerts that get all of the attention, and while I hesitate to use a “yin-yang metaphor,” I do think that having both is for the better.

In many ways it reminded me of the times I’ve attended chiptunes concerts (another scene where I have interest but cannot call myself a true fan). Like the Vocaloid scene, chiptunes creators and fans embrace a type of music often associated historically with a visual component that arguably puts less emphasis on the music or the musicians (video games in the case of chiptunes), but people who love chiptunes use concerts to show their appreciation and their talents outside of what is typically expected out of a “performance.” I have to wonder if Vocaloid concerts and the like are at an intersection between the rock concert where band members are sometimes viewed as gods, the orchestra where the relative significance of the individual musicians versus the composers’ original scores can be argued, and the club where a DJ combines beats and melodies together to form something new.

Super Robot Wars UX is Full of Whippersnappers

A new Super Robot Wars game was announced yesterday, Super Robot Wars UX for the Nintendo 3DS, and the amount of new and unexpected entries makes me want to talk about it, as well as some other SRW-related thoughts.

I think you can roughly categorize Super Robot Wars into two types of games: the flagship titles, and the experimental ones. The former consists of the titles with the best animation and the most-anticipated anime entries into the franchise. The latter can go in a number of directions, from aesthetics (3D models instead of 2D sprites in Super Robot Wars GC) to gameplay (a switch from turn-based to real-time strategy as with Super Robot Wars Scramble Commander), but often times “experimental” simply ends up referring to the titles chosen for that game.

That’s pretty much where UX is. Just look at the debut works for this version.

  • Kishin Houkou Demonbane
  • Fafner in the Azure: Heaven and Earth
  • Wings of Rean
  • Cyber Troopers Virtual On’s Fei-Yen HD
  • Mobile Suit Gundam 00: A wakening of the Trailblazer
  • SD Gundam Three Kingdoms Legend: Brave Battle Warriors
  • Mazinkaiser SKL
  • Heroman

When you include the other titles that are in this game, the first thing that jumps out is just how new most of the anime are. Not only is the Mazinger franchise represented by its latest one-off OVA series, but the actual oldest anime in the entire game (and the only two from the 1980s) are Aura Battler Dunbine, and then Ninja Senshi Tobikage of all things. If it were a flagship title, there would have to be certain staples, but with a “lesser” SRW like this, it’s possible to inject a ton of new blood into it and not offend anyone.

Not only that, but when you look at some of the recent titles chosen for UX, they seem to be among the least likely candidates even among non-flagship SRW games. Brave Battle Warriors is actually an already-super deformed Gundam anime done entirely in 3DCG and based on classical chinese literature, the sort of title one would least expect to represent Gundam even with the fact that SEED Destiny and 00 are there. Though I’m sure it’s based on the anime version, Demonbane‘s inclusion may be the first instance (and correct me if I’m wrong) of a visual novel appearing in SRW, which opens the gate for things like Muvluv Alternative.

Heroman I wasn’t even sure counted as a giant robot anime, though I guess if you think about it, it’s basically a combination of Tetsujin 28/Giant Robo with Gold Lightan (though Gold Lightan has yet to make its debut). Possibly craziest of all is the inclusion of Virtual On in the form of a Fei-Yen dressed like Hatsune Miku. Virtual On in SRW Alpha 3 paved the way for non-anime/manga to appear in Super Robot Wars games, and this takes it to another level, as I’m pretty sure Miku Fei-Yen is nothing more than a model kit!

It might sound like I’m complaining, but I’m really not. I actually love it when SRW games go a little wild like this, though one complaint I do have is that the DS SRW games have never been the most impressive when it comes to animation. My issue isn’t even with the quality of the sprites or an unfair comparison to the exquisitely animated Z series of SRW, but that a lot of the shortcuts taken to try to make the games look better actually end up making them look worse. In particular, I’m referring to the way the DS games including UX incorporate cut-ins, and detail shots. Instead of creating the images to better match the sprites and the visuals of the rest of the game, the DS SRWs basically take screenshots directly from the original anime, and while this means things look accurate, it also sticks out in an odd way and messes with the way the attack animations end up looking in a manner which didn’t quite affect previous games with worse sprite animation.

But it might just be that with a game with this daring of a series list, some things have to give. In that case, I’ll take it, but will still hope for better the next time around.