LewdGamer is proud to present our newest feature, Behind The Lewds, in which we will be taking a look behind the development curtains to reveal the creative processes of some of the most interesting and ambitious lewd games. We start our journey with the highly anticipated Together with the Wolf Girl by seismic.
Before we actually get into the meat of things that is Together with the Wolf Girl, it is prudent to give our readers a short introduction to seismic’s assorted works and how they came to their current prominence.
The one-man development studio known as seismic came to almost instantaneous popularity late in 2010, when its first interactive animated erotic work Cats Attacks based on the light novel, manga and anime series Asobi ni Iku yo! (Let’s Go Play!) was released. Even back in 2010 seismic’s work featured crisp and smooth animated 3D models that combined the visual style of 2D Japanese animation with highly detailed three-dimensional renders. The high frame rate and fluid erotic scenes made in flash combined with designs that had both cute and sexual appeal made for some delicious fap material and as the interest in seismic’s initial work grew so did the library of his titles.
As years passed seismic created even more high-quality products that were still based around anime and game series. Their Puella Magi Madoka Magica inspired animation Sweet Mami got so intensely popular seismic was encouraged to make more animations featuring the voluptuous and sometimes headless magical girl Mami. Apart from derivative creations based on anime, seismic also created an animation based on the JRPG Criminal Girls, a different animation with Rabi en Rose from the anime and manga series Di Gi Charat, a simple mobile game with touching and fondling mechanics called Just a Girl touch and at one point even hinted at creating something with Patchouli Knowledge from the Touhou series of shmups. Suffice to say, the developer dipped his toes into a lot of parody works, which in the end only contributed to his rising recognition.
Seismic’s newest work Together with the Wolf Girl starring the always energetic and eager to please werewolf girl Liru from the comedy anime Renkin San-kyuu Magical? Pokaan has been in the development grinder from quite some time now. The anticipation for some good old monster girl sex only grew after the studio released a demo version of the game in August of 2014. What you can find right below this paragraph is the translation (courtesy of Joe, one of our lewd Japanese translators) of seismic’s latest posts from his official development blog where he goes in-depth about his workflow, the newest features of Together with the Wolf Girl, his general attitudes towards the erotic game industry and lastly provides an explanation why his latest work has been victim to a series of delays.
###Various info on the new features
Hey guys, I thought I’d go ahead and show a blog update here. Although my work in progress, “Together with the Wolf Girl” didn’t meet its first deadline, I’ve included some new features that I thought I should talk about here.
#####First sex of the year with Liru…uh, I mean celebrating Girls’ Day.
Here’s my blanket excuse as to why the game’s late, and what’s new that you can expect!
This was already present in the demo, but this game would be my first time implementing lip syncing. “Lip syncing” is basically just having the character’s mouth move in response to how they’re speaking. At first, I didn’t plan to have any lip syncing, but the further I went, the more I realized how many more conversation scenes there were in this than my previous stuff, and I began to think “Hey, wouldn’t it be cool if I could make this more like an anime?” I tried it out, and like I figured, it looked much better; so I decided to put it in.
Average lip sync and seismic’s implementation
Let’s take a moment to think about your average lip syncing for 2D characters. First you have TV anime, which has fixed lines and speech timing. Basically, they just match those up and insert the correct look for the characters’ mouths. On the other hand, with galge, there are usually characters just standing upright in a still shot but with more diverse lines and options, so they match those up and sync them. However, in those cases, the base they’re building off of is a still image. They usually just take that image and overlay the different phases of the mouth based on what they’re saying, and sync it that way.
Now, let’s think about this circle’s preexisting works. In other words, it would be “Use a pre-rendered video, and just insert 6~8 voices over it.” This would mean one video would have lines with different timing and length in it. How would you lip sync that…if you matched it to the longer lines, you’d have it keep going after the cut-off when the shorter lines are being read. Not to mention how much of a problem it would be if you were to cut and overlap lines onto the video loop. Quite a problem indeed.
At first, I planned on just taking out the dynamic, changing dialogue during the love scenes and replace them with simple, scripted dialogue, but felt that really detracted from the overall enjoyment of the game. And I would’ve felt kinda bad for the voice actress. Plus, if there had been set lip sync for all the lines in all the scenes like that, it would have really bloated the amount of 3DCG work I’d have to do.
Then let’s do it like this
Thinking this could be solved on the video software side, I prepared two types of videos, one being idle videos (meaning the ones that have no lip syncing), and the other being ones with the lines lip synced. I took those two types and interchanged them according to what lines were being said at what time. That idea called for some pretty heavy work on my end, but the result looked so natural it surprised even me. Not only that, but the tedious work of getting the facial motions just right was pretty much nonexistent with this method, though it did double the amount of work that would have to go into the character rendering and composition. Before my predecessors could even know it, the stuff they used to just leave to the PC itself became manual labor for me, so I feel the process is very different.
That’s done! → So let’s fix it up
So in accordance with that, though the trial used C#＋MonoGame（OpenGL）, in order to get the environment up to snuff, I had to used Flash (AS3.0). The sound data and stuff from AS3.0 required that I studied up on it a bit more to get totally familiar with it, but in the end it came out great.
Although that massive work that would have went into hand-crafting the facial animations so that they fit the lines was gone, the work was still there, as I had to basically investigate the entire timing of the voices with the video, making sure there was a good synchronization rate.
So unlike the usual anime process of adding the sound after the “film” is already made, the process ended up more like something Takahata Isao-san does, with prescoring the lines before the “film.” Doing this as an individual though is some real work. Right now I’m working on the other scenes, using 3dsmax (which I haven’t used much) to make the motion scenes that support the voices.
#####Now that I think on it, I hadn’t used it much.
As always, I try recording the voices before doing anything else. Sometimes I look for a good-sounding voice actress with a free schedule, and I bookmark their page. But in the few years since I’ve started working on doujin stuff, there’s a lot of them that have stopped voice acting, have a dead home page, or my bookmarked links are just plain dead for some reason. I know I should wait for the CGs and stuff to be done before looking, but I feel like if I did that I might miss out on a great opportunity to get a good voice actress…let me tell ya, it’s a hard situation.
My past stuff had an option to select your situation (position), and that made a chapter select option redundant. However, this time we’ve got this flow of “meet Liru at the door after work > eat dinner > get in the bath >fun time in the bed”, so if you’re looking for a scene real quick, a chapter select option really would be necessary.
I’d also say I think this would increase the overall quality standards from previous releases, so I had this “I want the quality to be even better by adding this!” feeling too… and yeah, this accomplished that pretty well. It’s also possible to select scenes that only come about because of the dynamic dialogue that changes during scenes, so that’s really useful.
As a side note, there’s actually a chapter select button in the upper left corner near where the auto-mode on/off button is, so you really can just go to whatever scene you’d like at any time. It comes in handy for me while I’m developing too if I need to debug or check a certain video part.
###In regards to being able to switch in English translated text
This time, there’s been a lot of interest and a push for this from people overseas, so I looked into getting the text translated…is what I’m pretty sure I said before, right? I think it was around the end of February, wasn’t it? Actually, I did end up using DLSite’s translation service. It cost about 80,000 yen, like I figured (TL Note: this would be roughly $675.). Again like I figured, our data exchange and whatnot went pretty smoothly and I got it done in around a week.
Of course there’s a feeling of wanting to fulfill the hopes of oversea fans, but it’s not like I’m particularly trying to impress them or use their enthusiasm or anything. It’s more of an “I wonder how an English translated version would sell” kind of thing. I became so interested that I figured it was worth spending 80,000 yen to know. The material is our long awaited Liru after all. But it’s not like I’m that interested in the actual profits itself. I don’t really mind if people call me a money-obsessed asshole behind my back…but come on, aren’t you interested in how this will go?
I also wanted to try using DLSite’s new service, because they’ve been very helpful to me so far, so I couldn’t be happier if this whole thing would contribute to the doujin business and DLSite worldwide. I couldn’t care less about the CG creation business, but I’d really like for the doujin world to have some revitalization.
#####You can already change the config screen just with XML.
Here’s what happened in regards to the work on my end to support the English translation. Around the time I asked DLSite for the translation, I edited the Excel data I prepared for the voice actress. We emailed back and forth about it, but that actually only took a day or two. However, actually inserting the English required me to fiddle with the program, match the voices up with the display text, and other small stuff. That took about a week or so to get done.
And so here’s how I plan to do this. There’s going to be a Japanese edition that you can download, with extra data after that, which includes the program and files necessary to add English subs. The implementation will therefore come right after the Japanese version is released. Also, there won’t be any delay to the release because of the English translation. Instead, it will be the same as purchasing a normal Japanese version, except any overseas users will be able to download extras that will make the game subbed in English. ok?
###“Shout!”, a new feature
Now I’ll explain the Shout! feature, which I didn’t put in the pleminary announcement on DLSite even. It’s a feature that allows you to use a microphone to play the game without any mouse clicks.
The idea for the feature came from the user comments on the auto mode of seismic’s previous games. If a fighting game’s controller is an arcade stick, then an eroge’s is the mouse, right…actually no, it’s obviously that stick right between your legs, isn’t it? So it makes sense that anyone would want a way to control the game hands-free, which led to the development of the auto mode.
Now, the older games have a sort of random divergence in the auto mode (creampie, bukkake, etc.), and the loop cuts for the video are fixed (through XML rewriting you can play around with them though). I was thinking; what if we could expand on that, and have the player determine what divergence they want (without using the mouse), how many loops the video has, and whatnot?
However, just having the obvious “decide what predetermined course the auto mode will follow” and “decide how many loops to make of the video” would just make it a normal movie, and that’d be pretty boring. The objective was simply to make the game advance without using your hands. Isn’t there any way to control the game without using the mouse…? After thinking about that for a while, I came up with “Shout!”
#####Shout! feature demonstration
“Shout!” picks up your voice if you speak loud enough into the mic, and effectively pushes that button’s choice. If you keep talking when there’s multiple choices, the cursor will move to them and select them.
Explaining shout in 3 seconds
speed up that piston > vaaa- > creampie
Of course, there are the people that like to yell “I’M CUMMING!” and act it out like a voice actor, and there are also people that need to use both hands to fap, or have their hands slippery with lotion and don’t want to touch the mouse. All of those are now fine to do. In addition, for those extra high-grade gentlemen, you can even do your computer work or play an FPS in another window while having sex with Liru, because the latter doesn’t require your mouse or keyboard.
######The latest version has a sensitivity option.
…to be totally honest, while I really was worrying over the auto text thing…that’s not the only reason I put this in. While doing the motion work, I saw that AS3.0 had really easy mic support, so I’m was like “hmmm, that would be cool to put in”, so yeah. The idea might have come from there.
While this isn’t a new feature per sey, the smoothing when you went fullscreen with the demo wasn’t great, and the jaggies stood out more than the previous games. This will be improved in the release version, with an option to turn it off for low spec machines (default is on).
Lastly, seismic provides a thorough explanation why Together with The Wolf Girl has been subject to a lengthy period of delays and at the same time provides us with a new potential release date to look forward to.
In regards to a release date, judging from my skills, work speed, and physical and mental condition, the end of June seems like a reasonable estimate…is what I’d like to say, at least. I know this is a pretty big wait compared to my initial estimated completion date, and I’l like to sincerely apologize to those that had high expectations for the original date.
If you happened to fall in love with this particular developer’s work after reading our article or if you’re just curious, all of seismic’s erotic animation games (some with an English translation) can be purchased on the English version of DLSite.
We hope you enjoyed taking the virginity of our newest “behind the scenes” feature into game development. You can rest assured that we have more juicy material in store for you in the not too distant future. As for Together with The Wolf Girl, we sincerely hope it won’t get delayed again and doesn’t become the Half-Life 3 of 3D animated porn.
For news on lewd game development, be sure to follow us on Twitter.
If you enjoy our content and would like to support us, please consider donating to our Patreon.