Progress Update
A little birdy told us that maybe we should've provided some more info on the progress of Episode 2… and yeah, fair, we should have.
And so, a word from our programmer:
Hello! Aebard here. (It's actually always me, but now I'm speaking as me, not as the combined entity that is Cyberframe). I'm the big hold up on the release of Episode 2, and the one that kept providing (extremely inaccurate) estimations on release dates.
Firstly, I'm extremely sorry for hold up, and flat out wrong time estimations. At the time, I really thought they were going to be accurate… even up to the hours leading up to the deadlines. I know that seems insane, considering how much longer it has been, but it's still true. I wish I'd had a better handle on the time things would take (and still do), because the last thing we want is to let people down. Even so, that's what happened, and that's on me.
I haven't given a new updated estimated date because, again, I just can't trust my own sense of how long things will take yet. That said, I really don't want to leave people hanging, so here's the next best thing.
Here's the overall progress of Episode 2 (as of 2/19). I'll be updating this daily as I finish up work on the game. (Follow this link to see the current status[docs.google.com])
As you can see, the last big hold out is scene posing. I've spoken about this briefly in some previous posts and on discord, but this is by far the most time consuming part of creating LT, and it's what gives the game it's specific look and feel.
But why is it taking so long?
There are a couple reasons, which I'll briefly explain here.
Please don't take any of this as excuses. I still should've managed it better. This is just to give some insight into how I could heck up the estimations so badly.
The biggest thing was that I'd been using Episode 1 as the gauge for how long things would take. The posing in Episode 1, while time consuming, was actually very simple in composition: mostly simple back and forth shots between two characters. The majority of Episode 2, however, features three or more characters on screen, all of which need to move and pose on screen, interacting, with camera shots that make logical sense, etc… and all that turned out to be exponentially more time consuming than the back and forth shots of Episode 1.
In a similar vein, Episode 1 is largely setting the stage and introducing characters. This makes it a little easier to transition between scenes, without things feeling too jarring. In Episode 2, we already know our main characters, and now we're traveling with them through the world, scenes leading one into another, and "fade to black" just doesn't cut it most of the time.
Then there's Live2D. Live2D is awesome. It's what allows us to give the characters so many poses and expressions, instead of just 3 or 4, and what lets us animate them. However, Live2D is incredibly time consuming to setup & create the characters. Episode 2 requires 4 new characters (and some change, some of the CGs use Live2D), as well as modifying the current characters to support specific motions and poses that they couldn't in Episode 1. All that wound up eating up a few months of development.
Which leads us to the posing. The primary issue with this is… After Effects. For those of you who know Ren'py (the game engine we use), you'll know that it's all code, with no visual editor. This means that all scenes have to be written out as code, trial and error style. Lustful Temptation would be impossible (or at least, look much worse), if we had to trial and error every shot. So, to streamline things, I wrote a tool that would let me layout the scenes, plan motions, and pose characters in After Effects, and then export that directly into code, which I can paste into the game. And it works great! Except… After Effects is notorious for being laggy, and the lag just gets worse the more complicated the shots become (and I suspect the Live2D plugin is also tanking the performance). Saving currently locks me out of the program for a full 40 seconds, which doesn't sound like a lot, but any After Effects veteran will let you know that the lag and save interruptions are enough to break a man.
There were other things here and there. Bug chasing, getting sick, stuff that worked during the first pass that just didn't hold up once we put all the posing and animations in, etc. Little, unexpected things, but they add up.
And finally, my own dang brain lying to me about how fast I can get things done. Even now, my gut tells me that I can complete the rest in an amount of time history has shown me is incorrect.
And how are you going to make these aren't problems going forward?
This has been at the forefront of my mind this whole time, ever seine missing that first deadline. We've mentioned before about bringing on some more people to help with some of the tasks that I (as the programmer) don't need to be doing (art cleanup, Live2D, audio mastering, etc), and that's going to be a big help. I'll also be handing off some of the posing as well, splitting up that work load.
As to After Effects? These past months, as I've been cursing Adobe daily, I did figure out a way to build a tool to let me do the Live2D posing in-game, rather than having to go through After Effects. This won't eliminate After Effects from the process entirely (yet), as I'll still have to do the shot layout there, but it will speed things up considerably.
And so, once again…
I'm terribly sorry for how long this episode has taken. It's not where I thought I'd be, and not what our players deserve. I appreciate everyone that's supported us, even with the delays and disappointment.
So keep an eye on that chart, I'll be updating it everyday, getting those last percentages knocked out and closer to release. Episode 2 is looking fantastic, and I'm excited for everyone to see it… and doubly excited to start in on Episode 3.
As always, I'm available on our discord[discord.com] and happy to talk about the game's development and answer any questions, and I'll keep an eye on the comments below!
And so, a word from our programmer:
Hello! Aebard here. (It's actually always me, but now I'm speaking as me, not as the combined entity that is Cyberframe). I'm the big hold up on the release of Episode 2, and the one that kept providing (extremely inaccurate) estimations on release dates.
Firstly, I'm extremely sorry for hold up, and flat out wrong time estimations. At the time, I really thought they were going to be accurate… even up to the hours leading up to the deadlines. I know that seems insane, considering how much longer it has been, but it's still true. I wish I'd had a better handle on the time things would take (and still do), because the last thing we want is to let people down. Even so, that's what happened, and that's on me.
I haven't given a new updated estimated date because, again, I just can't trust my own sense of how long things will take yet. That said, I really don't want to leave people hanging, so here's the next best thing.
Here's the overall progress of Episode 2 (as of 2/19). I'll be updating this daily as I finish up work on the game. (Follow this link to see the current status[docs.google.com])
As you can see, the last big hold out is scene posing. I've spoken about this briefly in some previous posts and on discord, but this is by far the most time consuming part of creating LT, and it's what gives the game it's specific look and feel.
But why is it taking so long?
There are a couple reasons, which I'll briefly explain here.
Please don't take any of this as excuses. I still should've managed it better. This is just to give some insight into how I could heck up the estimations so badly.
The biggest thing was that I'd been using Episode 1 as the gauge for how long things would take. The posing in Episode 1, while time consuming, was actually very simple in composition: mostly simple back and forth shots between two characters. The majority of Episode 2, however, features three or more characters on screen, all of which need to move and pose on screen, interacting, with camera shots that make logical sense, etc… and all that turned out to be exponentially more time consuming than the back and forth shots of Episode 1.
In a similar vein, Episode 1 is largely setting the stage and introducing characters. This makes it a little easier to transition between scenes, without things feeling too jarring. In Episode 2, we already know our main characters, and now we're traveling with them through the world, scenes leading one into another, and "fade to black" just doesn't cut it most of the time.
Then there's Live2D. Live2D is awesome. It's what allows us to give the characters so many poses and expressions, instead of just 3 or 4, and what lets us animate them. However, Live2D is incredibly time consuming to setup & create the characters. Episode 2 requires 4 new characters (and some change, some of the CGs use Live2D), as well as modifying the current characters to support specific motions and poses that they couldn't in Episode 1. All that wound up eating up a few months of development.
Which leads us to the posing. The primary issue with this is… After Effects. For those of you who know Ren'py (the game engine we use), you'll know that it's all code, with no visual editor. This means that all scenes have to be written out as code, trial and error style. Lustful Temptation would be impossible (or at least, look much worse), if we had to trial and error every shot. So, to streamline things, I wrote a tool that would let me layout the scenes, plan motions, and pose characters in After Effects, and then export that directly into code, which I can paste into the game. And it works great! Except… After Effects is notorious for being laggy, and the lag just gets worse the more complicated the shots become (and I suspect the Live2D plugin is also tanking the performance). Saving currently locks me out of the program for a full 40 seconds, which doesn't sound like a lot, but any After Effects veteran will let you know that the lag and save interruptions are enough to break a man.
There were other things here and there. Bug chasing, getting sick, stuff that worked during the first pass that just didn't hold up once we put all the posing and animations in, etc. Little, unexpected things, but they add up.
And finally, my own dang brain lying to me about how fast I can get things done. Even now, my gut tells me that I can complete the rest in an amount of time history has shown me is incorrect.
And how are you going to make these aren't problems going forward?
This has been at the forefront of my mind this whole time, ever seine missing that first deadline. We've mentioned before about bringing on some more people to help with some of the tasks that I (as the programmer) don't need to be doing (art cleanup, Live2D, audio mastering, etc), and that's going to be a big help. I'll also be handing off some of the posing as well, splitting up that work load.
As to After Effects? These past months, as I've been cursing Adobe daily, I did figure out a way to build a tool to let me do the Live2D posing in-game, rather than having to go through After Effects. This won't eliminate After Effects from the process entirely (yet), as I'll still have to do the shot layout there, but it will speed things up considerably.
And so, once again…
I'm terribly sorry for how long this episode has taken. It's not where I thought I'd be, and not what our players deserve. I appreciate everyone that's supported us, even with the delays and disappointment.
So keep an eye on that chart, I'll be updating it everyday, getting those last percentages knocked out and closer to release. Episode 2 is looking fantastic, and I'm excited for everyone to see it… and doubly excited to start in on Episode 3.
As always, I'm available on our discord[discord.com] and happy to talk about the game's development and answer any questions, and I'll keep an eye on the comments below!
More from Vaygren - Lustful Temptation
