Update 45

tl;dr - Long post today! I've been trying to adjust Twyla's design for LMO to little avail; meanwhile I've had more success with designing new images for PAL and testing the alpha, although that's still proving to be a slow and deliberate process.

This week's image:

This image isn't a recent one for me - in fact the original pose is one of the earlier PAL images I designed - but I have recently readjusted it and redone it in Iray, so here's the new version. Obviously it's a follow-up to the image I showed in update 41. I'm fond of this pose; it's the middle one of a three-part series (first without the hands, third tweaking the nipples) and I like the comedy of the scene.

Actually I was hesitant to share this one because it felt a little too similar to some other stuff I've shown you, but I'm still being careful not to blow too many spoilers. Since I've shown you something more or less like this before it's a relatively harmless thing to show off; there's no plot developments disclosed here that you wouldn't already expect.

Let Me Out design work:

I shared with you guys some of what I've been working on lately - I've been trying to get the design of Twyla to work in the Iray rendering engine. Basically I've found there's a significant uncanny valley problem with Twyla and I've been really trying hard to change her expressions, facial features and skin tone in subtle ways to make her look less weird. I haven't been able to entirely work out why some characters just don't work in Iray while others are significantly improved. Twyla turned weird while Kimiko looks terrific - why?

After all that work I'm really starting to think that I need to revert back to 3Delight rendering for Let Me Out. I just can't see a good way to make Twyla look right in Iray, and I've had to make a lot of compromises in that design that I don't like all that much because of the quest to make it work. If I go back to 3Delight this sacrifices some texture and lighting quality, making the whole thing look more cartoony and less realistic.

On the other hand, however, the original Let Me In was all done in 3Delight and I felt it was pretty visually good, so there would be consistency there and I think it would probably still work. Also the more cartoony designs of LMI/LMO might just be flat out better suited to that engine. Oh, and the rendering time will be a lot shorter, which would allow me to complete the game slightly quicker (which is nice).

I'll show you some more test shots soon to keep you in the loop of what I'm doing and to solicit your opinions about the best course. Ultimately it's up to me to judge which is best, of course, but some transparency here is probably a good idea, and you guys have been great so far giving me feedback on the uncanny valley issue (which I can't always tell when I've been working on the image - it's like a magic eye picture). If you already have some opinions on the issue they'd be welcome too.

PAL image design work:

Okay, back to the main project. :) I've also made a number of new images this week; mostly revolving around the Ai sex scene, the Rio sex scene, one of the threesome endings and some replacements for a few every day images that I just didn't think were all that great. The best achievement was for the threesome ending, where I did a cool little bit of tinkering with one of the character's skin textures to create a nice effect (some stuff written on her skin). Since it worked well I decided to use that texture a couple more times and came up with some good images that I'm pretty sure you're going to like.

I actually did more PAL image design this week than I did LMO work, but I had more to say about LMO because I'm less sure about it. All the PAL stuff went pretty well.

PAL testing and programming:

I've continued my test of the PAL alpha... slow as it is. I've made a bunch of bug fixes; basically the Ai scene wasn't working at all before I tested it. The good thing is that now I think it's functioning well and I was able to improve it significantly with a few strategically placed extra images. So Ai is ready to go (and I also fixed some issues with her romance scenes) so it's on to the Rio scene!

There have been a bunch of other bug fixes needed too. Beyond the game killer bug in the lake day scene that I fixed a while back I found more problems with the lake day that I've been able to correct, such as the girls not actually showing up to the lake when they were supposed to. I've also restructured it slightly so that it should flow a little better.

Yesterday I came across a bug that was preventing Fumi's bullseye spell from working and I still haven't solved that one yet. I'm going to have to dive back into it, hopefully tonight when I get some time, and see if I can dig out the reason.

The future:

I've been toying with ideas for my future projects. I am very up in the air about one in particular that I think will likely be the next after LMO, because I want it to be building on what I've done before but also something fresh and innovative that doesn't just feel like you're playing the same games over and over again. I think I've got some interesting ideas, although they're not really nailed down yet. It's not a big deal as I need to focus on PAL anyway, but that's where I'm at, anyway.

Now it's full steam ahead until... Monday. Unfortunately that's when I'm about to get smacked right in the face with a two-by-four of day job work. I'll keep plugging away on the game, naturally, (that's how I've scheduled my life at the moment) but this week is going to be a busy one.

Oh, and don't forget that you can now vote in the Erin Awards for 2015 here.

Suggestions? Thoughts? Questions? You know what I'm going to say next - I'd love to hear them in the comments section!