Archive for January, 2017
Animations & sounds done.
I rewrote the jacket entirely during the friday afternoon to move optimize the memory usage. Darien gave me a few ideas I should test on the core TouchBound code later.
I haven’t recorded sound new sound effects in a long time. It took me several tries but I got some pretty convincing struggling and strapping sounds.
I’ve also made a set of animations for the different struggling minigames. They DO make struggling a little harder because of how the avatar move and I might possibly dial the difficulty down as a result.
Tomorrow: little code tweaks and I’m going to start working on the final materials (final canvas, but also leather nd latex with nylon straps 🙂
Struggle minigame refining.
Sooo… Mashing a button didn’t feel like it provided any skill or reward really (who would have thought?).
Then I came up with a way to spice it up: Now you have to hit a specific key each stroke to move the struggling progress forward, hit the wrong key and you get a time malus. After some testing, I also added a “prediction” that displays what the next keystrokes will be, so you can actually hit the keys a little earlier and essentially “ride” the latency.
The first version had the prediction separate from the “current key” but having them both together feels a lot more readable.
Now I need to bribe a few skilled DDR players to establish a benchmark of what the average difficulty should be.
Memory is still dangerously low on the main script but this is all offloaded to the main module so I’m not too worried.
Can you feel the struggle!
Yes I make horrible puns.
I have the base of a struggle mechanic, and ran out of memory twice on the main script. Not worrying at all…
Currently it plays like a button masher: hit pagedown rapidly to struggle against your bonds, you fail if you do not reach the goal before getting exhausted, or if you move (or someone moves you). If you succeed you loosen up the straps one “notch” and can struggle again after a brief resting period.
Balancing this is a bit difficult as it turns out and I might make the mashing a bit more complicated… Maybe something about hitting a specific key and if you mess up, you gain more exhaustion? I don’t know yet… I’m also gonna need to make a few animations&sounds to really “sell it”.
Oh yeah it’s using hovertexts for progress bars.
New update for the KDC Chat Conditioner
Yes, again! It’s annoying, but what’s even more annoying is bugs in products!
- Added: Word substitution filter.
- Fixed: a conditioner activation bug.
- The wearer impersonation now looks better and uses viewer URI to make the names clickable.
- Added: “Cedille c” to the microscript filter.
- Fixed: mute filter wouldn’t use the normal text impersonation method.
- Removed one prim.
Much love to the chat conditioner.
Today was another round of debugging, improving, refactoring performed on your friend the chat conditioner, I still need to do more testing before releasing a patch but the 3 people who actually use it should be quite happy.
The reason for me jumping to this, is that before starting on the jacket I left the TouchBound codebase in an “open heart surgery” state and the last patch that I released for the conditioner introduced new shiny bugs, so I wanted to set this stuff straight before starting the jacket core code proper.
Allright, time for code.
I spent the past two days cleaning up the model and tweaking the rigging some more, I think I’m done at this point. I could probably spend 6 more months on this and get it absolutely perfect but well… the law of diminishing returns is showing its ugly head.
Unless a glaring problem occurs, I just uploaded what I consider the “final” female version.
Now for the coding: I already wrote a fairly meaty texture applier/state changer system for it that allows me to swap between the different model parts and hide the unused ones. I’m trying a new approach of using alpha masking instead of setting things to 100% transparent, it appears to use less resources.
More tweaking of the tied version.
Still no end in sight… I corrected the upper body weighting so now the shoulders and upper arms look right when tied, however I still have issues with the back because this strap has to be weighted fully to the chest bone (it’s the locking one).
Another issue that isn’t visible in this picture is the weighting right where the arms touch the belly.
It’s… perfectible… to say the least.
Another inworld test!
Things are finally moving forward! This is a really terrible test rig, there are a lot of things that i still have to fix, but it is very promising.
Shoudlers are doing some terrible things to the model in this configuration so I’m going to get rid of them. I also need to figure out how to restrict weight copying to specific vertices.
Deform bones and posing
As I mentioned before (probably) my plan is to use two distinct models for the tied/untied version, the tied version having its arms stuck in place and not rigged to the torso instead, to ensure that it all looks as good as possible with different avatars.
I’ve fixed some UV errors I had appear between the bake version and this one, not entirely sure how they came to be, but to be sure I pinned all my UVs this tile.
I’ve also got some good fitmesh deformations on the chest, And I hope it’s good enough for femboys at the very least. I’ve purposely limited the breast size you can achieve through the sliders.
Fitmesh first pass
I redid the weighting from scratch again, with fitmesh this time. Good results, when Avastar doesn’t completely glitches out…
The bottom edge is surprisingly … wide compared to my hips so I must have done something wrong there. Additionally I noticed a few odd gaps that are probably due to some bad edge splits.
The texture for the back straps is the wrong one because I’m lazy. I also added a few extra edge loops to the front loop, it looks a bit less jagged now. The fitmesh stuff doesn’t quite work on the chest yet either, that’s another thing I need to work on. Don’t expect full breast size range tho, I did some testing and it simply looked terrible.