VN Ren'Py Game Dev Journal: Inner Demons (Ch 4 progress)

[GrayTShirt]

Results of render tests. Looks like 30 minutes is the good spot. Very happy.
2 hours, 47 minutes 4100 samples to 30 minutes 569 samples.

So now I can get more than 5 renders in the time it takes for one. Still using the CPU unfortunately, but the scenes haven't been optimized either. At least using the RenderQueue asset while sleeping and at work makes it all so much easier.
-GrayTShirt

 

[GrayTShirt]

Been able to reduce the render time even further. But I think it may be dependent on a few things. But it's for almost 8 minutes. We'll see if that continues through the remaining images. Attached image is before any post work is done.

 

[GrayTShirt]

Working on the new renders. I'm pretty excited for this update. 185 images down, got about 20 in the render queue. Need to do some improvements to a couple images.

 

[GrayTShirt]

New teaser. Working on one of the new character models. Liking it so far. And, I'm about to work on the first full-on lewd scene. Only took like 250 to get to...
-GrayTShirt

EDIT: Removed the images I posted on Monday since I re-did the "Madam" model. Looks a lot better now, but I'll keep it as a surprise.

 

[GrayTShirt]

Here's another image I was working on. The Bathhouse/brothel. I had to use a few assets to make it. Favreal MICK model, hot tubs, lots of vases from different places, garden arches, chandeliers, candles (fuck tons of candles), center fountain, potted flowers... It took forever...

 

[GrayTShirt]

Finally! Starting on the lewdidy! Here's a couple of teasers...
The first one is a 45sec render, so totally pixelated. The 2nd is just the viewport screenshot.

 

[GrayTShirt]

So now, I'm having another issue. THis time with collisions. I can use breastacular to set only one item to collide at a time. So, I need to use the invisible gloves since they're more complicated. What I need to do is find out how to make a "tunnel" in between the breasts so it doesn't look like the dick is just... there.

 

[GrayTShirt]

Well, I kind of solved it. Just set the collision for the penis and positioned the hands in such a way that you couldn't tell they weren't colliding with the breasts.
And here's a little lewd treat- HANDHOLDING! ('cause they're nothing more that a scientist could want.)

 

[GrayTShirt]

Using moisture shells and wet hair. Hoping these renders turn out well.

 

[GrayTShirt]

So it looks like I'm missing my target render quota by like, half. Upgrading to Daz 4.14 didn't help. All of my lighting changed. You can see stark difference between the 2 images where it updated. That was images 289 and 290. So, I either go back and re-render 289 images to get a complete upgrade across the entire chapter thus far, or suck it up and continue as is.
I think I had a character model change some of it's HD face shape also, but that I'm just going to have to accept. It's really only a character you see for a quarter of a chapter (albeit an important character) thus far.
Now, I have about 300 more planned images to do. Then the write-up, Ren'Py programming, and builds.

 

[GrayTShirt]

Okay, so I seriously underestimated how long these corrections are taking. I found the problem between the lighting differences and now I'm having to go back through over 100 renders to fix them. THEN, I found some more clipping issues, inconsistencies between shots, and of course, those ever-so-important money shots.

Fucking money shots, man. It's easy just to throw a bunch of cum on someone, it's harder to make it look good though. (Never in my life did I think i would ever type a sentence like that.) Attached is an example of how bad they look now. Need to look into surface and parameter settings. And how to effectively use the db xxx dForce cum packs.

EDIT: If there's a render setting that works well for you, you can "...save as" the render setting and use that. Just remember to look at what environments you're using. I didn't realize I changed the sky HDRI from a starry black sky to a huge moon purple sky.

 

[GrayTShirt]

So, I'm finally able to go back and correct EVERY SINGLE BATHHOUSE render. The example shown here is the main reason why I'm having to go back through so many renders. I swear, at some point I will finish these things and get to the script and Ren'Py builds.
It's things like this that make some games late with their updates. Especially if it's their first game. A lot of problems pop up that you never think of. Again, thank you ManFriday Render Queue.

 

[GrayTShirt]

Finally finished the last image of the first lewd scene! This is insane to me. I would have thought I'd be farther along by now. But finding all kinds of little details that need to be corrected, posing individual fingers, and test rendering takes so much longer than I thought.
Now I need to make some post-lewd (normal) shots. As far as what I wanted to do for this next update, I'm only half way done. But I think it will go much faster since I finally know what I'm doing and how to do it.
Here's a teaser shot of the last lewd scene, pre-render.
Still need to do all of this fuckin' post-work and Ren'Py build though...
-GrayTShirt

 

[GrayTShirt]

Everything is looking good so far. Still have a bit to go. A few pre-post-work shots. Some of my favorite shots. I suppose you can tell though. And a transition shot.

 

[osanaiko]

Hey, it's my first time to check in to this thread in a while.

Well done on making a release. I see the many of the commenters were just as kind and gentle as you might expect around here . You handled the replies well, so as the wise Ms Swift says, "hater's gonna hate, shake it off".

Looking at the more recent posts, you've made some great progress, I can see you also hit some stumbling blocks that forced you to do "re-work". That's unfortunately the nature of learning a new skill, especially one like this where there is nothing but trial-and-error for many things. Optimising for render speed and Daz scene setup speed are key to churning out a lot of image content.

Anyway, I'm very glad to see you are still pushing forward, which is the most important thing; pure skill or passion doesn't get a project like this done, persistence and continually plugging away at the tasks is the most important.

 

[GrayTShirt]

Working on set pieces is a real pain in the ass, ya know? For anyone wanting to work on a new game, plan your environments! I thought I had everything worked out, but turns out I need to work on about 3 different sets just to get through this chapter.
In the mean time, here's a teaser for the Army Captain.

 

[osanaiko]

Working on set pieces is a real pain in the ass, ya know? For anyone wanting to work on a new game, plan your environments!

I know what you mean! there's a good reason you keep seeing the same environments over and over again in Daz based games, the available set of environment/props seems large, but once you get into a narrow genre, can be quite limited. There looks like only two coffee shops in the whole f95 game world....

It's possible to mix things up a bit with different wall textures or some limited kitbashing, maybe creative angles.

 

[GrayTShirt]

I knew going into this I'd need a few set pieces. I didn't realize just how much of it I'd need to customize. The tavern, bedroom, dungeon are stock, but I need to kit-bash so much more.
Now I have to make the outside of a garrison with tables, background architecture, and misc stuff. Probably take a few days to do.
Using Fugazi1968, Roguey, Muelsfell, and Faveral props and scenes. Not to mention all of the HDRI iRay landscapes and skies.
Ugh...

 

[osanaiko]

i always have to remind myself that players aren't necessarily there to ogle the great props you have in the background.
Keep it simple, use some depth of field to hide the details a bit, layer some simple banners/weapon racks/barrels or whatever over some flat walls with textures stolen from another environment, you can get there!

 

[GrayTShirt]

So, figured out how to (successfully) use MICK architecture. There is no rotation or translate sliders for the actual building materials. But the snap x, y, z sliders work wonders. Once I was able to figure it out, it took everything in me to STOP building the garrison. This thing was going to be massive if I didn't limit myself.