In developing Workin' Out, I've been playing with some concepts of storytelling via narration and visual minimalism. The following post describes some of the approaches/techniques I've been trying out. I am also critiquing some storytelling techniques I've encountered in other VN's that I've played and enjoyed - without naming names, because I feel it's unfair to them for me to be critiquing their approach (I don't feel it's my place). 

Am I correct?  Probably not, storytelling is product of the author more than anything else - an authors style is their style, and you either like it or you don't... But, opinions are like assholes, everyone has one - and while I don't plan on showing mine, I will share it.

Quick warning - this post contains extremely NSFW content.

Critique: wasteful images in storytelling

There is one AVN that I really enjoyed when it first launched; enough to buy it as an early subscriber. The art style was one of the more common ones seen on itch - I'd guess Daz3D models/Blender renders, but the models were pretty and the storyline started with some real interesting elements. However, as the subsequent chapters were released the storytelling started to bog down. I noticed a trend by the author that began to really bother me. The author posted that the renders for their chapter were taking a long time, so they weren't making progress as fast as they planned. But at the same time for some of their releases were bloated with many still images to tell a segment of the story they could tell in one or two. I think the most egregious example was including something like 10 pics depicting walking from the passenger side to the driver side of a car. Why? It's not a movie, it didn't add anything to the story -there was no dialogue or monologue while the pictures were showing. It was just a waste of time and resources (IMO).

So, as I've been writing and developing the imagery for Workin' Out, I've been trying to edit myself. In the later chapters I've been generating more and more visual content, but I keep asking myself "Is that picture really needed? Does it add to the story?" For a metric, here are the first 4 chapters of Workin' Out, the number of images, and the size on disk of the image directory (4 is not yet complete):

This is a short story, currently sitting at 12,000 words, but I really don't have an idea what the proper ratio of word to picture should be...

• Chapter 2 is very dialogue heavy, where the relationship between the characters develops, and I use a few key images many times.
• Chapter 3 is when the adult content really starts, so it's a more picture heavy chapter.
• Chapter 4 is, again adult content heavy, but it's turning out to be a bit of an anomaly, there are a choice few images and I've rendered most of them at 2x resolution because of how I'm using them, which I'll explain below.

Critique: dependency on pre-rendered videos

Some other VN's that I've subscribed to include lots of full motion renders. I'm not criticizing the FMV's, they take a lot of work to set up and render... but in this medium, do they really benefit the story. Again, IMO, not always - especially if they are looping or if they have to be heavily compressed (to the point of affecting quality). I've downloaded some games that are 4, 6 or 8 GB. There is no way the script is that long, so there must be a lot of visual elements being included... and static pictures aren't that big.

I'm not against including renders. I've really enjoyed some stories with quite good FMV's: nice layout, good quality, and they add to the storytelling. So if the author has decided that's how their story is going to be told - that's their choice... But I've been wondering if there are alternatives, especially since Workin' Out is supposed to be a short AVN; to me that means it should also have a small download/install size, which precludes adding FMV's. Again, I'll explain below how I am trying to get around the limitation of no FMV's while still adding some expressive interest to the story.

Critique: creating too many images

This is something I'm guilty of! It's similar to creating a wasteful number of images described above, but also different. I've been working on some of the images for chapter 4 of Workin' Out, and the story I wrote required a picture of Gwen receiving cunnilingus followed by a picture of Gwen being fingered. So I fired up Stable Diffusion, created a bunch of image concepts of Gwen lying on her back, chose my favorite and then upscaled it. Then I blocked out the proper region mask for inpainting and instructed via prompt to add a hand/forearm fingering Gwen... but why? Why go through all the trouble trying to get SD to create another full size image here when I could work smarter and maybe even get a better result?

Solution: using Ren'Py features and layered sprites

Rather than explicitly adding a separate image or FMV of the POV performing oral on Gwen, is there a more clever way to imply it using the Ren'Py tool set? I think so... but ultimately, it will be up to the readers of the VN to determine if this is a successful approach.

Generally, I create the final images for Workin' Out chapter 4 at 2x screen resolution (3840x2160 for a screen size of 1920x1080) and use the oversample=2 attribute of the Image object. That way,  I can direct the script to zoom in and maintain good image fidelity. Let's try something...

To set the scene, the script reads:

ra "slowly you continue to stroke her clitoris, gradually increasing the pressure and contact until your tongue is lapping the top of her outer labia and clitoris"
ga "Gwen continues to writhe, her musky aroma fills your senses, her opening beginning to pucker and become slick"
g "Mmmmmmm... you\'re driving me crazy"

I hope the way I wrote the scene sets it up properly... a good author gives enough information while still allowing the readers imagination to fill in the blanks... but this is a Visual Novel... so can I do better? To imply performing cunnilingus I added the following image controls to the script:

show gwen_lying_bed_9 at truecenter:         
    zoom 2.0 yoffset -1400         
    ease 2.0 yoffset -1600         
    ease 2.0 yoffset -1400         
    repeat

Followed by 

ra """
   you continue...

Which creates the following effect (geez, had to compress the shit out of this image, and it hitches - but it demonstrates the concept. As implemented, it's seamless in game, hopefully you get the idea):

Overall, I'm pleased with the effect, I think combined with the way the story is written it adequately conveys an intimate embrace of Gwen.

Next, the script reads:

ra "cautiously, you probe her with a single finger, dragging it up and down her engorged inner labia before pressing slightly into her"

I originally tried rendering this as an additional unique image, but can we cheat here and get the same result with less expense? I think so:

I rendered up a disembodied forearm, wrist and finger:

Then, adding to the scene as a sprite layer, the following animation was added:

show ro_right_hand_1:         
    zoom 2.0 yoffset 500 xoffset 950         
    ease 2.5 yoffset 900 xoffset 1000        
    ease 2.5 yoffset 500 xoffset 950         
    repeat

Which creates this effect (yea, there's a typo in the text... I'm a noob):

Again, I'm pretty happy with how this came out, but, as I said - ultimately it's going to be up to the audience to decide. 

Overall, both of these story/graphical elements were added to the Workin' Out script via a few lines of code, reusing the same primary image and added barely an additional size cost (the arm weighs in at  275 KB). 

Since this is my first AVN and I'm still learning the tools I am going to keep exploring these ideas and techniques as I complete Workin' Out. 

I am curious if any other projects have experimented with the Ren'Py tools to tell their story in a similar way, or if there are any other good techniques that have been implemented.

Regards,

-Insomniac