Forum - Let's Talk About Pixelscaling

Introduction

So let me preface this by noting this.

I feel calculations should only be used when:

• A proper number is given/can be inferred.
• A proper scale can be inferred, for example if someone's official weight and speed is noted and they directly ram into a person, a KE calc is perfectly fine in my book.
• Or for example, Yor throwing a volleyball into space, you can take the average weight of a volleyball and take that it leaving into space requires escape velocity to be reached, thus you have KE there too. While they're not officially listed numbers, they're something that can be proven beyond a reasonable doubt to have credence.

With this in mind, there is nowhere I can see Pixelscaling work in this regard, I've tried for pretty much a decade to understand why this method is used and none of the reasons are valid in my book.

But before I go into the revisions, let me explain the history as I understand it, behind pixelscaling.

What is Pixel Scaling?

In short, the term is essentially based off Pixel-art scaling algorithms, where you define shapes and enhance the images of pixel art. On a screen there's dozens of tiny boxes and each box measures in what is essentially 1 px give or take. Pixelscaling is using this knowledge and comparing pixel sizes to try and gauge height. This is a very simplified definition mind you that doesn't break down the more complex processes behind how pixel art is done and such.

The Problem?

The problem with pixel scaling is the very concept behind it requires one to assume everything is drawn to scale in a scene, take for instance, an area in EarthBound or in Undertale. Frisk and Ness have a sprite and when attempting to measure them with houses or such in the background, they will normally get results that are far smaller or larger then a house realistically should be in size.

But this problem doesn't exist just for video games. One of the most infamous height inconsistencies in cartoons comes from Steven Universe, where characters with canon heights will just randomly swap and change in size. Another example is Koichi from JoJo's Bizarre Adventure, he has a canon height that does not at all match up with his size compared to others with a canon height.

Take into account stylized art styles, perspective shots, camera works, and more that all serve to make a cool scene rather then an accurate drawn to scale size.

But it doesn't end here! Sure you can say that characters have a canon height and maybe the character's height next to them is consistent with their canon height, but what about buildings, the ground, the area they're in, the specific axis the camera or visual is at, etcetera? Most authors do not include a canon height for everything and attempting to find the size through areas that are normally not drawn to scale or absolute scale at the bare minimum is just asking for a bunch of assumptions.

Calculations themselves are already a very headcanonish. As an example, Butch Hartman was asked if the level of power Danny did when destroying a golf field was intended and he just said "No we weren't thinking about that, we just thought it would look cool." So they're already hard enough to justify as they fall out of the intention of the source material, now with Pixelscasling, you are taking that already dubious method and applying far more assumptions into it, leading to calcs that massively inflate or deflate results.

I'm not sparing myself in this either, I was part of the problem. This calculation for Golem genuinely requires so many assumptions all based off him having a listed canon height. It's one of my least favorite things on official pages here and it was only kept up by the sheer luck of there just randomly being a neutron bomb in this series that has roughly the same ballpark tier. Which I should also note I had to keep re-doing this calc because the numbers kept being far too high for me to realistically agree, like numbers at 6-C and such.

The Solution?

Nuke Pixelscaling completely, yeah that's it. I could try and come up with some overly convoluted system to where it immensely deflates results and stuff, but not only would that be a waste of time and not really fix anything in my eyes. It also still has the issue of being a concept that's inherently flawed.

This also adds another thing that should be an added requirement for calculations, they should be under a "beyond a reasonable doubt" scale for validity. I think it's self-explanatory but it can be explained more in the future.

TL;DR Nuke pixelscaling and pixel-based calcing, it's the definition of headcanon.