VSBATTLE WIKI REVISION: "INCONSISTENT SIZES"

[Arnoldstone18]

The issue of consistent sizes has been bugging my mind for a long time but recently a couple of threads have made me realize that something had to be done, First I would like to propose a few things we can do to tackle this issue. This is going to be a fairly short thread.

---

We already give a bit of leeway to authors who draw objects a bit off-scale. However, some of the CGMs tend to take a few portrayals that are depicted way off the original calced measurement without considering several factors that may have made those portrayals the way they are. I would like for us to tackle the most prevalent issues so I propose that we need official discussion guideline(s) to follow so as to lessen lengthy debates:

LARGE SIZE VS SMALL SIZE

Firstly,

When discussing the consistency of sizes of a measured object, using shots of the less focused measured object in the background of other more focused objects far smaller/larger in comparison should not take precedence over shots that focus directly on said measured object.

Spoiler: EXAMPLES

For Example

Let's say we intend to measure a head of a dragon. It should be wrong to compare a shot that focuses more on a far smaller object (a human) to a shot that only focuses on the dragon itself and depicts it higher than any building or any mountain and seen from beyond the horizon simply because the dragon is in the background of the former.

Let's say we intend to measure the size of this cloud split, It should be wrong to use shots that focus on smaller objects compared to the cloud split to undermine a shot that focuses on the cloud split itself.

These Mangakas are only human, we can't expect them to keep drawing to scale while keeping their story meaningful.

From @Therefir, “
The only reason human-sized characters are still visible compared to supergiants like Sage Centipede is because of visibility, it would look extremely weird if an unrecognizable dot was fighting a massive creature that can dwarf cities.

So the author has no choice other than shrinking the monster/making the human-sized character much bigger in the panel, even if the creature dwarves cities and islands like Sage Centipede and the Dragons from Black Clover.”

---

THE FINAL OPTION

In the event the sizes are still deemed inconsistent after enacting the aforementioned policy, I propose that we should measure several acceptable shots of the same object and simply average each result as the final compromise.

The acceptable shots must be detailed and focused on the object we intend to measure.

---

I don't believe we should arbitrarily ignore feats based on the consistency of their size alone. Mangakas/Authors are on a strict deadline and many of the popular ones are often overwhelmed, so certain margins of error when it comes to size portrayals are to be expected. Averaging reasonable portrayals should be our final option instead. I also believe that this should be applied to any form of graphic media.

So please I know some of us already follow this as I discussed with a few CGMs, but this needs to be official.

---

Agree: CloverDragon03, DarkDragonMedeus, Therefir, KLOL, Deceived, M3X, DemonGodMitch, Armorchompy, DMUA, TheRustyOne, Dalesean, Possibly Jasonsith, Damage,
Neutral: Mr. Crabs
Disagree: Don’tTalkDT


Feel free to suggest how the rule addition should be worded.

 

[CloverDragon03]

I've always been in agreement with something like this. It just makes a lot of sense and I really don't like the idea of using inevitabilities like this to just completely ignore feats

It should also be noted, though, that stated sizes will always take the highest priority

 

[Arnoldstone18]

It should also be noted, though, that stated sizes will always take the highest priority

I agree.

 

[DemonGodMitchAubin]

Following to see where it goes

 

[DarkDragonMedeus]

I am confused, what's the summary on what is being proposed? Something about size inconsistency yes, but on which end does it tend to ruin calculations?

 

[Godernet]

I am confused, what's the summary on what is being proposed? Something about size inconsistency yes, but on which end does it tend to ruin calculations?

Some verses have feats that are disregarded due to size inconsistencies of attacks, characters, etc.(Black Clover, One Piece, and Naruto, to name a few notable ones.)

The OP is proposing that if there is no author-confirmed size for something and the subject of the calc is shown as inconsistent in different panels, that we go over acceptable shots and average out the values so feats aren't completely disregarded.

 

[CloverDragon03]

Some verses have feats that are disregarded due to size inconsistencies of attacks, characters, etc.(Black Clover, OP, and Naruto, to name a few notable ones.)

The OP is proposing that if there is no author-confirmed size for something and the subject of the calc is shown as inconsistent in different panels,that we measure acceptable shots and average out the value so feats aren't completely disregarded.

To add to this, taking an average is only the last ditch effort. In most cases, we should take a panel where the large size object in question is the focus rather than using instances of it being a minor detail as proof against its merit

 

[DarkDragonMedeus]

Fair enough then. I agree with that then.

 

[Arnoldstone18]

Thank you!!

 

[Antvasima]

@AKM sama @DontTalkDT @Celestial_Pegasus @Wokistan @Ultima_Reality @Mr._Bambu @Elizhaa @Qawsedf234 @ByAsura @Sir_Ovens @Damage3245 @Starter_Pack @Abstractions @LordGriffin1000 @Colonel_Krukov @SamanPatou @GyroNutz @Firestorm808 @Everything12 @Maverick_Zero_X @Executor_N0 @Spinosaurus75DinosaurFan @Therefir @DMUA @Jasonsith @Armorchompy @Migue79 @Psychomaster35 @KLOL506 @M3X_2.0 @Dark-Carioca @AbaddonTheDisappointment @Aguywhodoesthings @Dalesean027 @DemiiPowa

What do you think about this?

 

[Damage3245]

This feels a bit relevant to a thread I've got going on at the moment...

Let's say we intend to measure the size of this cloud split, It should be wrong to use shots that focus on smaller objects compared to the cloud split to undermine a shot that focuses on the cloud split itself.

As I brought up in my thread on this specific feat, I can't agree with that. I intend to continue arguing my point over there soon.

To add to this, taking an average is only the last ditch effort. In most cases, we should take a panel where the large size object in question is the focus rather than using instances of it being a minor detail as proof against its merit

I feel like this approach shows an inherent bias if you're assuming that the object is "large size" to begin with, you're prioritizing panels where the object is shown to be large over perfectly valid scans that show that not to be the case.



Ultimately I feel like this kind of policy would just be used to dismiss evidence that would be perfectly valid otherwise to use IMO. Using this rule you can just dismiss someone else's argument out of hand without considering it because "it's not the focus of the panel".

 

[Suigetsuhyugs]

I would like to hear the opinion regarding something that is relevant to this.

Currently the size of onigashima from One Piece is ridiculously high they took 4 or 5 different scans using objects that are in the background to calc one of its horns to have a width bigger than Marineford the horn is decorative and its much bigger than its width.
There was at least one thread talking about how inconsistent it was and there are other ways to reach a different more reasonable value.
Their counter argument is that the sizes of the characters change so going for background objects is "more accurate".

 

[Therefir]

I completely agree with this sentiment, panels where the object in question is the main focus should have priority over ones where they are not.

 

[M3X_2.0]

I always thought we used the panel from the feat itself and not something pages or chapters later.

Though I have to say, unless something has been shown to be really, really inconsistent, like for example Armin's Colossal Titan, wich has 50m but looks more like 200m or more, we shouldnt call it inconsistent and ditch the feat.

We're too fanatics with this. We calc every single panel and every single pixel to find a size and if the size isn't consistent across the panels we say the author is shit and can't draw. If you have any draw and panelling notion you know this isn't possible and should stop bitching about this. Unless this is a live action or videogame where the actor or model size keeps the same size all the time.

 

[Deceived3596]

I would've thought this was common sense tbh. Of course shots which are intended by the author to visually show the scale of a certain object, structure or entity would hold more precedence over close up shots which are forced to cram everything onto a much smaller and concentrated page just for coherency sake, making author intended large sized things look much smaller in the process.

Agree.

 

[KLOL506]

I agree with Therefir and Clover that using the scan in which the feat happens is usually the best place to calculate the size from where the feat itself is happening. At least for page-based or animation-based material. Video games are less likely to fall under that because usually the model sizes remain constant there and there is little to no chance of it changing in the long run.

 

[Arnoldstone18]

Though I have to say, unless something has been shown to be really, really inconsistent, like for example Armin's Colossal Titan, wich has 50m but looks more like 200m or more, we shouldnt call it inconsistent and ditch the feat.

Yeah, let's just all keep in mind that canon sizes take the most precedence to prevent stuff like this from happening.

but If the colossal was never given a canon height and its depiction across multiple shots focused on it were still inconsistent to inane proportions then they should just be all measured and averaged.

 

[Jasonsith]

Yeah, let's just all keep in mind that canon sizes take the most precedence to prevent stuff like this from happening.

but If the colossal was never given a canon height and its depiction across multiple shots focused on it were still inconsistent to inane proportions then they should just be all measured and averaged.

I would further add that some outliers may need to be discarded on a case by case basis.

 

[Nierre]

Of course stated sizes >>>>>>> any and everything else although there might be some outliers but I guess that should be case by case.

As for a scenario with no stated sizes, I definitely agree with using the shot/page that actually depicts the feat in question - instead of using smaller objects/different perspectives to "undermine" the feat in question. For scenarios like this, I think it's perfectly fine with using multiple ends imo. Like each perspective shot having different ends - and then the supporters/CGM discussing which ones should be valid.

I don't really agree with averaging because it doesn't really solve the issue imo. I believe the shots should be discussed with the supporters, cgms, and staff. Using that MHA feat as example, I'd personally choose the shot where we actually see Star and Stripe's hand right before the cloud split because it is the clear focus of the page - rather than the other perspective shot from far away which doesn't focus on the feat but rather we're seeing an object's perspective on the feat.

 

[CloverDragon03]

I feel like this approach shows an inherent bias if you're assuming that the object is "large size" to begin with, you're prioritizing panels where the object is shown to be large over perfectly valid scans that show that not to be the case.

Well that’s only really a problem if you start claiming a large sized object isn’t actually a large size. And my point is that those “perfectly valid” scans are not actually valid to begin with

Ultimately I feel like this kind of policy would just be used to dismiss evidence that would be perfectly valid otherwise to use IMO. Using this rule you can just dismiss someone else's argument out of hand without considering it because "it's not the focus of the panel".

Well just as I’ve said before, the point is that such “evidence” should not actually be valid. Using an unfortunate inevitability as an actual counterpoint should not be a thing to begin with

So yes, such “evidence” can and should be dismissed

 

[Qawsedf234]

Ultimately I feel like this kind of policy would just be used to dismiss evidence that would be perfectly valid otherwise to use IMO. Using this rule you can just dismiss someone else's argument out of hand without considering it because "it's not the focus of the panel"

I was initially leaning towards agreeing, but youbmake a good point here. Ultimately this is just making confirmation bias a rule. A thing I think is big should be big.

Well just as I’ve said before, the point is that such “evidence” should not actually be valid.

I disagree here. If we're saying that authors can't do scale then we shouldn't assume a bigger size is better over a smaller size. Authors can't death and scale is a common trope for exactly that reason. They over or underestimate stuff and to dismiss a low end because we don't like it shouldn't work in my view.

 

[Qawsedf234]

I always thought we used the panel from the feat itself and not something pages or chapters later.

There's plenty of country size calcs that leap frog off multiple planet's or different maps aren't there?

 

[M3X_2.0]

what

 

[CloverDragon03]

I disagree here. If we're saying that authors can't do scale then we shouldn't assume a bigger size is better over a smaller size. Authors can't death and scale is a common trope for exactly that reason. They over or underestimate stuff and to dismiss a low end because we don't like it shouldn't work in my view.

A full-on panel where the focus is the large-sized object should always take priority over random panels where it’s significantly less of a focus, as the former is far more likely to represent the intended size. It also, again, prevents the willful ignorance of blatant feats.

 

[Damage3245]

A full-on panel where the focus is the large-sized object should always take priority over random panels where it’s significantly less of a focus, as the former is far more likely to represent the intended size. It also, again, prevents the willful ignorance of blatant feats.

The solution to ignoring blatant feats is not too willfully ignore all counter-evidence either.

We can't get inside author's heads and just assume that a single panel in particular represents the authors intended vision for the size, especially if they draw it differently to that on multiple other occasions.

 

[CloverDragon03]

The solution to ignoring blatant feats is not too willfully ignore all counter-evidence either.

That would be the case for actual counter-evidence...

Which this is not

We can't get inside author's heads and just assume that a single panel in particular represents the authors intended vision for the size, especially if they draw it differently to that on multiple other occasions.

You can at least make logical guesses, some which have more actual logic to them than others. A big panel focusing on the object, for example, is far more likely to represent its intended size given that it's the focus of it and thus doesn't need to concern anything else. Other panels that focus on other things need to account for said things

 

[TheRustyOne]

I disagree here. If we're saying that authors can't do scale then we shouldn't assume a bigger size is better over a smaller size. Authors can't death and scale is a common trope for exactly that reason. They over or underestimate stuff and to dismiss a low end because we don't like it shouldn't work in my view.

I don't think anyone is saying the bigger size is better. I know a few feats were the more detailed panel is smaller in fact. Smaller panel doesn't mean bigger object.

In fact I think that's just as common as the reverse.

But I disagree if you're saying a bigger and more detailed panel isn't better than a smaller and less detailed panel, where the object in question isn't even in focus.

 

[Qawsedf234]

But I disagree if you're saying a bigger and more detailed panel isn't better than a smaller and less detailed panel, where the object in question isn't even in focus.

I'm saying that we shouldn't dismiss panels of smaller sizes or when the object isn't fully in the foreground under the assumption that a big thing should bi big.

 

[CloverDragon03]

I'm saying that we shouldn't dismiss panels of smaller sizes or when the object isn't fully in the foreground under the assumption that a big thing should bi big.

Well it depends. As mentioned by TheRustyOne, a more detailed panel of the thing in question should take precedence as it's being given full attention.

But also... for example, if the big object is stated to be big, then it should be big

 

[Damage3245]

Well it depends. As mentioned by TheRustyOne, a more detailed panel of the thing in question should take precedence as it's being given full attention.

But also... for example, if the big object is stated to be big, then it should be big

Statements and canon information take priority, yeah.

 

[CloverDragon03]

Statements and canon information take priority, yeah.

Well yes but not just direct size statements but also if characters talk about how big the object is, clearly it's meant to be big

 

[Dalesean027]

Statements and canon information take priority, yeah.

Damage I don't believe they're saying statements and canon information don't have priority I'm pretty sure we agree those have the upmost priority as it'll help determine something if the sizes are consistently off looking HOWEVER in cases where that information isn't provided at all and we're to take the size of the presented feat or showcase, it'd be best to not try and use other panels where there'd be inconsistencies in size rather the the more detailed spreads or panels that the feat or large sized being or object itself is presented in when it takes priority.


Imagine now a case where there's a creature stated to be "as big as a mountain" and in the more detailed panels or spreads wherebirs the focus its depicted as being massively large and is seen even over the horizon but then suddenly when interacting with human panels where those humans are the focus and ot suddenly looks a smaller and is plagued by inconsistencies in size, I'd always be more inclined to take the feat itself and large sized thing when its the focus itself and is presented more detailed and large than when that same thing suddenly shrinks to accommodate for human characters.

 

[DontTalkDT]

I disagree with the OP. I feel like in the whole hierarchy of things to determine which scaling is best to use, this is the least important thing to focus on.

Like, in the example brought up, the issue that the original panel has no actually good measuring stick (cloud height should only be used as a last resort, as it's immensely inconsistent) takes absolute priority to such a sentiment.

 

[Dalesean027]

We literally dealt with this same thing with Asura's Wrath where people tried using some inconsistencies in size to invalidate feats yet those people had never touched a model or did any work with 3d game assets in their life yet tried to argue with this same kinda stuff like Wyzen's (a guy who's as large as Earth) finger when shown in comparison in Earth is as large as a country yet when it switches scenes to Asura's perspective it's only as large as a couple of small mountain ranges.

Yet people complained that the because the game had inconsistencies like that it should invalidate taking the higher end interpertations yet as per my point before these people have never once touched a 3d model or game asset to know that for something like Wyzens poke you aren't and literally cannot model an entire country's length of distance for a finger poke and neither was the actual model for wyzen or the Earth as large as the real one because that'd be an impossible task in game development costing an unseemly amount of resources for one scene so its unrealistic to complain about size discrepancies in situations like that.

And this is just for the 3d side of things so now translate this over to 2d mediums like manga and comics and you'll see its kinda foolish to try and invalidate every large sized feat by this kind of logic but that said there should be some level of scrutiny with some of this stuff and we can't always otp to big size go brrrr as that'd be disingenuous.

 

[Damage3245]

I disagree with the OP. I feel like in the whole hierarchy of things to determine which scaling is best to use, this is the least important thing to focus on.

Like, in the example brought up, the issue that the original panel has no actually good measuring stick (cloud height should only be used as a last resort, as it's immensely inconsistent) takes absolute priority to such a sentiment.

Agreed. In the hierarchy of determining accurate size scaling, the important part to focus on is consistency.

Highlighting a specific panel and excluding all others is just removing contradicting data points to create an illusion of false consistency. If you always make your sample size equal to 1.... then of course there's going to be nothing else that can contradict it.

 

[CloverDragon03]

I disagree with the OP. I feel like in the whole hierarchy of things to determine which scaling is best to use, this is the least important thing to focus on.

Like, in the example brought up, the issue that the original panel has no actually good measuring stick (cloud height should only be used as a last resort, as it's immensely inconsistent) takes absolute priority to such a sentiment.

I don't think I follow. Why shouldn't we prioritize a focused shot of the object in question over random panels where it's essentially an afterthought or times where it needs to account for the presence of human beings?

In my opinion, this really shouldn't be controversial.

Agreed. In the hierarchy of determining accurate size scaling, the important part to focus on is consistency.

Highlighting a specific panel and excluding all others is just removing contradicting data points to create an illusion of false consistency. If you always make your sample size equal to 1.... then of course there's going to be nothing else that can contradict it.

You always talk of consistency yet never acknowledge the possibility of something being consistently wrong. Such a term exists, after all, and focused shots are always going to be most indicative of intended portrayal, regardless of what other, less focused shots show.

 

[Damage3245]

You always talk of consistency yet never acknowledge the possibility of something being consistently wrong. Such a term exists, after all, and focused shots are always going to be most indicative of intended portrayal, regardless of what other, less focused shots show.

I acknowledge that can be the case but to make that kind of declaration we'd need objective information to compare it to. Taking the cloud dispersal feat for example, if we had a canon statement that the size of the hole was 20 km, but we measure it ourselves and get figures like 4.7 km, 5.2 km, 8 km.... Then yeah, the artwork would be consistently wrong compared to what it is supposed to be.

But we're almost never operating with objective information here. We do have to rely on some basic assumptions like things drawn in the manga are supposed to be consistent in size elsewhere in the manga unless they've had reason to change.

Assuming that a object being focused on in panel means it is the only data point that can be used to detirmine its size isn't something I agree with. And this can vary with art style a bit too. Sometimes things being highlighted in a panel makes them more inaccurate and appear larger than they're supposed to be, see Fist of the North Star for example or certain chapters of One Punch Man.

 

[DontTalkDT]

I don't think I follow. Why shouldn't we prioritize a focused shot of the object in question over random panels where it's essentially an afterthought or times where it needs to account for the presence of human beings?

In my opinion, this really shouldn't be controversial.

It wouldn't be (very) controversial if the shots were otherwise completely equal. But they're not.

It being a more direct shot takes the least priority. The fact that you can use pixel scaling from something that we know the size of, instead of having to make a guess based on unreliable cloud size, takes priority as it's the better scaling method.
To be clear: I'm not saying to use the hole size from later panels. Not due to the stuff OP says, but because the hole size can change with time.
But it makes more sense to scale "laser/giant -> cloud thickness -> from cloud thickness this panel to get hole size" rather than just using cloud height and then the panel.

 

[CloverDragon03]

I acknowledge that can be the case but to make that kind of declaration we'd need objective information to compare it to. Taking the cloud dispersal feat for example, if we had a canon statement that the size of the hole was 20 km, but we measure it ourselves and get figures like 4.7 km, 5.2 km, 8 km.... Then yeah, the artwork would be consistently wrong compared to what it is supposed to be.

But we're almost never operating with objective information here. We do have to rely on some basic assumptions like things drawn in the manga are supposed to be consistent in size elsewhere in the manga unless they've had reason to change.

Assuming that a object being focused on in panel means it is the only data point that can be used to detirmine its size isn't something I agree with. And this can vary with art style a bit too. Sometimes things being highlighted in a panel makes them more inaccurate and appear larger than they're supposed to be, see Fist of the North Star for example or certain chapters of One Punch Man.

Well there's a difference between a focused shot and an exaggerated shot. A focused shot, as the name suggests, mainly shows the full scope of something. An exaggerated shot, as the name implies, exaggerates the sizes of one thing or another for higher emphasis while still showing other objects. In a sense, you could say they're the opposite of shots where the large sized object needs to account for human beings and thus the size gets reduced.

Now, I know you're going to then ask: "Well how can you distinguish a focused shot from an exaggerated shot?" And to that I say, the biggest giveaway should be whether or not other objects are present in the shot. Focused shots usually answer this with "no," while exaggerated shots typically answer "yes." Another big giveaway is how close these exaggerated objects are to the PoV, such as when Star and Stripe claps her hands. This is an exaggerated view of her clapping her hands to showcase the action itself, further supported by the fact that we have a canon height of Star and Stripe at 1.93 meters. Unless her hands suddenly became hundreds of kilometers long somehow.

 

[CloverDragon03]

It wouldn't be (very) controversial if the shots were otherwise completely equal. But they're not.

It being a more direct shot takes the least priority. The fact that you can use pixel scaling from something that we know the size of, instead of having to make a guess based on unreliable cloud size, takes priority as it's the better scaling method.
To be clear: I'm not saying to use the hole size from later panels. Not due to the stuff OP says, but because the hole size can change with time.
But it makes more sense to scale "laser/giant -> cloud thickness -> from cloud thickness this panel to get hole size" rather than just using cloud height and then the panel.

If you're not saying to use the hole size from later panels, then we're actually in agreement here. We did in fact adjust the calc to get an actually pixel scaled cloud altitude, and the calc was adjusted as a result. That much is fine.

But now I'm talking about generally when it comes to getting the size of things. I find it a rather dubious practice to use shots that need to account for the presence of human beings for one reason or another as a reason to either drastically reduce the size of the object or just discard the calc entirely