One Piece: ANOTHER Dressrosa Size Revision

[noninho]

Ngl I'm kinda hyped for Damage's answer and I'm remembering that day that a guy posted a Google doc with 8000 words for a Bleach thread answer lmao

 

[Damage3245]

Ngl I'm kinda hyped for Damage's answer and I'm remembering that day that a guy posted a Google doc with 8000 words for a Bleach thread answer lmao

I appreciate that, but please don't fill this thread with posts that do not contribute anything to it.

Off-topic posts will be deleted.

 

[SnookB]

I know calcs are probably being made in the background, but...

Why the **** are we relying on pixescaling instead of just>kinemon statement of barely seeing the target+a clear shot of window perspective

There's literally no reason to do anything else because it's a waste of time and using further out shots of the palace for anything but angling will result in a massively lowballed result.
There's been an entire thread about why using zoomed out shots for scale is unfair due to certain things being drawn bigger just for visibility, and I got a feeling we'll be doing that here.

 

[Damage3245]

@SnookB Apologies if I don't respond to your post right away. I have limited time available and trying to make sure I respond to KT's arguments first.

@KingTempest Apologies for the delay in getting a response up. Hope you don't mind waiting while I work on a response. I haven't forgotten about the thread.

 

[KingTempest]

No rush

 

[Damage3245]

I have a larger post being prepared as a response to the OP, but in order to avoid it getting bogged down with multiple topics, I'll address first the counter-argument raised against the current calc.

I'm heavily against this

First of all, it's common knowledge that there are times that Oda flat out stretches the size of characters in some unclear panels just to note where they are in reference to things, but that isn't the main debunk. It's the support.

No artist is perfectly consistent. I think we all know that not every single individual panel of the manga is going to be appropriate for size scaling.

However instead of dismissing the panel out of hand, what we need is evidence that can be looked at to judge whether this panel is unsuitable or not.

You can't even tell that's Trebol in that Panel. It's so unclear that it's crazy.

I think we can safely judge that is Trebol in the panel. The figure in the panel there is holding a staff, which Trebol is in the scene. Oda included that detail which means we can say that he drew Trebol there.

It's like using this panel and saying it's good to calc Big Mom from it all cause "you can see her".

Or this of Kaidou and Yamato

Oda does this a lot. Get characters and "enlarges" them just to note where they're at in reference to large structures.

Here's the thing; you've provided panels of other instances of Oda's artwork and the reason why we don't use those particular panels isn't just because the panels are "zoomed out" or the characters are not filling the panel, it is because we have other views of these structures or locations that we can use to contradict these individual panels.

You haven't provided any other view of the royal palace's central tower here that would suggest against the size of it being 67.9 meters across.

You claim that Trebol's height is being "stretched out" here. What size do you propose the central tower should be that would prove that Trebol's height isn't being depicted accurately here?

It's so bad we can't even see the character's clothing color.

We don't need to see the character's clothing color in order to scale off of their height.


Unless there is actual evidence against using Trebol's height to scale the size of the tower as being the size it is currently accepted to be, I don't believe the currently used calc should be withdrawn from consideration.

 

[KingTempest]

Here's the thing; you've provided panels of other instances of Oda's artwork and the reason why we don't use those particular panels isn't just because the panels are "zoomed out" or the characters are not filling the panel, it is because we have other views of these structures or locations that we can use to contradict these individual panels.

You haven't provided any other view of the royal palace's central tower here that would suggest against the size of it being 67.9 meters across.

Ok

 

[Damage3245]

Ok

For the record, I'll keep an open mind about Dressrosa's size and I'm open to discussing alternative options for it. I'm not saying that the current calc has to be kept above all else.

At present I'm just arguing that it be kept in consideration instead of being dismissed entirely. I'm not currently convinced that the calc is outright too flawed to be used.

 

[Deceived3596]

I have some contentions with Damage's response, I'll address them later.

 

[Arc7Kuroi]

I'll address them later

Make this your signature

 

[Deceived3596]

Make this your signature

I'll do it later.

 

[Damage3245]

I have some contentions with Damage's response, I'll address them later.

No worries, take your time.

But the main gist of my post is that I'm just asking for more proof which I don't think is unreasonable.

 

[Kachon123]

Bump

 

[KingTempest]

Bump

I'm working

 

[Arc7Kuroi]

I'm working

Hi working, I'm Arc

 

[KingTempest]

mannn

 

[Damage3245]

Since the main contention with the original method seems to be using a character's height from a zoomed out panel as the basis for the scaling. I'll try using something not based on any character's height.

Based on the pixelscaling KT has already done, the height of the windows of the royal palace are around 2.25 meters tall:

Window Pane = 39 px = 0.4 m

Full Window Height = 220 px = 2.2564102564103 m

---

We can use the height of the windows to find the height of the palace walls themselves:

Windows = 5 px = 2.2564102564103 m

Wall = 130 px = 58.6666666667 m

---

And with the height of the palace walls, we can find the diameter of Dressrosa:

Palace Walls = 13 px = 58.6666666667 m

Dressrosa Width = 2028 px = 9152 m


This method only involves three steps of pixelscaling and isn't dependent on Oda's depiction of any of the characters.

 

[Arc7Kuroi]

I type this post up knowing it will do nothing but lead to Nierre and Deceived making art of the deal jokes. However, I truly do think both sides bring up legitimate points that have me not 100% swayed to either side as objective. I'm not going to hanker much on the pixel scaling size stuff of all the affected calcs, as I feel like in this issue, that borders on the equivalent of a semantics debate for calcs. But let me explain.

Regarding the King7Tempest claims, I agree with the notion that it is more than likely that Kinemon's bitch ass cannot resolve the windows with eyes. The statement that he can barely see the castle, which is definitively much larger than the window leads me to believe that there is a highly likely probability that to him he probably cannot see the windows, and thus they would be at most 0.02 degrees in angular size. I also don't believe that narratively speaking a ~30 km Dressrosa is inherently inconsistent either. Being that Dressrosa is based off of Spain, and tmk Dressrosa is an island nation with multiple cities. So there being ~30 km of length to fit a handful of ~Spanish cities isn't crazy to think of at all. According to google Madrid is like ~600 km^2, and a 30 km Dressrosa is like ~3000 km^2, so it could fit like 5 of em.

That being said, I think Dam7age brings up some valid points as well. According to our calculation guide, angular sizing is explicitly less reliable than pixel scaling, at least by our standards. So, looking at it through that lens, Damage's calc using the same window method but strictly sticking to pixel scaling is likely more reliable as it adheres more to our standards. This isn't to say that I think KT's version is inherently incorrect now, but merely a comment on our standards. Furthermore, I do believe matters of consistency should be taken into consideration with stuff like this, in which case it seems like there are more calcs that aren't bad hanging around less than 10 km rather than near 30 km.

All of that being what it is, I think there's a few ways we could go about reaching an acceptable compromise. We could average all the acceptable ends, I do this a lot with size related calcs of my own that have different depictions showing slightly different sizes. I believe this works well to capture the valid aspects of all ends while adequately addressing the skepticism for all ends as well, in most cases at least. There is also the option of doing stuff like "Damage's size, possibly KT's size" as well in the advent that it ends up being decided that both ends aren't that far off in validity, compromise memes aside. At this point, I don't believe I can fully support the ~30 km size as an objective full on rating, but I do think a sort of compromised middle ground could be a fine way of concluding this, even if atm our standards have me leaning slightly more in Damage's favor.

 

[Damage3245]

Being that Dressrosa is based off of Spain, and tmk Dressrosa is an island nation with multiple cities.

To clarify, Dressrosa does not have multiple cities inside of it, but multiple towns such as Acacia and Sevio.

And having a Spanish aesthetic / culture is not necessarily indicative of the island being the size of Spain or its towns being the size of any of Spain's cities.

Also, considering that an island such as Manhattan has 53 distinct neighbourhoods within it, Dressrosa does not have to be 30 km in diameter to have 5 or 6 neighboring towns within its borders.

 

[Arc7Kuroi]

Ok that's a fair point, I was under the impression they were cities, but if I'm wrong on that I'll cross that out.

 

[Deceived3596]

I type this post up knowing it will do nothing but lead to Nierre and Deceived making art of the deal jokes

As for the thread itself, I'm neutral on Damage's new calculation. I'll wait until others address it before I give my full opinion on it.

For a short summary now; I'm currently fine with either calculation.

 

[That_moron2]

Could we get some more mods on this thread before it turns into Chernobyl?

 

[Damage3245]

Could we get some more mods on this thread before it turns into Chernobyl?

I plan to summon some more mods soon.

 

[Arc7Kuroi]

Kuchiyose no Jutsu, CGM

 

[Kachon123]

I plan to summon some more mods soon.

 

[SlendVeny]

You could have used a gif of Raizo

 

[Kachon123]

I plan to summon some more mods soon.

I didn't know it took a week to ping staff members.

 

[Damage3245]

I didn't know it took a week to ping staff members.

I became a bit pre-occupied with other threads.

 

[Kachon123]

I became a bit pre-occupied with other threads.

Now that you're here though let's get the ball rolling again

 

[KingTempest]

I got some stuff in the back rolling

But in the meantime I was asked to ping these guys

@Dalesean027 @DemonGodMitchAubin @CloverDragon03 @Executor_N0
And specifically the one mf who hates HST
@KLOL506

 

[CloverDragon03]

And specifically the one mf who hates HST
@KLOL506

That's TOUGH

 

[DemonGodMitchAubin]

This is fine

 

[Damage3245]

This is fine

Thanks for responding to the thread.

Sometime tomorrow I'll try to have summary post up regarding why I don't think the calc in the OP should be the one that we use.

 

[CloverDragon03]

That's TOUGH

Hold up I could've sworn I gave my stance on this thread-

Well, clearly I didn't, so I'm here to say I'm siding with KT's version atm

 

[Damage3245]

Hold up I could've sworn I gave my stance on this thread-

Well, clearly I didn't, so I'm here to say I'm siding with KT's version atm

I was wondering why you hadn't commented already, haha. Thanks for leaving an evaluation.

 

[KingTempest]

Sometime tomorrow I'll try to have summary post up regarding why I don't think the calc in the OP should be the one that we use.

ain’t you already do that?

 

[Damage3245]

ain’t you already do that?

I posted some arguments in the thread, but haven't gotten into a proper comparison summary, which I'm working on now - albeit I've been out for most of the day so haven't had a lot of spare time to work on it yet.

EDIT: Not forgotten about the thread btw; just out of time tonight and will get my response up in the morning.

 

[Dalesean027]

The OP seems fine to me

 

[Damage3245]

The first thing I want to highlight is a comparison of the method's between this calculation and the calculation in the OP.

Both of them use the same foundation; the window pane in the windows of the outer palace walls being stated to be 40 cm tall.

Both versions essentially use three steps:

My version:

1) Pixelscales the window pane to the full height of the window.
2) Pixelscales the window to the height of the wall.
3) Pixelscales the height of the wall to the diameter of Dressrosa.

KT's version:

1) Pixelscales the window pane to the full height of the window.
2) Angsizes the distance between the plateau and the window on the palace walls.
3) Pixelscale the above distance to the diameter of Dressrosa.

However, KT's version depends on the assumption that the height of the window is beyond the angular resolution of 0.02 degrees.

And in our Calculation Guide, it is mentioned that angsizing is a less reliable method than pixelscaling:

A less reliable, but often useful technique is angular sizing or angsizing for short.

That page also mentions:

As a general guideline, it's preferable to use a scaling where the measuring stick used is of similar size to the object being scaled, rather than a scaling where the measuring stick and object are of very different sizes. However, this guideline should not take priority over other criteria that may cause similar or greater uncertainty in the scaling.

For instance, one method may involve more scaling steps than another, or one method may use angsizing while another uses exclusively pixel scaling. Additionally, the size of one measuring stick may be much less certain than the other or the perspective in one method may prevent precise results. The general consistency of the sizes obtained by various methods should also be taken into account.

The inclusion of this line indicates to me that a method that involves angsizing is generally less reliable than a method that uses exclusively pixel scaling.

But I'm not solely going to fixate on the difference between these two methods. If the requirement of the assumption for KT's version and the usage of angsizing over just pixelscaling isn't enough of a concern, then the second point I want address is what is mentioned at the bottom of that quoted section:

The general consistency of the sizes obtained by various methods should also be taken into account.

My version produces a size of 9152 m, and KT's version produces a size of 28996 m.

Do any other methods produce any results close to either of these two values?

Well, I've tried three alternative methods here, using either pixelscaling or a mixture of pixelscaling and angsizing, that give values of 7503 m, 9835 m, and 11393 m, averaging out to 9577 m. Each of the results is a whole lot closer to my version of the calc than the OP's version.

As far as consistency goes, I believe that the lower end result for Dressrosa's size here is more heavily supported than the higher end 28996 meter end in the OP.

I think there is enough reason to not consider the OP's version to be the most ideal method, and from a consistency standpoint, the OP's value for Dressrosa's size is the least supported. The list of angles that KT links in the OP as support for his version only provides distances of a several hundreds of meters between the plateau and the royal palace walls; not the 6+ km in his primary calc.

What do you think about this? @DemonGodMitchAubin @CloverDragon03 @Dalesean027 So far you've just voted that the OP seems fine, but what do you think is the ideal solution based on our site standards?

 

[KingTempest]

Hey, that's weird.

Not to throw shots at you but it seems like a lot of your calcs just produce, hilariously low results. Consistently low results if I say so myself.

Let me fix that for you.

So let me break down your calcs.

Spoiler: Calc 1

Also, your whole point is that "angsizing isn't very accurate, but I'm gonna angsize for a calc to go against yours". That's cute.

So your first calc uses a very bad panel for the window, showing a window we can barely see that seems stretched out just so we can see the windows.

If you were to substitute that with this panel with this one, showing the window in a much clearer light, even showing the panes clearly.

Window Pane Height: 0.4 m = 1 px
Minimum Height of Palace: 267 px = 106.8 m

So that alone cuts out a whole step of pixelscaling from your first calculation, just making it
"Pane to wall"
"Angsize"
"Distance to dressrosa"

And if I actually follow through with that instead of just showing you your comedically small wall, I'd get the distance between them to be 2817.2 meters.

You'd probably say "1 pixel doesn't make sense", which is why I zoomed in and screenshotted a pane across the screen showing that the pixel above is 1 bar, and the pixel below is another bar, showing that this is a bar.

And if I use that to find Dressrosa with your scan in the other blog using the 13 (palace height) to 2028 (dressrosa), it'd be 16660.8 meters.
Which is really nice as support for mine.

The funny parts about this calc, the calculation showcases a pane that's so small that we can't even see a clear panel of the window pane because the others are affecting it. On top of that, I didn't even get a full window.
So if I wanted to say my slightly smaller window pane was slightly less than a pixel and the rest of the wall got me, hmm, 4/3rds the size of the wall I just got, it'd bring Dressrosa to 22214.4 meters.

Closer to mine than yours.

Also uses far less steps. The last thread got the calc accepted because it used 2 pixelscaling steps, yet this calc also uses 2 steps, uses a clear disparity in the window sizes, and shows a much bigger difference.

Spoiler: Calc 2

So this calc shows "I have no counter against KT's calculation, nor do I have a counter towards his argument towards my calculation. I will stand on my calc which he has countered, not counter his argument, and reuse it in another message. I have the same amount of steps, and I'm not angsizing."

I don't feel like countering this, so I won't.

Spoiler: Calc 3

This is funny.
Because when this same calculation used the old battleship size, you didn't like it... but now it's smaller and it's one of your main methods?
Yeah nah.

If we look at your calculation here, which this thread spoke about and which DT himself spoke about, we see that your calc is fine, it's just that others utilize small steps.

So now you have a calc which has the same amount of steps, but because prior the large size was too much, but now it's perfectly fine?
Nada.

We see that your own calc using Issho's scaling to the beach uses much clearer panels of the things in comparison, and that would provide a value much further away than the others.

But then we see Kobster's old calculation, finding the full width of the beach from a straighter angle and accounting for slightly more of the width.
27485.92 meters. Very close to my 28996.105714286 meter value don't you say?

Now a heavy thing that I noticed you spoke about was "consistency". Yeah, that's bs.

Because before we worked on "which was a factually better calc", but now it's "what calcs bring a closer value".

So now since we're doing that, I wanted to double it.

You brought 3 calcs ranging from 7km to 11km. I bring to you 6 calcs ranging from 50km to 200km.

With these calculations, some of them use 2 pixelscaling steps, one of them uses 1, some use a few more than wanted.

They bring an average of 135154.221 meters, or 135.154221 km.

And before you go and dissect every single calc, you already know at least 3 of these are getting accepted.

So the question is, do you really want to argue with me of all people about consistency?

Because now we have a motherload of calculations utilizing
A. Less panels and steps (like literally everybody in this thread agreed is better)
B. Much clearer panels.
C. Consistent values