27

I've only noticed this today: when the chemical station appeared in the grey-backed loading screen, there appeared to be a transparent image used on the inside of the Florence flask (the big round one at the corner of the worktable).

I managed to capture an example in-game, but it's far easier to see in the loading screen.

Does anybody know what (or where) the image is of? Or how to find the original in the game data?

To me it looks like a 360 degree image taken on the bank of a river, of a town with white buildings and dark red roofs. It does appear to rotate as you move your character around the flask.

Fallout 4 chemistry station, zoomed in on the Florence flask

The image does not appear to be a reflection of the world around the character, as I took this photo in Diamond City.

So far in my own research, I have found out that the flask type is called a Florence flask (or a boiling flask). Searches on Google images for "fallout 4 chemistry station image in flask" and "fallout 4 chemistry station Florence flask background image" (and others) have not proven very useful.

Joachim
  • 20,337
  • 15
  • 73
  • 103
Skrrp
  • 373
  • 3
  • 5
  • My first thought was the Nakano residence, but after looking at more pictures I don't think its that one. I'll keep looking though. Really wish I knew how to modify/adjust pictures to make them easier to see, like with a TV's brightness and contrast. – WiZΔRD Mar 21 '22 at 05:58

1 Answers1

46

This is the default cubemap for objects so that they 'reflect' outside scenery, hence the reason they 'follow' the player. It is the same one as used for the 'reflection' of the buggy water puddles, and objects like the scopes of guns.

I guess the reason for this being an outside scenery is that these objects are likely to be used more frequently outside.

The cubemap file is called mipblur_DefaultOutside1.dds (this .dds graphics filetype can be decompressed by the GPU [more info here], and is therefor preferable over other graphic image files).
The file can be found in Fallout 4's Data folder, in the file Fallout4 - Textures1.ba2. This file has a folder hierarchy (similar to the loose files in the folder structure in 'Fallout 4/Data'), and the cubemaps are located in 'Textures\Shared\Cubemaps'.

This is the file as opened in Photoshop:

file opened and exported in Photoshop

And this is the file after some transforming, to make it look how cubemaps are supposed to look:

rearrangement of the sides of the cube, so that they form a coherent image

While the image may have elements of Concord or similar towns in the Commonwealth, it is more likely a collage of different views: if you follow the lines of the different sides of the 'cube', you can clearly see blurred edges, strange transitions, and implausible geometry (the patch of grass in the center image is a good example).
In all likelihood the different sides are different screenshots from the (early) game photoshopped together.

Joachim
  • 20,337
  • 15
  • 73
  • 103
  • Fake environment map then? – user1095108 Mar 21 '22 at 12:44
  • @user1095108 Kind of. A cubemap is a type of environment map (the "map" refers to a pre-calculated image, and "fake environment map" seems like a pleonasm to me). Cubemapping (or environment mapping) was the de facto method before being replaced by real-time reflections, because these are much cheaper to calculate (and are still often used on smaller objects, or on objects that only have a vague reflection (old and dirty armour, for example)). – Joachim Mar 21 '22 at 12:52
  • Any reason why, say, a sphere map wasn't used? – user1095108 Mar 21 '22 at 13:15
  • 1
    @user1095108 Cube mapping wiki 2nd par.: "In the majority of cases, cube mapping is preferred over the older method of sphere mapping because it eliminates many of the problems that are inherent in sphere mapping such as image distortion, viewpoint dependency, and computational inefficiency. Also, cube mapping provides a much larger capacity to support real-time rendering of reflections relative to sphere mapping because the combination of inefficiency and viewpoint dependency severely limits the ability of sphere mapping to be applied when..." – anjama Mar 21 '22 at 14:31
  • Note: I am not qualified to judge the accuracy of that quote – anjama Mar 21 '22 at 14:32
  • 5
    @anjama Please do not use code formatting for quotations, or any other non-“code” text. It’s not just a visual highlight, it’s also a semantic marker indicating that the text is code and this can cause alternative browsing technologies (e.g. screen readers for the blind) to render it differently, which may make it harder to understand (may explicitly note punctuation which may be critical to code, there have even been cases of reading letter-by-letter). Quotation marks do their job very well; if you insist, italics or bold are acceptable. – KRyan Mar 21 '22 at 14:40
  • @KRyan Yeah, I was trying to get blockquotes to work initially but they didn't seem to want to work in comments for me, and code formatting was the first thing to come to mind as an alternative. I don't know why I didn't think to just italicize it, especially since I did have italics on my mind for the follow-up note. If I could edit it now, I would. – anjama Mar 21 '22 at 15:05
  • @anjama the environment map is "fake" and distorted as it is, and this is why I asked. – user1095108 Mar 21 '22 at 15:32
  • 1
    @user1095108 Because a sphere map is a two-dimensional image of a sphere, whereas a cubemap is (rendered as) a spherical image of a cube :) Cubemapping is preferred for dynamic reflections, because as the angle at which it is looked at changes, the definition/resolution of the cubemap won't change as much as that of a sphere map. Think of it this way: if you take a photo of a smooth mirror ball and use that photo for a render of a model of that ball, it will rarely look good, whereas projecting a cubemap (from the inside of the ball) will give a more convincing reflection. – Joachim Mar 21 '22 at 15:42
  • @Joachim very good explanation, but my point was, that cursory inspection of the reflection reveals that it is fake (i.e. scenery by a river bank), so it may not matter whether it is convincing, that's all I had in mind. – user1095108 Mar 21 '22 at 16:06
  • @user1095108 Ah. Well, in any case, the behaviour of this type of map is less eye-catching than having a non-spherical map with hard edges (especially since the flask is transparent). – Joachim Mar 21 '22 at 16:56
  • There are clearly levels of fakery in reflections. At one extreme you have full dynamic reflections of the actual current scene at the other extreme you have cases like this where the reflections are totally unrelated to the scene the object sits in. Somewhere in the middle you have reflections based on a simplified version of the scene the object sits in. – Peter Green Mar 23 '22 at 21:07