x

Pixels

 
  • Recommend tagsx
+720
Views: 36698
Favorited: 248
Submitted: 11/02/2013
Share On Facebook
Add to favorites Subscribe to Vadi submit to reddit

Comments(152):

[ 152 comments ]
What do you think? Give us your opinion. Anonymous comments allowed.
User avatar #12 - retrofresh (11/02/2013) [-]
All that was in my mind was... FIND FAMOUS PEOPLE *******
#114 to #12 - buthow (11/03/2013) [-]
I was barely reading because all I thought was that.
User avatar #136 to #12 - womd (11/03/2013) [-]
It was the second you seen the name Emma Watson wasn't it? Me too.
#107 - razorlupus **User deleted account** (11/03/2013) [-]
You can also look at bewbs
You can also look at bewbs
#90 - deletedmyaccount (11/03/2013) [-]
And yet instead of these possibilities, we make gifs like this.
And yet instead of these possibilities, we make gifs like this.
#128 to #90 - steammadewalrus (11/03/2013) [-]
Or this.
Or this.
User avatar #64 - skubasteve (11/02/2013) [-]
Would it show me what it was like before the big bang?
#67 to #64 - anon (11/02/2013) [-]
User avatar #21 to #20 - croski (11/02/2013) [-]
All of the books ever written and all of the books that will be written.
User avatar #22 to #21 - mooproxy (11/02/2013) [-]
Everything. Completely everything.
User avatar #23 to #22 - croski (11/02/2013) [-]
I know, I've been thinking about this for quite some time now...

The possibilities...
User avatar #24 to #23 - mooproxy (11/02/2013) [-]
It strikes me that even though it's non recurring, it must repeat the same digit billions and billions of times consecutively.

The possibilities...
User avatar #27 to #24 - croski (11/02/2013) [-]
I was once trying to find a streak of 5 in I think 10000 digits of Pi. Don't remember the result though...
You could also translate those numbers into binary code and then into pixels and then into, for example, movies. You could find a movie of your birth and of your death as well.

And because it is infinite there would be, at one time, a loop of videos displaying every possible way you can die.

The possibilities...



User avatar #34 to #24 - quotes (11/02/2013) [-]
theres a spot with six 9's in a row
so you memorize to that point and read it off
read hte last 9's and it sounds like pi ended
#44 to #20 - anon (11/02/2013) [-]
It has not been proved that every possible sequence of numbers is contained in pi. It could for example have every possible sequence except 123456789. Non-repeating is not as powerful a property as it may seem.
#70 to #44 - anon (11/02/2013) [-]
every single unique substring of numbers is by definition contained in the digits of any irrational number at some point

it has to occur somewhere. not only that, other properties of pi showing statistically proved randomness guarantee that every substring can be found by a calculable point based on the size of the substring. the statistical randomness and normal distribution of digits is a very significant property that you need to account for here.
User avatar #46 to #44 - mooproxy (11/02/2013) [-]
If it goes on infinitely, then the probability of any sequence occurring is 1.

Direct proof here: en.wikipedia.org/wiki/Infinite_monkey_theorem

#49 to #46 - zobzob (11/02/2013) [-]
Yes, it occurs with probability 1, but unfortunately probability 1 does not guarantee that it will occur.
#71 to #49 - anon (11/02/2013) [-]
it guarantees it will occur because it has a statistically normal distribution of digits

if it didnt, then just being irrational and infinite isn't enough, but this property guarantees it will occur.
absolutely every conceivable substring does occur, and you can calculate a maxima for the point any substring has to occur by
User avatar #60 to #49 - mooproxy (11/02/2013) [-]
Well yes, which is why the strange idea of infinity cannot possibly exist in the real world, only as an abstract mathematical concept.
#51 to #49 - zobzob (11/02/2013) [-]
Sorry, I'm not actually sure if it happens with probability 1 with pi, but even if it did, it still would not imply actual existence.
#8 - cjwers (11/02/2013) [-]
If a program/website were made to continually show every combination of pixels:
Everything from FunnyJunk would show up
Every movie clip would show up
Every porn clip would show up
Even the most bizarre porn clips would show up
The nastiest stuff never thought imaginable would show up
If I had a point, I think I made it.
If not, enjoy the thoughts
#26 - lech (11/02/2013) [-]
Ok, so let's go over this, carefully.
My screen is 1920x1200, but because that's a bit unusual, I'll use 1920x1080 (That's 1920 pixels wide, 1080 pixels tall)
Area of a rectangle is length x width. Meaning, our 1920x1080 monitor has 2,073,600 pixels.
Now, let's look at a single pixel, shall we?
The picture shows that there's usually a red, green, and blue led for each individual pixel. Each color, red, green, or blue, has a number that correlates to the power of the led. That number is between 0 and 255. Because of limitations, computers use binary to represent data. A way of representing data is called hexadecimal. Which is 2^8, or 256. Since 0 can also be a number, we have a system where numbers can only go from 0 to 255.
Since we're dealing with 3 leds (blue, green, and red), we have: 2^8 * 2^8 * 2^8.
Which is 16,777,216 different combinations of the pixels.
Now we have to multiple 16,777,216 with the amount of pixels we have, 2,073,600. This turns out to be 34,789,235,097,600 different combinations. This is almost 35 trillion, but it's NOT infinite. It's a lot of pictures. But 34,789,235,097,601 is bigger than the answer we got.
User avatar #40 to #26 - cryingchicken ONLINE (11/02/2013) [-]
You were right until the end where you multiplied them. You aren't meant to multiply them. you should put the pixel value to the power of the total pixels. The answer would be:


16,777,216^2,073,600

which is some mad number like a guhzillion or some **** .
#122 to #26 - fancyjokes (11/03/2013) [-]
35 trillion is "not" a large number. Computers work at GHz of operations.

1,000,000,000 < 35,000,000,000,000

It would take a 1Ghz computer (simplified concept of actual speed) 35,000 seconds/583 minutes/9.7 hours to find all your images, if your logic made any sense. A "super computer" would probably take 400 ms.

The amount of pictures an 8-bit color system on a 1920x1080 monitor may create is actually:

256^2073600

This, now, is a number so ******* large, there's no physical reference known to man that may take that number to a understandable scale. Actually, only other numbers created under multiverse theories are as large. The largest number being the finite possible universes which is around 10^(10^(10^7))

Funny Fact: 10^82 is the estimated number of atoms in the universe.
Funny Fact 2: A computer trying to create Black and White pictures of a 15x10 monitor would take at a rate of 150Hz 46,138,562,195,008,110,600,774,753,760,087,749,172,181,189,607,929,628,058,548,5 17,099,604,563,033,706,075 years to create all possible images and at 20 PetaHz would still take 46,138,562,195,008,110,600,774,753,760,087,749,172,181,189,607,929,628,058,548,5 17,099,604 years.
User avatar #127 to #122 - lech (11/03/2013) [-]
(As people who have posted before you pointed out, I did have improper logic at the end of it. You do not need to point it out again)
It's not a large number, because you can easily come up with an infinite amount of other numbers.
This is more of a mathematical concept than a real world concept.
If you think you've found the largest number ever possible.
Well, I can always add 1 to it, meaning, that's not the largest number.

Sure, it may be bigger than any physical reference we have. But, with a purely mathematical view point, it's not large at all.
User avatar #147 to #26 - zaxzwim (11/03/2013) [-]
lets say we are dealing with a 32x32 pixel area of only black or white pixels, how many is that?
User avatar #153 to #147 - lech (11/03/2013) [-]
(If you'd continue reading, people pointed out I did a bit of a mistake when I multiplied that out. I was supposed to have one as an exponent)
This many
You need to login to view this link
User avatar #156 to #153 - zaxzwim (11/03/2013) [-]
yes i saw and i thought that you would now have corrected and would have made fixes and would be able to tell me how many
#52 to #26 - anon (11/02/2013) [-]
YOU MULTIPLY NUMBERS OUT OF YOUR ASSHOLE AND GET RETARDED ANSWERS, I CANNOT TRUST A WORD YOU ARE SAYING, BUT JUST TO SHOW YOU THE MAGNITUDE TO WHICH YOU ARE WRONG I LL TELL YA THIS:

WE ASSUME A PIXEL HAS 3 POSSIBLE COLORS
WE HAVE 2073600 PIXELS,
THIS MEANS THERE ARE 3 AT THE POWER OF 2073600 POSSIBLE COMBINATIONS, A NUMBER SO IMMENSE IT ELUDES YOUR WRONGFUL MIND.
User avatar #62 to #26 - zzforrest (11/02/2013) [-]
... OP said "finite" like 30 times did you not catch that?
User avatar #33 to #26 - goldenfairy (11/02/2013) [-]
Uhm... You've got that wrong.

Lets say there are only 2 pixels, each pixel having 16,777,216 colors, 16M for simplicity. According to you there would be 16 * 2 = 32M different combinations.

However lets imagine there is only 1 pixel, with 6 colors. 2^8 * 2^8 * 2^8 * 2^8 * 2^8 * 2^8 = 281,474,976,710,656, which is actually 16M ^ 2 = 256M.

Therefore real answer is 16,777,216 ^ 2,073,600 = 1.5 × 10^14981179.

There are ******* 15 million digits. I don't even know what damn number that is.
User avatar #55 to #33 - lech (11/02/2013) [-]
Oh damn, I was wondering why that number was so small. Thanks!
By the way, I calculated it with Mathematica
You need to login to view this link
#95 to #55 - unholyjebus (11/03/2013) [-]
God dammit... 5 minutes later and my computer is still trying to load that txt file.
User avatar #98 to #95 - lech (11/03/2013) [-]
Open with notepad++
Notepad sucks
#36 to #33 - collegedood ONLINE (11/02/2013) [-]
**collegedood rolled a random image posted in comment #6363641 at Safe For Work Random Board **
you beat me to the correction. Yeah, ****** crazy.
#134 to #33 - themapestree (11/03/2013) [-]
To push this one level further, I tried to figure out how long it would take to display every possible image on a full HD screen.

The standard refresh rate for a TV is 120 HZ, or 120 images a second.

So the math works out to be 16,777,216^2,073,600 (the number of possible images) divided by (120*60*60*24*365), which gets us to images per year if you have the TV running 24/7.

This works out to about 3.9648*10^14981169 years.

For those interested, we would have to have a TV running from the big bang until now 2.8735*10^14981159 times over to see every possible image.

I was going to calculate the case where every HDTV in the world was also running, but it reduces the number so little that it doesn't matter.

tl;dr We will NEVER, EVER experience all possible images that a 1920x1080 TV could show.

Just imagine a 4k TV!
#66 to #33 - parman ONLINE (11/02/2013) [-]
This is the correct answer. But now to the real mind blowing thing.

Let's assume you have older 1024x768 screen.
Total number of pixels is 786432 (as opposed to over 2 millions in case of 1920x1080).

Therefore total number of every possible image displayed is (256^3)^786432 = 8.2 × 10^5681750.

Compare it to 1.5 × 10^14981179 for 1920x1080 screen.
On 1920x1080 res you can display roughly 10^(14981179-5681750) = 10^9299429 times more images than on 1024x768 ! It's frikkin 10 with almost trillion zeros!

BUT. I think we'd all agree that even on smaller screen we could display the very same number of images. The only difference would be quality.

mind=blown
0
#39 to #33 - nexman has deleted their comment [-]
#15 - icametochewgum (11/02/2013) [-]
Another paradox is what's known as "Gabriel's Horn"

If you take the the equation y = 1/x from [1, infinity), and rotate it about the x-axis, you produce an object of infinite surface area, but finite volume.
What this means is that you could fill the horn with a finite amount of paint, but filling the horn with that much paint still would not provide enough paint to cover the outside of the horn.
User avatar #30 to #15 - cheesezhenshi (11/02/2013) [-]
Technically there's still an infinite volume. the asymptote at y=0 means that it would never close, so while the volume is increasing infinitely small at infinity, it's still increasing. Granted, the surface area would become larger than the volume, but that's not really a paradox, it's a well known fact that as you decrease the size of an object the surface area decreases at a slower rate than the area, because surface area is x^2 while area is x^3. It's why cells are tiny instead of huge, they need a certain ratio of volume to surface area. What am I missing that this is a paradox?
#38 to #30 - icametochewgum (11/02/2013) [-]
I'm hoping that this will explain it for you!

We're taking the integral of y=1/x on [1,infinity) and rotating it about the x-axis, creating a 3-D horn-shaped object. We're going to do this by taking the integral from x=1 to x=a where a>1
The volume of that horn is:

V = pi * [ the integral from 1 to a of ] dx / (x^2)
So the volume is:
V = lim(a --> infinity) of [ pi * (1 - (1/a)) ]
Since the domain is x=[1,infinity), the volume of the horn is the limit as "a" goes to infinity of that equation; as "a" increases towards infinity though, the value of (1/a) decreases towards 0. This means that as "a" goes to infinity, (1-(1/a)) goes to 1. So the volume then, as "a" goes to infinity, is:
V = pi * 1 = pi
While pi is irrational, it is definitely finite.
0
#37 to #30 - emrakul has deleted their comment [-]
User avatar #93 to #15 - sidekickman (11/03/2013) [-]
Can you explain how it has finite volume, but infinite surface area? I'm really ******* confused.
#132 to #93 - icametochewgum (11/03/2013) [-]
Try thinking of it this way:
Pick any random interval in the domain of length 1, so say from x=7 to x=8.
The length of the actual curve there (the graph y=1/x) is greater than or equal to 1 (since a line parallel to the x-axis would have a length of one, but a line that had a negative slope would have to still cover a distance of one unit along the x-axis, but the hypotenuse of that triangle - or top of the curve, whatever - would be longer than one)

However, the area underneath the curve there is log(8/7)=0.133531
As the interval moves further along the x-axis (say x=3700 to x=3701), the area under the curve gets smaller and smaller.
However, the curve itself is still adding a unit of length onto it for everyone of those intervals.
So, to use the interval x=3700 to 3701, the length of the curve there is roughly 1, but the area under that curve is 0.000270234.

Hope that helps clarify things!

So as
#104 to #93 - ennemi (11/03/2013) [-]
Just check out koch snowflake. It's in 2D so it's easier to see. This object have a finite area and an infinite perimeter.
User avatar #148 to #104 - sidekickman (11/03/2013) [-]
But I'm still confused, because that line never intersects the axis, so how can it have a finite volume if it never comes to an absolute vertex?
#152 to #148 - ennemi (11/03/2013) [-]
alright, so I don't know if you studied series, but basically, every series like that :

sum(1/n^k) where k > 1 converge, meaning that when n -> infinity, sum(1/n^k) -> C where C is a constant.

Now when you rotate 1/x around the x-axis, and you want to find it's volume from 1 to infinity you get :

pi*int(1/x^2) from 1 to infinity = pi*[1/x] from 1 to infinity = pi * (1/1 - 1/infinity) = pi.

You're right that the line never intersects the axis, but because you have a 1/x^2, the volume added by increasing the x become so small so fast, that for big x value, what they add is nothing compare to when x was of value 1 or 0.5 ( for example when x = 1 000 000, you add only : 0,000 000 000 001. If you want to get a better grasp at the concept without the integral, you should look up convergent and divergent series.
User avatar #154 to #152 - sidekickman (11/03/2013) [-]
OH. I was kind of missing the point, but now I get it. Thanks!
User avatar #17 to #15 - Vadi (11/02/2013) [-]
Please read the first sentence...
User avatar #82 - thedarkestrogue (11/03/2013) [-]
Woo. people figuring out how a screen works. So ******* mind blowing.

I might be a cynical asshole.
User avatar #84 to #82 - listerthepessimist (11/03/2013) [-]
I too am a cynical asshole, but I never thought of it in those terms
User avatar #6 - torchrose (11/02/2013) [-]
And yet we can't view the true color of cyan on a screen

GG technology
User avatar #41 to #6 - zeroqp (11/02/2013) [-]
the irony of this comment is that I wanted to find out what cyan is supposed to look like, and the fastest way to do so is by Googling it. So that I can see it on my computer screen. Where it won't be displayed like it truly is. But still it will look almost(?) perfect.
#48 to #41 - torchrose (11/02/2013) [-]
Well here you go
User avatar #149 to #48 - zeroqp (11/03/2013) [-]
wow, that looked amazing. I didn't think much was going to happen, but as soon as I started moving my head back..! And what a beautiful color.
User avatar #137 - iamgandalfsagbag **User deleted account** (11/03/2013) [-]
im too ******* high for this ****
User avatar #141 to #137 - revengeoftheowl (11/03/2013) [-]
i had to read it three times, and it blew my mind every time
#63 - anon (11/02/2013) [-]
lets cut this down to size
say we're dealing with only black and white images

not only that, lets keep them small, a simple monochrome 16x16 grid, each pixel is either black or white, a simple 256 bit image. It seems EXTREMELY small. But even with this really minimal set up, you still have 2^256 = 1.1579*10^77, or approximately equivalent to ten percent of the number of atoms in the observable universe.

If you had a billion monitors on earth, each displaying a bunch of these possibilities in a 100x100 grid of distinct images (1600x1600) pixels at 60fps, it would take much longer than the lifetime of the universe to display every possible combination. (600,000,000,000 images per second is many orders of magnitude shorter).

you can optimize this down by allowing multiple substrings per line (pixels 0-16, pixels 1-17, etc being allowed, similar to entering a passcode in a stateless lock, where 12341 checks the passcode 1234 as well as the passcode 2341) but that barely makes a dent.
#97 to #63 - haidbz (11/03/2013) [-]
You know, I intended to post something like this, but didn't want to do the math, so thank you for saving me the work.
#69 to #63 - saxong (11/02/2013) [-]
This is the single most well thought out, intelligent thing I've ever seen posted by an anon.
This is the single most well thought out, intelligent thing I've ever seen posted by an anon.
#115 - dkedr (11/03/2013) [-]
Yeah, the number of possibilities is massive though
2 ^ (1920*1080*24)=2 ^ 49766400= google doesn even want to tell me the number so I had to split it up

2^(497) = 4.091738e+149
2^(664) = 7.654505e+199
2^(100) =1.2676506e+30
2^(1000) = 1.071509e+301

At this point I don't really care about all the numbers before the e since they won't make a heck of a difference.

So now we just take the 2^(497) * 2^(1000)*2^(100) and add 2^(664) *2^(1000)
2^(497) * 2^(1000)*2^(100) = 5.557803743e+480 and 2^(664) *2^(100) = 9.703237856e+229
This one here ^ isn't really worth thinking about since it's so much smaller than the other one.
So if you could view a million of these pictures a second and decide if they were usefull or not, it'd take you 5.557803743e+474 seconds or
1.7623680058e+467 years
Even if you take it to a billion pictures a second you'd just shave another e+3 off that e+467, so it'll still take longer than anything really.
Now if we take a smaller screen it could be easier, and reduce the number of colours to make it even easier.
So lets take a 32 x 24 pixel black and white screen, as in no grey, no colours, just black and white. That's still 1.5e+231 possibilities. Will still take ~e+215 years to look through.
Even if we reduce the picture size to just an 8 by 8 pixels, it'll still give you 1.8e+19 possibilities, which will take e+3 years to filter through, and that's a thousanish years.
So it is finite, mathematically, but really, it's infinite.
User avatar #109 - weapsycho (11/03/2013) [-]
emma watson beating my nuts with her third boob
User avatar #139 - goll (11/03/2013) [-]
they have a name for this, it's called photoshop.
#138 - orgasmpirate (11/03/2013) [-]
<MFW
< This screen image, page, and comment are also some combinations.
#76 - icametocomment (11/03/2013) [-]
Dude...
User avatar #43 - sadisticsalmon (11/02/2013) [-]
Yeh, sure Emma watson beating the **** out of a bear on live TV..... Thats totally the first Emma Watson related footage I'd conjure up if I could do that.
[ 152 comments ]
Leave a comment
 Friends (0)