JoeV
Thin Air, Bright Sun
Is photography closed? Are there an infinite number of possible images, implying an infinite wealth of potential creativity yet to be discovered?
How many possible photographs are there?
As a side question, are you old enough to remember the early days of electronic calculators? I do. I had an interesting collection at one time, including ones with red LED displays, others with green fluorescent tube displays that ate batteries like crazy. I'm left with three HPs using RPN logic, and a red LED version once sold in hardware stores. Every phone, tablet and computer has a calculator, these days.
These are the kinds of thought experiments that will challenge even the best scientific calculator. But here goes.
Assume square format, 800 pixels on a side, 8-bit RGB channels. Pretty much the standard image viewed on a phone, tablet or computer.
Total possible photos = (Total Color Levels) ^ (Total Pixel Count)
That's total number of color levels raised to the power of total pixel count. If this isn't obvious, imagine a simple photo of just two pixels, each with three color levels. That's 3^2 or 9 total possible images.
Three 8-bit color channels implies 2^24 color levels, which is: 1.677x10^7.
800x800 pixel image size implies 640,000 pixels total.
Total possible photos = (1.677x10^7) ^ (640,000)
Converting to scientific notation (Google will be of little help here, as are most scientific calculators; you'll have to dust off your high school maths):
Total number of photographs = 5.4075 x 10^4,623,820
That's 5.4 followed by 4,623,823 zeroes.
That's a lot of photographs, each one unique. Many will differ by only one pixel. But it's a finite number of images.
How long would it take to view all those photos, assuming a rate of one photo per second (I know people who surf the Net faster than that)?
First, assume the universe is 14.8 billion years old. That's 4.667 x 10^17 seconds.
Obviously, at one photo per second, it will take much, much longer than the age of the universe for one person to view them all. That's assuming other problems are solved, like extending a person's lifespan, and supplying huge quantities of grilled steaks, dark beers and great coffee (because, that's why).
How about if we use 5 billion people, each viewing photos at a rate of one per second, each for 80 years? It would still require 4.29 x 10^4,623,801 earth-fulls of people for 80 years to view them all.
I'll leave it to the morbidly curious to calculate the total possible number of 15-minute YouTube videos. (Hint: divide by 30x60x15; 25x60x15 for Europe). For the total number of Vine videos, divide by 30x6.
Such thought experiments are useful, actually. For one, there are a large number of seemingly near-identical images for every seemingly unique one. Which raises the point of how do we agree on uniqueness. How many pixels have to differ before two photos are no longer practically unique?
You can also see that there will be a huge number of images that will appear to be random grain, similar to what we used to call "snow" in analog TVs tuned to a dead channel.
There will also be a huge number of images that will be non-random but abstract, along with another huge set that will be in the gray area between random and abstract. You can begin to see that random and abstract are not necessarily the same thing.
Any image that can be represented as three 8-bit RGB channels, at 800 pixels on a side, will be included. Including every 800-pixel rendering of every painting ever made. And every '80s-era 8-bit video game graphic ever made. And every one never made.
Every photograph yet to be made will be included, even ones that never could be made without Photoshop, like Jesus riding a unicorn atop a pizza slice inside a snow globe, for example. Or substitute your favorite diety. Or every possible version of Ansel Adams's "Moon Over Halfdome," including the alternate ones with purple dolphins jumping off the waterfall, or "Moonrise over Hernandez, New Mexico" with massive mothership UFOs in the sky.
Every photograph ever made will be included, along with the alternate 640,000 versions of each with a dead pixel (not counting color variations amongst the dead pixels - you can figure that out for yourself).
So, yes, photography is theoretically closed. But still pretty darned open to future creative potential.
And so now I'll end it here, because I have more photos to make, and view. Just don't take any of mine. Now, where's that steak and beer?
~Joe
Addenda: Corrected video numbers. But heck, this is all theoretical anyway, so what's a few thousand between friends?
How many possible photographs are there?
As a side question, are you old enough to remember the early days of electronic calculators? I do. I had an interesting collection at one time, including ones with red LED displays, others with green fluorescent tube displays that ate batteries like crazy. I'm left with three HPs using RPN logic, and a red LED version once sold in hardware stores. Every phone, tablet and computer has a calculator, these days.
These are the kinds of thought experiments that will challenge even the best scientific calculator. But here goes.
Assume square format, 800 pixels on a side, 8-bit RGB channels. Pretty much the standard image viewed on a phone, tablet or computer.
Total possible photos = (Total Color Levels) ^ (Total Pixel Count)
That's total number of color levels raised to the power of total pixel count. If this isn't obvious, imagine a simple photo of just two pixels, each with three color levels. That's 3^2 or 9 total possible images.
Three 8-bit color channels implies 2^24 color levels, which is: 1.677x10^7.
800x800 pixel image size implies 640,000 pixels total.
Total possible photos = (1.677x10^7) ^ (640,000)
Converting to scientific notation (Google will be of little help here, as are most scientific calculators; you'll have to dust off your high school maths):
Total number of photographs = 5.4075 x 10^4,623,820
That's 5.4 followed by 4,623,823 zeroes.
That's a lot of photographs, each one unique. Many will differ by only one pixel. But it's a finite number of images.
How long would it take to view all those photos, assuming a rate of one photo per second (I know people who surf the Net faster than that)?
First, assume the universe is 14.8 billion years old. That's 4.667 x 10^17 seconds.
Obviously, at one photo per second, it will take much, much longer than the age of the universe for one person to view them all. That's assuming other problems are solved, like extending a person's lifespan, and supplying huge quantities of grilled steaks, dark beers and great coffee (because, that's why).
How about if we use 5 billion people, each viewing photos at a rate of one per second, each for 80 years? It would still require 4.29 x 10^4,623,801 earth-fulls of people for 80 years to view them all.
I'll leave it to the morbidly curious to calculate the total possible number of 15-minute YouTube videos. (Hint: divide by 30x60x15; 25x60x15 for Europe). For the total number of Vine videos, divide by 30x6.
Such thought experiments are useful, actually. For one, there are a large number of seemingly near-identical images for every seemingly unique one. Which raises the point of how do we agree on uniqueness. How many pixels have to differ before two photos are no longer practically unique?
You can also see that there will be a huge number of images that will appear to be random grain, similar to what we used to call "snow" in analog TVs tuned to a dead channel.
There will also be a huge number of images that will be non-random but abstract, along with another huge set that will be in the gray area between random and abstract. You can begin to see that random and abstract are not necessarily the same thing.
Any image that can be represented as three 8-bit RGB channels, at 800 pixels on a side, will be included. Including every 800-pixel rendering of every painting ever made. And every '80s-era 8-bit video game graphic ever made. And every one never made.
Every photograph yet to be made will be included, even ones that never could be made without Photoshop, like Jesus riding a unicorn atop a pizza slice inside a snow globe, for example. Or substitute your favorite diety. Or every possible version of Ansel Adams's "Moon Over Halfdome," including the alternate ones with purple dolphins jumping off the waterfall, or "Moonrise over Hernandez, New Mexico" with massive mothership UFOs in the sky.
Every photograph ever made will be included, along with the alternate 640,000 versions of each with a dead pixel (not counting color variations amongst the dead pixels - you can figure that out for yourself).
So, yes, photography is theoretically closed. But still pretty darned open to future creative potential.
And so now I'll end it here, because I have more photos to make, and view. Just don't take any of mine. Now, where's that steak and beer?
~Joe
Addenda: Corrected video numbers. But heck, this is all theoretical anyway, so what's a few thousand between friends?
aizan
Veteran
does it matter? what's at stake?
Ko.Fe.
Lenses 35/21 Gears 46/20
My photography calculated in inches of wet prints.
Keith
The best camera is one that still works!
I see it as I see music which in theory probably has it's own mathematical limitations but I'm happy in knowing that they will never be reached in my lifetime or anyone else's.
lukitas
second hand noob
Upsizing only a little grows your numbers explosively.
I save my jpegs to 1000 pixels on the long side, a lot of people go bigger than that, and, I keep my originals at full size : at 5 times linear from your 800 x 800 image.
That gets us a lot closer to infinity, especially if you include all the different sizes digital images can be.
You could go at the problem from the other angle : how many subjects are there, and how many different photos can you make from each subject.
There are about 7 billion faces on earth. And every one of these faces can be portrayed in thousands of different ways : light, age, expression, apparel.
To make all possible photos, you'd have to fill the universe with cameras shooting 60 frames a second all the time, to account for all the angles and all the movements and all the changes.
The amount of pictures we can take might as well be infinite, but the number of photos we will never take is even bigger.
I'll share that steak and beer with you.
cheers
I save my jpegs to 1000 pixels on the long side, a lot of people go bigger than that, and, I keep my originals at full size : at 5 times linear from your 800 x 800 image.
That gets us a lot closer to infinity, especially if you include all the different sizes digital images can be.
You could go at the problem from the other angle : how many subjects are there, and how many different photos can you make from each subject.
There are about 7 billion faces on earth. And every one of these faces can be portrayed in thousands of different ways : light, age, expression, apparel.
To make all possible photos, you'd have to fill the universe with cameras shooting 60 frames a second all the time, to account for all the angles and all the movements and all the changes.
The amount of pictures we can take might as well be infinite, but the number of photos we will never take is even bigger.
I'll share that steak and beer with you.
cheers
Lss
Well-known
Dark beers and great coffee.
giellaleafapmu
Well-known
This remind me of a sentence I read sometimes in somebody's signature concerning the number of possible novel of a certain length. It went like this: "There used to be a theory which said that if enough monkeys type long enough they will eventually reproduce Shakespeare's work: the internet proved this to be false". Should we paraphrase the same and say that even tough there are only finitely many possible pictures of a certain size, RF proved that if we take photographs at random we will only get cat and benches' pictures? (Just kidding!)
GLF
GLF
Godfrey
somewhat colored
Wrong approach to the problem. Photographs are a frozen bit of light in a slice of time, so there are as many photographs as there are moments in time.
G
G
back alley
IMAGES
i always seek the simplest answers...how many left? at least one more!
JoeV
Thin Air, Bright Sun
Thanks for the responses, everyone. I had intended this to be both lighthearted and thought-provoking.
Lukitas, you're correct, upsizing the image size dramatically increases the set size.
Thinking further about this last night after I posted (I know - dangerous!), what amazes me about this thought experiment is, although this theoretical set of images includes all entirely uniquely different ones, how many images end up being near duplicates of others, so that while they might be technically unique, in practical terms they might not be. This has more to do with how the human brain can so easily find similar patterns between technically unequal images.
This mirrors my own experience, that while a person can randomly blast away taking seemingly random photos, far fewer of them are worth considering a second glance, while a tiny subset of those are truly great images.
I can't imagine a way to calculate such things as the subset of all possible images that are considered "great", because how does one assign numeric values to aesthetic considerations? One simply can't. But my gut feel is that subset, though still huge, is a small fraction of the total number of possible images, mirroring my personal experience as to how small a fraction of great images I've discovered out of the sum total I've shot over the years.
~Joe
Lukitas, you're correct, upsizing the image size dramatically increases the set size.
Thinking further about this last night after I posted (I know - dangerous!), what amazes me about this thought experiment is, although this theoretical set of images includes all entirely uniquely different ones, how many images end up being near duplicates of others, so that while they might be technically unique, in practical terms they might not be. This has more to do with how the human brain can so easily find similar patterns between technically unequal images.
This mirrors my own experience, that while a person can randomly blast away taking seemingly random photos, far fewer of them are worth considering a second glance, while a tiny subset of those are truly great images.
I can't imagine a way to calculate such things as the subset of all possible images that are considered "great", because how does one assign numeric values to aesthetic considerations? One simply can't. But my gut feel is that subset, though still huge, is a small fraction of the total number of possible images, mirroring my personal experience as to how small a fraction of great images I've discovered out of the sum total I've shot over the years.
~Joe