Compression: Crash Course Computer Science #21 CrashCourse English English (auto-generated)

Compression: Crash Course Computer Science #21YouTube↗ IA↗(long shot)
Education 340.5K views 9.2K likes 128 dislikes 26 Jul 2017 Show Contributors 2

CrashCourse 17.1M subscribers United States YouTube↗

Contributed subtitles

Korean byongcheol Ko

Chinese (China) Zhihao Jiang

Click to Play Video

This episode is brought to you by Curiosity Stream.

this episode is brought to you by

Hi, I'm Carrie Anne, and welcome to Crash Course Computer Science!

curiosity stream hi I'm Kerry Ann and welcome to crash course computer science

Last episode we talked about Files, bundles of data, stored on a computer, that are formatted

last episode we talked about files bundles of data stored on a computer

and arranged to encode information, like text, sound or images.

that are formatted and arranged to encode information like text sound or

We even discussed some basic file formats, like text, wave, and bitmap.

images we even discuss some basic file formats like text WAV and bitmap while

While these formats are perfectly fine and still used today, their simplicity also means

these formats are perfectly fine and still use today their simplicity also

they’re not very efficient.

means they're not very efficient ideally

Ideally, we want files to be as small as possible, so we can store lots of them without filling

we want files to be as small as possible so we can store lots of them without

up our hard drives, and also transmit them more quickly.

filling up our hard drives and also transmit them more quickly nothing is

Nothing is more frustrating than waiting for an email attachment to download. Ugh!

more frustrating than waiting for an email attachment to download the answer

The answer is compression, which literally squeezes data into a smaller size.

is compression which literally squeezes

To do this, we have to encode data using fewer bits than the original representation.

data into a smaller size to do this we have to encode data using fewer bits than the original representation that

That might sound like magic, but it’s actually computer science!

might sound like magic but it's actually

computer science [Music]

Lets return to our old friend from last episode, Mr. Pac-man!

let's return to our old friends from last episode mr. pac-man this image is 4

This image is 4 pixels by 4 pixels.

pixels by 4 pixels as we discussed image

As we discussed, image data is typically stored as a list of pixel values.

data is typically stored as a list of

To know where rows end, image files have metadata, which defines properties like dimensions.

pixel values to know where rows end image files have metadata which defines

But, to keep it simple today, we’re not going to worry about it.

properties like I mentions but to keep it simple today we're not going to worry

Each pixel’s color is a combination of three additive primary colors: red, green and blue.

about it each pixels color is a combination of three additive primary colors red green and blue we store each

We store each of those values in one byte, giving us a range of 0 to 255 for each color.

of these values in one byte giving us a range of 0 to 255 for each color if you

If you mix full intensity red, green and blue - that’s 255 for all three values - you

mix full intensity red green and blue that's 255 for all three values you get

get the color white.

the color white if you mix full

If you mix full intensity red and green, but no blue (it’s 0), you get yellow.

intensity red and green but no blue zero

We have 16 pixels in our image, and each of those needs 3 bytes of color data.

you get yellow we have 16 pixels in our image and each of those needs 3 bytes of

That means this image’s data will consume 48 bytes of storage.

color data that means this images data will consume 48 bytes of storage but we

But, we can compress the data and pack it into a smaller number of bytes than 48!

can compress the data and pack it into a smaller number of bytes from 48 one way

One way to compress data is to reduce repeated or redundant information.

to compress data is to reduce repeated or redundant information the most

The most straightforward way to do this is called Run-Length Encoding.

straightforward way to do this is called and run length encoding this takes

This takes advantage of the fact that there are often runs of identical values in files.

advantage of the fact there are often runs of identical values and files for

For example, in our pac-man image, there are 7 yellow pixels in a row.

example in our pac-man image there are 7 yellow pixels in a row instead of

Instead of encoding redundant data: yellow pixel, yellow pixel, yellow pixel, and so

encoding redundant data yellow pixel

on, we can just say “there’s 7 yellow pixels in a row” by inserting an extra byte

yellow pixel yellow pixel and so on we can just say there are 7 yellow pixels in a row by inserting an extra

that specifies the length of the run, like so:

byte that specifies the length of the

And then we can eliminate the redundant data behind it.

run like so and then we can eliminate the redundant data behind it to ensure

To ensure that computers don’t get confused with which bytes are run lengths and which

that computers don't get confused with

bytes represent color, we have to be consistent in how we apply this scheme.

which bytes and run links from which bytes represent color we have to be consistent in how we apply this scheme

So, we need to preface all pixels with their run-length.

so we need to preface all pixels with

In some cases, this actually adds data, but on the whole, we’ve dramatically reduced

their run links in some cases this actually adds data but on the whole

the number of bytes we need to encode this image.

we've dramatically reduced the number of bytes we need to encode this image we're

We’re now at 24 bytes, down from 48.

now at 24 bytes down from 48 that's 50%

That’s 50% smaller! A huge saving!

smaller huge saving also now that we

Also note that we haven’t lost any data.

haven't lost any data we can easily

We can easily expand this back to the original form without any degradation.

expand this back to the original form without any degradation a compression

A compression technique that has this characteristic is called lossless compression, because we

technique that has this characteristic is called lossless compression because

don’t lose anything.

we don't lose anything

The decompressed data is identical to the original before compression, bit for bit.

the decompressed data is identical to the original before compression bits a

Let's take a look at another type of lossless compression, where blocks of data are replaced

bit let's take a look at another type of lossless compression where blocks of

by more compact representations.

data are replaced by more compact

This is sort of like “don’t forget to be awesome” being replaced by DFTBA.

representations this is sort of like don't forget to be awesome being replaced by dftba to do this we need a

To do this, we need a dictionary that stores the mapping from codes to data.

dictionary that stores the mapping from

Lets see how this works for our example.

Coates to data let's see how this works

We can view our image as not just a string of individual pixels, but as little blocks

for our example we can view our image is not just a string of individual pixels

but as a field blocks of data for

For simplicity, we’re going to use pixel pairs, which are 6 bytes long, but blocks

simplicity we're going to use pixel pairs which are 6 bytes long but blocks

can be any size.

can be any size in our example there are

In our example, there are only four pairings: White-yellow, black-yellow, yellow-yellow

only 4 pairings white yellow black

and white-white.

yellow yellow yellow and white white

Those are the data blocks in our dictionary we want to generate compact codes for.

those are the data blocks in our dictionary we want to generate compact

What’s interesting, is that these blocks occur at different frequencies.

codes for what's interesting is that these blocks occur at different

There are 4 yellow-yellow pairs, 2 white-yellow pairs, and 1 each of black-yellow and white-white.

frequencies there are 4 yellow yellow pairs 2 white yellow pairs and one each of black yellow and white whites because

Because yellow-yellow is the most common block, we want that to be substituted for the most

yellow yellow is the most common block we want that to be substituted for the

compact representation.

most compact representation on the other

On the other hand, black-yellow and white-white, can be substituted for something longer because

hand black yellow and white white can be substituted for something longer because

those blocks are infrequent.

those blocks are infrequent one method

One method for generating efficient codes is building a Huffman Tree, invented by David

for generating efficient codes is building a Huffman tree invented by

Huffman while he was a student at MIT in the 1950s.

David Huffman while he was a student at MIT in the 1950s his algorithm goes like

His algorithm goes like this.

First, you layout all the possible blocks and their frequencies.

first you lay out all the possible blocks and their frequencies at every

At every round, you select the two with the lowest frequencies.

round you select the two with the lowest

Here, that’s Black-Yellow and White-White, each with a frequency of 1.

frequencies here that black yellow and white white each for the frequency of 1

You combine these into a little tree... ...which have a combined frequency of 2, so we record

you combine these into a little tree which have a combined frequency of 2 so

we record that and now one step of the

And now one step of the algorithm done.

algorithm is done now we repeat the

Now we repeat the process.

process this time we have three things

This time we have three things to choose from.

Just like before, we select the two with the lowest frequency, put them into a little tree,

just like before we select the two with the lowest frequency put them into a

and record the new total frequency of all the sub items.

little tree and record the new total frequency of all the sub items okay

Ok, we’re almost done.

we're almost done

This time it’s easy to select the two items with the lowest frequency because there are

this time is easy to select the two items with the lowest frequency because

only two things left to pick.

there are only two things left to pick

We combine these into a tree, and now we’re done!

we combine these into a tree and now

Our tree looks like this, and it has a very cool property: it’s arranged by frequency,

we're done our tree looks like this and it has a very cool property it's arranged by frequency with less common

with less common items lower down.

items lower down so now we have a tree

So, now we have a tree, but you may be wondering how this gets us to a dictionary.

but you may be wondering how this gets

Well, we use our frequency-sorted tree to generate the codes we need by labeling each

us to a dictionary well we use our frequency sorted tree to generate the codes we need by labeling each branch

branch with a 0 or a 1, like so:

with a 0 or a 1 like so with this we can

With this, we can write out our code dictionary.

write out our code dictionary yellow

Yellow-yellow is encoded as just a single 0.

yellow is encoded as just a single 0

White-yellow is encoded as 1 0 (“one zero”)

white yellow is encoded as 1 0 black

Black-Yellow is 1 1 0

yellow is 1 1 0 and finally white white

and finally white-white is 1 1 1.

is 1 1 1 the really cool thing about

The really cool thing about these codewords is that there’s no way to have conflicting

these code words is that there's no way

codes, because each path down the tree is unique.

to have conflicting codes because each path down the tree is unique this means

This means our codes are prefix-free, that is no code starts with another complete code.

our codes are prefix free that is no code starts with another complete code

Now, let’s return to our image data and compress it!

now let's return to our image data and

Our first pixel pair, white-yellow, is substituted for the bits “1 0”.

compress it our first pixel pair white yellow is substituted for the bits 1 0

The next pair is black-yellow, which is substituted for “1 1 0”.

the next Perris black yellow which is substituted

Next is yellow-yellow with the incredibly compact substitution of just “0”.

for one one zero next is yellow yellow with the incredibly compact substitution

And this process repeats for the rest of the image:

of just zero and this process repeats

So instead of 48 bytes of image data ...this process has encoded it into 14 bits -- NOT

for the rest of the image so instead of 48 bytes of image data this process has encoded it into 14 bits

BYTES -- BITS!!

not by its bits that's less than 2 bytes

That’s less than 2 bytes of data!

of data but don't break out the

But, don’t break out the champagne quite yet!

champagne quite yet this data is

This data is meaningless unless we also save our code dictionary.

meaningless unless we also save our code

So, we’ll need to append it to the front of the image data, like this.

dictionary so we'll need to append it to the front of the image data like this

Now, including the dictionary, our image data is 30 bytes long.

now including the dictionary our image data is 30 bytes long that's still a

That’s still a significant improvement over 48 bytes.

significant improvement over 48 bytes

The two approaches we discussed, removing redundancies and using more compact representations,

the two approaches we discussed removing redundancies and using more compact

are often combined, and underlie almost all lossless compressed file formats, like GIF,

representations are often combined and underlie all most all lossless compressed file formats like gif png PDF

PNG, PDF and ZIP files.

and zip files both run length encoding

Both run-length encoding and dictionary coders are lossless compression techniques.

and dictionary coders are lossless

No information is lost; when you decompress, you get the original file.

compression techniques no information is lost when you decompress you get the

That’s really important for many types of files.

original file that's really important

Like, it’d be very odd if I zipped up a word document to send to you, and when you

for many types of files like it'd be very odd if I zipped up a word document

decompressed it on your computer, the text was different.

to send to you and when you decompressed it on your computer the text was

But, there are other types of files where we can get away with little changes, perhaps

different but there are other types of files where we can get away with little

by removing unnecessary or less important information, especially information that human

changes perhaps by removing unnecessary or less important information especially

perception is not good at detecting.

information that human perception is not

And this trick underlies most lossy compression techniques.

good at detecting and this trick underlies most lossy compression

These tend to be pretty complicated, so we’re going to attack this at a conceptual level.

techniques these tend to be pretty complicated so we're going to attack this at a conceptual level

Let’s take sound as an example.

let's take sound as an example your

Your hearing is not perfect.

hearing is not perfect we can hear some

We can hear some frequencies of sound better than others.

frequencies and sound better than others

And there are some we can’t hear at all, like ultrasound.

and there are some we can't hear at all

Unless you’re a bat.

like ultrasound unless you're a bat

Basically, if we make a recording of music, and there’s data in the ultrasonic frequency

basically if we make a recording of music and there's data in the ultrasonic

range, we can discard it, because we know that humans can’t hear it.

frequency range we can discard it because we know that humans can't hear

On the other hand, humans are very sensitive to frequencies in the vocal range, like people

it on the other hand humans are very sensitive to frequencies in the vocal

singing, so it’s best to preserve quality there as much as possible.

range like people singing it's best to preserve quality there as much as

Deep bass is somewhere in between.

possible the bass is somewhere in

Humans can hear it, but we’re less attuned to it.

between humans can hear it but we're

We mostly sense it.

less attuned to it we mostly sense it

Lossy audio compressors takes advantage of this, and encode different frequency bands

lossy audio compressors take advantage of this and encode different frequency

at different precisions.

bands at different Precision's even if

Even if the result is rougher, it’s likely that users won’t perceive the difference.

the result is rougher it's likely that users won't perceive the difference or

Or at least it doesn’t dramatically affect the experience.

at least it doesn't dramatically affect

And here comes the hate mail from the audiophiles!

the experience and here comes the hate mail from the audio files you encounter

You encounter this type of audio compression all the time.

this type of audio compression all the

It’s one of the reasons you sound different on a cellphone versus in person.

time it's one of the reasons you sound different on a cell phone versus

The audio data is being compressed, allowing more people to take calls at once.

in-person the audio data is being compressed allowing more people to take

As the signal quality or bandwidth get worse, compression algorithms remove more data, further

calls at once as signal quality or bandwidth gets worse compression algorithms remove more data

reducing precision, which is why Skype calls sometimes sound like robots talking.

further reducing precision which is why Skype calls sometimes sound like robots

Compared to an uncompressed audio format, like a WAV or FLAC (there we go, got the audiophiles back)

talking compared to an uncompressed audio format like a wave or flag there we go got the audio files back

compressed audio files, like MP3s, are often 10 times smaller.

compressed audio files like mp3s are

That’s a huge saving!

often ten times smaller that's a huge

And it’s why I’ve got a killer music collection on my retro iPod.

saving and it's why I've got a killer music collection on my retro iPod don't

judge this idea of discarding or

This idea of discarding or reducing precision in a manner that aligns with human perception

reducing precision in the manner that aligns with human perception is called

is called perceptual coding, and it relies on models of human perception,

perceptual coding and it relies on models of human perception which comes

which come from a field of study called Psychophysics.

from a field of study good psychophysics

This same idea is the basis of lossy compressed image formats, most famously JPEGs.

this same idea is the basis of lossy compressed image formats most famously

Like hearing, the human visual system is imperfect.

JPEGs like hearing the human visual

We’re really good at detecting sharp contrasts, like the edges of objects, but our perceptual

system is imperfect we're really good at detecting sharp contrasts like the edges of objects our perceptual system isn't

system isn’t so hot with subtle color variations.

so hot with subtle color variations jpg

JPEG takes advantage of this by breaking images up into blocks of 8x8 pixels, then throwing

takes advantage of this by breaking images up into blocks of 8 by 8 pixels

away a lot of the high-frequency spatial data.

then throwing away a lot of the high-frequency spatial data for example

For example, take this photo of our directors dog - Noodle.

take this photo of our directors dog

noodle so cute let's look at a patch of

Let’s look at patch of 8x8 pixels.

8 by 8 pixels pretty much every picture

Pretty much every pixel is different from its neighbor, making it hard to compress with

is different from its neighbor making it

loss-less techniques because there’s just a lot going on.

hard to compress with lossless techniques because there's just a lot

Lots of little details.

going on lots of little details but

But human perception doesn’t register all those details.

human perception doesn't register all

So, we can discard a lot of that detail, and replace it with a simplified patch like this.

those details so we can discard a lot of it and replace it with a simplified

This maintains the visual essence, but might only use 10% of the data.

patch like this this maintains the visual essence but might only use 10% of

We can do this for all the patches in the image and get this result.

the data we can do this for all the patches in the image and get this result

You can still see it’s a dog, but the image is rougher.

you can still see it's a dog but the

So, that’s an extreme example, going from a slightly compressed JPEG to a highly compressed

image is rougher so that's an extreme example going from slightly compressed JPEG to a highly compressed one 1/8 the

one, one-eighth the original file size.

original file size often you can get

Often, you can get away with a quality somewhere in between, and perceptually, it’s basically

away of a quality somewhere in between and perceptually it's basically the same

the same as the original.

as the original the one on the left is

The one on the left is one-third the file size of the one on the right.

1/3 the file size of the one on the

That’s a big savings for essentially the same thing.

right that's big savings for essentially

Can you tell the difference between the two?

the same thing can you tell the

Probably not, but I should mention that video compression plays a role in that too, since

difference between the two probably not but I should mention that video compression plays our role in that too

I’m literally being compressed in a video right now.

since I'm literally being compressed

Videos are really just long sequences of images, so a lot of what I said about them applies

right now videos are really just long sequences of images so a lot of what I

said about them applies here too but

But videos can do some extra clever stuff, because between frames, a lot of pixels are

videos can do some extra clever stuff because between frames a lot of pixels

going to be the same.

are going to be the same like this whole

Like this whole background behind me!

background behind me this is called

This is called temporal redundancy.

temporal redundancy we don't need to

We don’t need to re-transmit those pixels every frame of the video.

retransmit those pixels every frame of

We can just copy patches of data forward.

the video which is copy patches of data

When there are small pixel differences, like the readout on this frequency generator behind

forward when there are small pixel differences like the readout on

me, most video formats send data that encodes just the difference between patches, which

frequency generates behind me most video formats send data that encodes just the difference between patches which is more

is more efficient than re-transmitting all the pixels afresh, again taking advantage

efficient than retransmitting all the pixels of fresh again taking advantage

of inter-frame similarity.

of inter frame similarity the fanciest

The fanciest video compression formats go one step further.

video compression formats go one step

They find patches that are similar between frames, and not only copy them forward, with

further they find patches that similar between frames and not only copy them

or without differences, but also can apply simple effects to them, like a shift or rotation.

forward with or without differences but also can apply simple effects to them like a shift or rotation they can also

They can also lighten or darken a patch between frames.

lighten or darken a patch between frames

So, if I move my hand side to side like this the video compressor will identify the similarity,

so if I move my hand side to side like this the video compressor will identify the similarity capture my hand in one or

capture my hand in one or more patches, then just move these patches around between frames.

more patches then just move these patches around between frames you're

You’re actually seeing my hand from the past… kinda freaky, but it uses a lot less data.

actually seeing my hand from the past kind of freaky but it uses a lot less

MPEG-4 videos, a common standard, are often 20 to 200 times smaller than the original,

data mpeg-4 videos a common standard often 20 to 200 times smaller than the

uncompressed file.

original uncompressed file however

However, encoding frames as translations and rotations of patches from previous frames

encoding frames as translations and rotations of patches from previous

can go horribly wrong when you compress too heavily, and there isn’t enough space to

frames can go horribly wrong when you compress too heavily and there isn't

update pixel data inside of the patches.

enough space to update pixel data inside

The video player will forge ahead, applying the right motions, even if the patch data

of the patches the video player will forge ahead applying the right motions even if the patch data is wrong and this

leads to some hilarious and tribute

And this leads to some hilarious and trippy effects, which I’m sure you’ve seen.

effects which I'm sure you've seen

Overall, it’s extremely useful to have compression techniques for all the types of data I discussed today.

overall is extremely useful to have compression techniques for all the types of data I discussed today I guess are a

(I guess our imperfect vision and hearing are “useful,” too.)

perfect vision and hearing a useful -

And it’s important to know about compression because it allows users to store pictures,

and it's important to know about compressions because it allows users to

music, and videos in efficient ways.

store pictures music and videos in

Without it, streaming your favorite Carpool Karaoke videos on YouTube would be nearly

efficient ways without compression streaming your favorite carpool karaoke videos on YouTube will be nearly

impossible, due to bandwidth and the economics of transmitting that volume of data for free.

impossible due to bandwidth and the economics of transmitting that volume of

And now when your Skype calls sound like they’re being taken over by demons, you’ll know

data for free and now when your Skype called sound likely being taken over by

what’s really going on.

demons you'll know what's really going

I’ll see you next week.

on I'll see you next week hey guys this

Hey guys, this week’s episode was brought to you by CuriosityStream which is a streaming

week's episode was brought to you by curiosity stream which is a streaming

service full of documentaries and nonfiction titles from some really great filmmakers,

service full of documentaries and nonfiction titles from some really great

including exclusive originals.

filmmakers including exclusive originals

Now I normally give computer science recommendations since this is Crash Course Computer Science and all

now I normally give computer science recommendations since this is crash course computer science at all and

and Curiosity Stream has a ton of great ones. But you absolutely have to check

curiosity stream has a ton of great ones but you absolutely have to check out

out “Miniverse” starring everyone’s favorite space-station-singing-Canadian astronaut,

miniverse starring everyone's favorite space stations being in Canadian

Chris Hadfield, as he takes a roadtrip across the Solar System scaled down the the size

astronaut Chris Hadfield as he takes a road trip across the solar system scaled

of the United States.

down to the size of the United States

It’s basically 50 minutes of Chris and his passengers geeking out about our amazing planetary

it's basically 50 minutes of Chris and his passengers geeking out about our

neighbors and you don’t want to miss it.

amazing planetary neighbors and don't

So get unlimited access today, and your first two months are free if you sign up at curiositystream.com/crashcourse

want to miss it so get unlimited access today and your fur two months of three if you sign up at

curiosity Steve calm slash crash course

and use the promo code "crashcourse" during the sign up process.

and use the promo code crash course during the signup process crash course computer science is produced in association with PBS Digital Studios at

their channel you can check out a playlist of shows like PBS idea channel physics girl and it's okay to be smart this episode was filmed at the chad and

stacey emigholz studio in indianapolis indiana and it was made with the help of all these nice people and our wonderful graphic scene thought cafe thanks for

watching I'll see you later [Music]

Similar videos:

The Internet: Crash Course Computer Science #29
CrashCourse
458K views 9.4K likes 163 dislikes 20 Sep 201767% match

Cybersecurity: Crash Course Computer Science #31
CrashCourse
696.6K views 12.7K likes 280 dislikes 11 Oct 201767% match

3D Graphics: Crash Course Computer Science #27
CrashCourse
326.5K views 10.2K likes 107 dislikes 06 Sep 201767% match

Crash Course Computer Science Preview
CrashCourse
2.3M views 38.9K likes 619 dislikes 15 Feb 201767% match

Computer Networks: Crash Course Computer Science #28
CrashCourse
778.8K views 14.8K likes 281 dislikes 13 Sep 201757% match

Early Programming: Crash Course Computer Science #10
CrashCourse
524.1K views 10K likes 126 dislikes 03 May 201757% match

Early Computing: Crash Course Computer Science #1
CrashCourse
2.6M views 51.5K likes 1K dislikes 22 Feb 201757% match

Intro to Algorithms: Crash Course Computer Science #13
CrashCourse
1.3M views 23.7K likes 550 dislikes 24 May 201757% match

Memory & Storage: Crash Course Computer Science #19
CrashCourse
588.4K views 11.1K likes 244 dislikes 05 Jul 201757% match

Instructions & Programs: Crash Course Computer Science #8
CrashCourse
640.4K views 12.2K likes 140 dislikes 12 Apr 201757% match

Intro to History of Science: Crash Course History of Science #1
CrashCourse
802.2K views 19.5K likes 390 dislikes 26 Mar 201827% match

🍌 What is Geography? Crash Course Geography #1
CrashCourse
338.7K views 10.3K likes 242 dislikes 30 Nov 202025% match

India: Crash Course History of Science #4
CrashCourse
1.4M views 35.9K likes 1.2K dislikes 25 Apr 201825% match

Crash Course Geography Preview
CrashCourse
226.3K views 9K likes 75 dislikes 17 Nov 202025% match

What is a Map? Crash Course Geography #2
CrashCourse
248.4K views 6.2K likes 182 dislikes 14 Dec 202025% match

Deep Time: Crash Course Astronomy #45
CrashCourse
2M views 35.3K likes 751 dislikes 14 Jan 201625% match

Plato and Aristotle: Crash Course History of Science #3
CrashCourse
1.1M views 20.1K likes 498 dislikes 16 Apr 201825% match

Karl Popper, Science, & Pseudoscience: Crash Course Philosophy #8
CrashCourse
1.9M views 38.3K likes 501 dislikes 28 Mar 201625% match

What is Engineering?: Crash Course Engineering #1
CrashCourse
984.8K views 23.2K likes 589 dislikes 17 May 201824% match

Crash Course Linguistics Preview
CrashCourse
264.8K views 13.5K likes 95 dislikes 04 Sep 202024% match

Crash Course Study Skills Preview
CrashCourse
848.4K views 17.1K likes 221 dislikes 01 Aug 201724% match

Compatibilism: Crash Course Philosophy #25
CrashCourse
1.3M views 28K likes 306 dislikes 22 Aug 201624% match

What is space and how do we study it? Crash Course Geography #3
CrashCourse
151.6K views 4.3K likes 77 dislikes 22 Dec 202024% match

Welcome to Crash Course
CrashCourse
2.5M views 9.2K likes -1 dislikes 12 Nov 201924% match

What is Physical Geography? Crash Course Geography #4
CrashCourse
147.4K views 3.8K likes 76 dislikes 12 Jan 202124% match

Crash Course Sociology Preview
CrashCourse
1M views 15K likes 491 dislikes 27 Feb 201724% match

What is Philosophy?: Crash Course Philosophy #1
CrashCourse
5.1M views 94.3K likes 1.4K dislikes 08 Feb 201623% match

How Computers Calculate - the ALU: Crash Course Computer Science #5
CrashCourse
1.3M views 24.2K likes 393 dislikes 22 Mar 201723% match

Utilitarianism: Crash Course Philosophy #36
CrashCourse
3.2M views 56.8K likes 865 dislikes 21 Nov 201623% match

Metaethics: Crash Course Philosophy #32
CrashCourse
1.4M views 24.2K likes 346 dislikes 25 Oct 201623% match

Settings Privacy About Contact Donate