I'm trying to read about midi files here:
http://www.sonicspot.com/guide/midifiles.html

It says in Big endian C8 is 11001000

So I guess 1000 is 8 and 1100 is C.

If I fill out all 4 digit binary numbers that makes sense
0000 0
0001 1
0010 2
0011 3
0100 4
0101 5
0110 6
0111 7
1000 8
1001 9
1010 A
1011 B
1100 C
1101 D
1110 E
1111 F

So what is big endian? I don't get it. I keep reading it has to do with the most significant bit. I was using little endian then how would my 16 numbers above read?

C8 hex is always 11001000 binary.

It is written in text with the most significant part on the left.

Big endian means most significant part first, while little endian means least significant part first.

If you can't understand the big difference between "on the left" and "first", you're going to be a little confused here.

In English we read left to right, so we think left is first. Inside a computer left and right don't have that kind of meaning.

Next trick is figuring out what "part" means. Big endian means the most significant part first. But it is ambiguous about what "part" means.

In that midi file document, "part" means "byte". They are not telling you the sequence of bits within a byte. They are telling you the sequence of bytes within any multi byte quantity.

By "read" you are pulling in a "left to right" component that only leads to confusion in this topic.

You also are looking at just half a byte at a time. Serial (bit at a time) transfer and storage of data in and between computers is normally little endian regarding the sequence of bits within each byte. But that level (such as transferring bits on the SATA cable between the disk and the ram) is almost entirely transparent to the level of programming MIDI files.

So if you insist on left == first then to describe bit level little endian, you would need to write the four digits of each of your binary numbers backwards. I prefer the view that says you write them exactly as you have but read them starting at the right.

But all of the bit level sequence is a nearly pointless side question, because that MIDI spec is just talking about byte sequence. The bit sequence within a byte is whatever is natural to the computer you are on.

I don't get it.

Well in the link it says the biggest value that can be represented by 4 bytes is 0x0FFFFFFF so it looks like we lose half a byte. I'd think the biggest value would be 0xFFFFFFFF. So if the 1st bit tells whether there is another byte of information then I guess it can't be part of the actual number?

So I think 01111111 is the largest byte and then if you had 11111111 then that would make no sense by itself because it means there is another byte coming.

Or maybe I'm completely wrong I don't know.

It talks about variable length formats saying
So obviously in any variable length data you lose one bit per byte, so you lose four bits from four bytes. What is difficult about that?

Notice also they say "lower 7 bits" and "top bit". They don't say "first" or "last" bits. At least in that section the writer was careful over the fact that they are not saying anything about the sequence of bits in a byte.

Top bit not 1st bit. But otherwise, yes. The bit that tells you whether another byte follows is not part of the encoded number.