Pro Tools 11 vs Pro Tools 11 HD w/ HDX or Native Headroom difference

[JFreak]

Originally Posted by Amack

OK

Right, so there you have it. I used 8 and 16 bit wordlengths because the principle doesn't change and it would be exhausting to write 32 and 64 bit wordlengths as an example. But now as you get this, you can also understand the reason why mixer word length is beneficial to be longer than what it mixes. For example;



You have a 16-bit mixer that knows numbers between zero (all zeros in the binary word) and 65535 (all ones in the binary word), that means if you have 256 tracks of full-scale values values (all ones in the binary word) it will lead to mixer having value of 65535 (16 one-bits, a.k.a full scale). THEN, if you added another track, you would clip the mixer. In other words, if you needed to mix another full-scale 8-bit word together, you would need a mixer that has word length longer than 16 bits. And if you don't, all that longer wordlength does is eat resources and have a lot of zeroes in it.



That's all of it in a nutshell. Floating point math makes this a little more interesting, but this is all you as a recording engineer should ever understand.

[Amack]

Really? Is that Avid's way of thinking?

Originally Posted by JFreak

Right, so there you have it. I used 8 and 16 bit wordlengths because the principle doesn't change and it would be exhausting to write 32 and 64 bit wordlengths as an example. But now as you get this, you can also understand the reason why mixer word length is beneficial to be longer than what it mixes. For example;



You have a 16-bit mixer that knows numbers between zero (all zeros in the binary word) and 65535 (all ones in the binary word), that means if you have 256 tracks of full-scale values values (all ones in the binary word) it will lead to mixer having value of 65535 (16 one-bits, a.k.a full scale). THEN, if you added another track, you would clip the mixer. In other words, if you needed to mix another full-scale 8-bit word together, you would need a mixer that has word length longer than 16 bits. And if you don't, all that longer wordlength does is eat resources and have a lot of zeroes in it.



That's all of it in a nutshell. Floating point math makes this a little more interesting, but this is all you as a recording engineer should ever understand.

[musicman691]

Originally Posted by Amack

Really? Is that Avid's way of thinking?

Really? Why are you even worried about all this? I know you put up a link to Katz's book on mastering but that has nothing to do with this at all. Going through all that you're trying to go through won't help you make better recordings, if it does anything at all. Or are you trying to dig up material as research for a doctorial paper?

There's academic research and there's recording music; what you're looking at is the former when all you really should be concerned with is the basics of getting a recording system to work for you. And what you're asking about has not one thing to do with that.

[JFreak]

Originally Posted by Amack

Really? Is that Avid's way of thinking?

Yes they think like that because that is basic (integer) mathematics. Avid cannot beat the laws of physics, which includes mathematics.

1+1 equals 2 (base-10)
01+01 equals 10 (base-2)

Computers talk binary numbers, which is why most people have hard time understanding "bits". They are, however, just numbers which is what I was trying to explain in layman's terms. You say 10 (base-10), computer says 1010 (base-2) and both of you are talking about the same number. And that number in 24bit wav file would be 00000000000000000000000000001010

DAW mixer does nothing else than that. It combines the sample values of all the tracks that need to be summed. Minimum value is 0 (absolute silence) and maximum value is 2^wordlenght. In my example, the mixer wordlenght is 16 so 2^16 equals 65536 (distinct values) so the mixer deals with sample values 0-65535 represented in binary numbers, a.k.a. "bits".

Floating point math is even more interesting, but if you fail to understand integer math, let's not go there.

[JFreak]

My point in joining this discussion was to talk about the concept of "headroom" in digital systems. Nothing more. It's just math. Most people won't understand -- AND MOST RECORDING ENGINEERS WILL NOT HAVE TO !!!

[musicman691]

Originally Posted by JFreak

My point in joining this discussion was to talk about the concept of "headroom" in digital systems. Nothing more. It's just math. Most people won't understand -- AND MOST RECORDING ENGINEERS WILL NOT HAVE TO !!!

Amen. I understand it because I'm a huge math geek but is it necessary to getting great recordings? Nope. Perhaps if I were wearing my EE hat and was designing new electronics or trying to improve on a design I'd get wrapped up in it but otherwise - nope again.

Perhaps the best advice to come out of all this discussion is to stay in as high a bit rate and bit depth as your system will allow and only when it comes to delivering the end product come down to something like 16 bit 44.1 KHz for a CD. Leave your mastering engineer a couple of dB headroom and you're golden.

[Amack]

Have you read the book?

Originally Posted by musicman691

Really? Why are you even worried about all this? I know you put up a link to Katz's book on mastering but that has nothing to do with this at all. Going through all that you're trying to go through won't help you make better recordings, if it does anything at all. Or are you trying to dig up material as research for a doctorial paper?

There's academic research and there's recording music; what you're looking at is the former when all you really should be concerned with is the basics of getting a recording system to work for you. And what you're asking about has not one thing to do with that.

[JFreak]

Originally Posted by Amack

Have you read the book?

Yes. It is more practical approach than yours. No offense :)

[musicman691]

Originally Posted by Amack

Have you read the book?

Enough to know you're making mountains out of mole hills.