How to put ur beats on the computer

Mr modnaR wrote: utilise the headroom in the first place, saving the need to normalise, so i guess we don't need this function!

bada bing!




and if you are going to normalize, don't do it to 0 dB, go to like -.1 or -.2 because there are always some intersample peaks that will overshoot that amount, and a lot of consumer CD players and such will clip at 0, not over.


And if your going to be sending the track for mastering, DEFINITELY don't normalize.

[quote="rickrock305"] and if you are going to normalize, don't do it to 0 dB, go to like -.1 or -.2 because there are always some intersample peaks that will overshoot that amount, and a lot of consumer CD players and such will clip at 0, not over. [quote]

Please some audio engineer pull me up on this, but I believe the above statement is just plain wrong.

If the process of normalizing involves scanning the whole audio file to find the highest peak, and then raising the whole file proportionaly to ensure this highest peak does not go over 0db, then how can there be any overshoots when everything else is below the level of the highest peak ?

Answer: there are no overshoots because everything else is below the highest peak which has been optimized to 0db.

"utilise the headroom in the first place, saving the need to normalise, so i guess we don't need this function"

This is the best technique, but it is very hard to record to a perfect 0db digitally, hence recording slightly lower and then normalizing.

edit - bad spelling

.

jigginz wrote:Please some audio engineer pull me up on this, but I believe the above statement is just plain wrong.

If process of normalizing involves scanning the whole audio file to find the highest peak, and then raising the whole file proportionaly to ensure this highest peak does not go over 0db, then how can there be any overshoots when everything else is below the level of the highest peak ?

There are no overshoots because everything else is below the level of the highest peak which has been optimized to 0db.

yes i agree with you, i'm pretty sure that most, if not all, modern audio editors will check the whole waveform for peaks, intersample ones included.

jigginz wrote:This is absolutely the best technique, but it is very hard to record to a perfect 0db digitally, hence recording slightly lower and then normalizing.

plus, if you're recording digitally, you won't have to worry about any extra noise (on top of the original sample's noise, that is)

Keep in mind that even many of todays commercial, pro mastered CD´s have samples breaking the 0 dB limit. At least according to Bob Katz and other mastering engineers that are very anal and neat with the numbers. It´s not audible if it´s not a couple of hundred samples in a row that exceed the limit.

And Jigginz, I think that the audio engineer your asking for might be rickrock305. If I remember correctly he´s stated in another thread that he´s a professional engineer.

ryssen wrote:And Jigginz, I think that the audio engineer your asking for might be rickrock305. If I remember correctly he´s stated in another thread that he´s a professional engineer.

uh oh, better not joust with the fella then!

i did it the way u said , but the seqences stay in the tempo the first sequence in the song is in, and in comes off wrong!! how do i fix that??

dj oasis916 wrote:i did it the way u said , but the seqences stay in the tempo the first sequence in the song is in, and in comes off wrong!! how do i fix that??

HAHA oh man the timing of that was classic

ryssen wrote:Keep in mind that even many of todays commercial, pro mastered CD´s have samples breaking the 0 dB limit. At least according to Bob Katz and other mastering engineers that are very anal and neat with the numbers. It´s not audible if it´s not a couple of hundred samples in a row that exceed the limit.

And Jigginz, I think that the audio engineer your asking for might be rickrock305. If I remember correctly he´s stated in another thread that he´s a professional engineer.

All true. But mastering guys have equipment that can do that...you don't. Its absolutely audible, barely. Most people don't hear it, they feel it. But if its only a few samples in a row then it won't audibly clip. Its a very long a complicated discussion, but bottom line its not good. And Bob Katz will be the first to tell you. But the labels and artists always want their stuff louder than the next guy, even though it doesn't matter at all, in fact the louder you try to get it the worse off you'll be when it hits the radio. (This applies to mastering really)

jigginz wrote:

rickrock305 wrote: and if you are going to normalize, don't do it to 0 dB, go to like -.1 or -.2 because there are always some intersample peaks that will overshoot that amount, and a lot of consumer CD players and such will clip at 0, not over.

Please some audio engineer pull me up on this, but I believe the above statement is just plain wrong.

If the process of normalizing involves scanning the whole audio file to find the highest peak, and then raising the whole file proportionaly to ensure this highest peak does not go over 0db, then how can there be any overshoots when everything else is below the level of the highest peak ?

Answer: there are no overshoots because everything else is below the highest peak which has been optimized to 0db.

"utilise the headroom in the first place, saving the need to normalise, so i guess we don't need this function"

This is the best technique, but it is very hard to record to a perfect 0db digitally, hence recording slightly lower and then normalizing.

edit - bad spelling

.

your right and wrong. Yes, normalizing searches for peaks, but it is only searching the samples for the highest one. If you take the highest sample and normalize it to 0dB, there will be intersample peaks that exceed 0.

rickrock305 wrote:All true. But mastering guys have equipment that can do that...you don't.

Hey, I´m on your line in this discussion, I told that about commercial CD´s to point out that it´s near impossible to know if you exceed the 0 dB if you don´t actually get in there analysing the file with proper software.

Talking from my own perspective I´ve had things recorded (final 2-track mixes from analouge mixer) into Cubase with -1 dB or a straight +/-0 levels which later peak above 0 dB when I edited (that is cut n´paste, no processing) it in Pro Tools.
It´s not an exact science unless you make it one, which requires a tight studio enviroment and lot s of work and knowledge.

An important thing that not many people consider, is that the analog sound that comes out of your computer is not nesceserally the same level as the digital samplepoints in your computer. The analog sound wave might be louder that the digital representation of it, and if you have a sound on your computer that's ~0 dB, it will most likely clip when it's DA converted. A lot of the times it's not noticable, but sometimes it sounds terrible.

That's why you should have some headroom all the time (both recording and mixing). Pro mastering engineers say they prefer that the mix has plenty of headroom, around 6 dB (meaning the signal should be @ -6 dB).

Wow, I need to do some reading up this, there's a bit more to this than I thought.

"Yes, normalizing searches for peaks, but it is only searching the samples for the highest one. "

Please enlighten me! So you're saying there is something more than just the samples? I suppose these are the "intersample peaks" you talk of? What the hell are these, and how can they clip if they are not samples?

I'm lost

jigginz wrote:Wow, I need to do some reading up this, there's a bit more to this than I thought.

http://www.digido.com/

Check the articles.

ryssen wrote:

jigginz wrote:Wow, I need to do some reading up this, there's a bit more to this than I thought.

http://www.digido.com/

Check the articles.

second that. Check this site out for all the answers to all your questions about digital levels and clipping. Bob Katz explains it much better than i can.

Bottom line, no need to slam levels, anywhere at anytime. Leave that to the mastering guys.

sub