Normalize Levels Over 100%

Has anyone else played around with this? I just took a kick sample which was ALREADY very loud, and normalized a few times with the level at 120%. This sample is now gritty and loud to a whole new level! I love it. Can anyone explain what's happening here? It doesn't really sound distorted, almost like compressed...

Did JJ slip in a secret fairy dust feature?

nawh man. it just simple normalize level I asked him to add. You see I come from Boss sp-505 sampler and it hd ability to normalize not only to the 100% but over it thus it allowed you to have samples louder.

if you try to normalize it many times with values over 100% you will start to hear distortion but if you do this once or twice it will simply put the main frequencies louder and those that are even higher are simply cut off.

so if you sampled a record with a loud crack in the middle if you tried to normalize to 100% sound will still be quiet but the crack gots louder. what you can do now is to normalize to even higher level to have the sample sound as loud as the crack is. nahmsayin' ?

i really missed this feature from 505 so i asked JJ to add this to OS and he thankfully did it!

Peace out,

Olby

Olby, yeah man! That's pretty much what I'm hearing, thanks for confirming. And thanks for suggesting it to JJ!

You're telling us that at normalize levels over 100%, it begins to look for increasingly quiet peaks of the sample to bring up to 0DB. So then, are the louder peaks going over 0DB? It doesn't really sound like it, but would seem likely. The other alternative would be that the loudest transients remain limited at 0DB, which would definitely explain the compression effect I'm hearing. Either way I love it!

BTW, the I used to rock an SP-505! The sound and effects were sweet, but that 4 track sequencer was a real let down.

yeah you understood me correctly. when you normalize all peaks become louder but as the crack and pops are not that loud and long to make distortion you will not hear any difference in the sound quality until the main sample frequencies start to distort.

olby,

whats happening in usual normalize - is that any sample levels below 100% will be raised up by the amount that the highest sample is from maximum. then each subsequent JJ normalize must be sending the values higher (esp when a increase of 100%+ is used) than 100%, and then is truncating values at 100%. so the resulting waveshapes are being pushed up from their original context.....giving a distorted sample (distorted doesnt necessarily mean ruined - as often the samples inherit a hyped up kind of feel), a bit like extreme limiting/compression techniques.

yeah arebee explained it better you see, but there is definitely no compression involved at all.

its called saturation or clipping or more accurate for digital domains: wrapping

yeay, wrapping, nice def.