How To Tell If A Song Is Clipped

[BigPimpin91]

I was unclipping the "The Dream - let me see the booty feat lil jon" song

It took 1 hour to remove all of the clipped parts.

There wasn't big chunks so it was easy to just zoom in and remove the nano second part of the clipped parts or so.

That song is hardly clipped at all.

I wouldn't have even bothered.

[Guest]

you can just select the whole song (ctrl-a) then go to effect, normalize, and make sure the -3 option is the only one selected

[moh.vze.com]

you can just select the whole song (ctrl-a) then go to effect, normalize, and make sure the -3 option is the only one selected

Holy crap dude! that worked a lot better!

I have 1 question though, does this make the song crappy in anyway?

[moh.vze.com]

Also, how come people are always asking if there is a "non clipped version" of "so and so song" when they can just load it up on audacity and do the normalize -0.3 trick?

Would a song still be considered "Clipped" If I reduce the max amplitude doing this trick?

[BigPimpin91]

Also, how come people are always asking if there is a "non clipped version" of "so and so song" when they can just load it up on audacity and do the normalize -0.3 trick?

Would a song still be considered "Clipped" If I reduce the max amplitude doing this trick?

Yes.

The Wave is still square at the top and bottom.

But Like I said, if its such a small amount, just ignore it.

[Decaf]

alright heres my contribution with plenty of zoom in and alterations that can be made to clipped music

t-rock - maryjane..... classic clipped song... viewed on windows media player osciloscope

song at ~20 seconds during a few clips

zoomed in to show how the wave has been cut off or clipped and isnt a continuous smooth wave

peak freq of 39hz @25db

this is what happens when you go to amplify and choose -3db on the sliding scale, see how all of the notes in the song went down 3db, but the clipping still exists, just with less amplitude, which translates to less power being put out by the amp with a weaker/lower signal db

after normalizing to -3db as suggested earlier in a post

and after applying a 78hz high pass filter (im assuming the slope is -12db so i choose the croosover point to be 1 octave up so at 39hz the signal has dropped at minumum -12db)... notice how the clipping is almost nonexistant....

lets zoom in and look at the peak again... notice it dropped 6db... so obviously the hpf is -6db

so lets apply it at 156hz to get the 12db slope... notice the drop from 25db from the unedited version to the now 12db peak... this results in again less power being put out by the amp since the signal at that note is even weaker....

the only problem with applying the high pass filter is that it affects all of those notes from 156hz down, which will result in a very muddly blurred depiction of the orignal data that sounds very empty and void of presence

this is why u have to be dedicated to finding the orignal copies of songs to assure u are getting what the artist intended u to get, (which 97% of the time is loud unclipped data)...

not some dj's mixtape that he added bass boost to so his hood figga followers get the boomy bass in their sealed off cadillac fifth-wheel-slanting trunks, and they wonder why their 12 10s in a sealed box arent loud and blow so frequenctly

[moh.vze.com]

Thanks for the info!

Hmm. This is kinda weird.

I am in a JR level Electrical & Computer Engineering: Signals and Systems class and I didn't understand the fundamentals of all of this

[moh.vze.com]

So I have a question

in the example when you said "the signal is getting weaker" at the 12db of that certain frequency...

What is a normal decible peak value? There are some songs that peak at 20db and some at 14db

what does this all mean?

[Decaf]

when refering to sub-bass... a good reference point to start with is 20db at each peak... anything over this will sound very loud...

jeezy, gucci, gorilla, unk, and others all have unclipped data approaching 25db+

dj billy and sine waves are 28db-30db at each frequency..... and are the highest peaking data available to play

the majority of downloaded music resides around 17db... commercially avail. cd's usually are ~21db... but thats why some artists albums are so much louder than others... artists like dj unk with the help of oomp camp productions find beats exceeding normal data ranges... he didnt mean beatin down yo block for nothin

with a wee bit of math introduced it explains it a bit more...

so when you double cone area or power... there could be a theoretical 3db gain... so if u have a song with 17db@35hz and a song with 21db@35hz... the second song could technically be delivering twice the amount of wattage than the first song... this is in a perfect world but u can get the idea

so lets say u used a 22db sine wave @50hz to set ur sub amp... when u play music with data exceeding 22db... u could possibly clip the signal, since there's more data, the amp would be set to sensitive to the stronger signal, or to high of gain setting

this is also why some people say to use a -3db sine wave when setting a sub amp since almost no music approaches the ~27-30db range, which most sine waves are

which means ur gain would be set a little too low to get the full potential out of music

[Eric B]

its tedious... is a little overkill.... but its really best to use the clipping finder on every song... then watch it on an osciloscope... and determine how severe the clip is and whether the song should be played or not

alot and i mean alot of gucci mane and souljaboy are clipped like crazy... be careful when dl from p2p

irs sad that big record companies put out albums with clipped shit.. i mean dont they pay that "sound tech" alot of money to clean everything up and have a good clean track? who honestly wants to buy a clipped to hell cd?