clipping question

cpc

Member
this is probably really dumb, so i put it in newbies.. anyways, i record into adobe audition 1.5 . i'll get the levels are right so when i record they're not clipping, but pretty loud and maxed out... then i'll add some compression and reverb with audition, which will raise the wave forms so that they are now over the clipping mark and hitting the red over 0 db area.... my question is, are they really clipping now even though i recorded them not clipping? like, will this make them sound worse? also i notice when i import otherwise good audio tracks, that they seem to be clipping according to audition (going into the red over the 0 db mark)
 
? i'm not sure what that meant.


what im saying is when i recorded i wasnt clipping, the volumes were fine, with some headroom. THEN i added virtual compression/reverb/effects within the software...which raised the wave of the already recorded audio into the clipping range... does this actually cause it to clip even though its AFTER it was already recorded?
 
Yes. You need more headroom. The individual tracks aren't "changing" or anything, but you're clipping somewhere in a group or mix buss. Just turn it down. You've got more headroom and a broader dynamic range to work with than at any time in history. Yet everyone seems to want to use it all up from the start... That's disturbing...

Think of it this way -

Your gear wants to run, and is designed to run, at 0dBVU.

0dBVU is "somewhere around" -14 to -18dBFS.

If you keep the meat of the signal around there from the start, you're *still* going to need to attenuate during mixdown assuming that there are a good number of tracks. Your gear will run better = your tracks will sound better. Your mixes will have more headroom (so your mixes will sound better), you won't be clipping (so your mixes will sound better) and your mixes will sound better.

This isn't analog - There's no "safety cusion" in there with tape and circuitry. Once that signal turns to all "1's" it's gone.

When tracking and mixing, my rule of thumb is to try to keep everything at or below 0dBVU - *Maybe* - MAYBE there might be a peak from some spectacular transient at -6 or -5dBFS. But for the most part, every track is riding around -12 to -14dBFS. Then during the mix, the same thing. Everything is attentuated until the entire mix is riding around -14dBFS, allowing for peaks. I'd feel comfortable turning it down even more. Much more. As long as the entire mix peaks somewhere above -20dBFS, it's fine.

In 24 bit digital audio, even that isn't a lot to ask.

Save the volume for the mastering phase. If you're going to use up all your headroom, use it up ONCE - Not at every step in the process.
 
sorry, im still lost...probably due to no fault of your own, you sound like you know what you're talking about, and thanks for trying to help... i'm just lost.

here's a rough example and picture of what i mean...

https://i21.photobucket.com/albums/b296/riznich/clip.jpg

the original audio is obviously not clipping...will adding the compression and reverb cause the audio to distort and clip.

there are various compression settings within my software like most i would imagine... some compression will raise the overall volume while compressing, while some will lower it. and i dont know which one im going to use until after i have the recording done, so how do i know which level to record at? if i were to lower the recording volume, and then use the compression that lowers the volume more...then when i raise the volume of the track i will have a whole bunch of unwanted hiss.

also like i said...when i import some perfectly fine NON clipped tracks into audition, they will go over the clipping mark within the program...when they are professional produced songs that arent clipped.
 
Just turn down the makeup gain in your compressor, probably labeled something like "output" or "gain". Or in many programs you can pull down the master fader.
 
The compressor is adding too much gain, that's all. Turn the make-up gain down.

And if you're getting hiss, there's a problem - Your recording levels aren't related to your noise floor - I'll explain...

If you record something with a signal to noise ratio of 50dB at -0dBFS, the noise will be at (you guessed it) -50dBFS.

If you record that *same signal* at -20dBFS, the noise will be at (you guessed it again) -70dBFS.

Add 20dB of gain to that, and you have your noise at -50dBFS -- Just where it would've been had you recorded it hotter.

That's a fairly extreme example, but the point is that adding volume will really only make hiss more audible if there's hiss there already. The level of the hiss in relation to the overall level won't change.

True - Once compressed and gain added, your signal to noise will come down with the dynamic range (dynamic range reduced, gain increased, difference between top and bottom is smaller). But still, only the hiss that was already there will increase in volume. If it's too audible, find out what's causing it and either gate it, fix it, or otherwise try to reduce it. Or compress it less. And of course, try to reduce it BEFORE you compress it.

[edit] Oh, hey MSH :D [/edit]
 
alright well i understand all of that...

but you're saying that even though its a software compressor, and its compressing an originally unclipped file... that it can still make it clip? that was my main question...and i guess it was answered as yes.. so thanks alot.

sometimes the preset reverbs i use will also add volume and put it over .. i guess i just have to practice recording at lower volumes so i can add volume/compression/reverb/and affects later within the software.. i was just always told to record fairly high and to take volume away, rather than to add.



also- i still dont get when i import some audio tracks from professionaly recorded/ produced cds that they are over the cliping area once imported in audition, and they obviously dont clip... maybe its just a glitch in the program?
 
yes, the answer is yes. i have a demo from one of my old bands that i can't even listen to anymore because i compressed the final mixes in audacity and they all clip (i didn't even notice at the time, sounded fine to me)
 
riznich said:
alright well i understand all of that...

but you're saying that even though its a software compressor, and its compressing an originally unclipped file... that it can still make it clip? that was my main question...and i guess it was answered as yes.. so thanks alot.

Maybe, but probably not. Signal processing in a DAW is usually done with floating point math, so there's a huge amount of headroom above 0dB through the effects chain, the software mixer, and any final output stages. If you don't clip while recording, odds are you won't get clipping anywhere until the audio leaves the processing chain.

However, there's a catch. You can still get clipping at the output stage, whether that's going to disk or to your audio interface. Thus, you must make sure that the final output level doesn't exceed the 0dB mark.

That said, it isn't particularly useful to have the individual levels exceed 0dB, so you probably shouldn't do so in general.
 
I have a related question....

I'm looking at my Home Recording For Musicicans For Dummies book (no hahas please). In the section for gain structuring, there is a lot of "How you set the levels is important" kinda stuff.

The prolem is, I get the impression that this section is geared towards an outboard recordist (maybe?). While there is gain staging material in the book, I am still dumbfounded.

Here's why. The way the book talks, I should be setting my preamp just below clipping and my computer at unity. Okay cool. But, shouldn't we be taking the later processing into account?

Here's an example...

mic preamp input gain (peaking at unity) to
mic preamp output gain (peaking at unity) to
recording interface (recording at unity) to
any gain adding effects.....

Or am I correct in assuming that we should not utilize the gainstaging of any faders or plugins to increase peak volume?... Is this proper conservation of headroom... Or is the headroom gone, as suggested earlier by M.E., as a result of recording at unity?

Another example...

WAV file of performance peaking at unity to
eq plug-in boosting ANYTHING to
a lowered fader to get it back down to peaking at unity...

Would I be correct in assuming that this signal chain is incorrect, due to the fact that the EQing will result in distortion before the fader ever has an effect on it?

Maybe I'm totally lost... I'm not asking for a seeing-eye-dog... Just looking for a "the door is over there" helper...

Thanks a million! :D

Edit: changed a biiiig mistake...
 
Last edited:
Okay..After some thought, I've decided to revise example 2...

WAV file of performance (0 dbfs) to
eq plug-in input level (-Y dbfs) to
eq boosting X hz @ +Y dbfs to
eq plug-in output (0 dbfs) to
a fader (0 dbfs)

Now this seems correct to me. Is this the answer?

Edit: Updating incorrect db references...
 
Last edited:
riznich said:
sometimes the preset reverbs i use will also add volume and put it over .. i guess i just have to practice recording at lower volumes so i can add volume/compression/reverb/and affects later within the software.. i was just always told to record fairly high and to take volume away, rather than to add.
Use reverb as a send effect, don't apply it to the file. That will take care of that problem.

The 'recording high and taking volume away' advise is very old. It doesn't pertain to 24 bit recording. It is also a bad idea because the sweet spot for most equipment is designed to be around line level which is about -12dbfs. Recording higher than that will put a strain on your preamps and the input of your converters.
 
Just to clarify the dbfs vs. db terminology (cause I should have learned this earlier):

(From the Wikipedia)
dB
The decibel (dB) is a measure of the ratio between two quantities, and is used in a wide variety of measurements in acoustics, physics and electronics. While originally only used for power and intensity ratios, it has come to be used more generally in engineering. The decibel is widely used as a measure of the loudness of sound. It is a "dimensionless unit" like percent. Decibels are useful because they allow even very large or small ratios to be represented with a conveniently small number. This is achieved by using a logarithm.

dBFS or dBfs
dB(full scale) — the amplitude of a signal (usually audio) compared to the maximum which a device can handle before clipping occurs. In digital systems, 0 dBFS would equal the highest level (number) the processor is capable of representing. (Measured values are negative, since they are less than the maximum.)
 
Farview said:
Use reverb as a send effect, don't apply it to the file. That will take care of that problem.

Wouldn't the end result be the same at mixdown?

Farview said:
The 'recording high and taking volume away' advise is very old. It doesn't pertain to 24 bit recording.

Why?

Farview said:
It is also a bad idea because the sweet spot for most equipment is designed to be around line level which is about -12dbfs.

Can you give me some idea of how to identify each piece of equipment's sweet spot?

Farview said:
Recording higher than that will put a strain on your preamps and the input of your converters.

I agree with the premise of preamps being strained to amplify a signal to unity gain... I don't understand why it would strain a converter to convert signals at unity gain..Can you help me understand?

Thanks!
 
I'll quote myself to answer most of that...

Your gear wants to run, and is designed to run, at 0dBVU.

0dBVU is "somewhere around" -14 to -18dBFS.

"Unity gain" does NOT translate to -0dBFS. Unity gain is "In = Out" - You amplify to a certain point. Arguably, that point should be around 0dBVU or -12,14,16,18dBFS (depending on how the gear is calibrated). THEN you worry about unity gain.

In the good 'ol days of tape, you wanted to record hot to tape and bring the levels down. That doesn't translate to digital. Digital audio came about for it's ability to record *quieter* without noise. Not louder. It has *more* headroom, not less - IF that headroom is utilized. Again, no "cushion" like tape.
 
Okay... I think it's coming into focus (Thanks!!!)

Please verify this....

mic preamp input meter is peaking at 0 dbVU >>>to
mic output meter is peaking at 0 dbVU >>>>to
recording interface (whose meter is prefader).....

Then the meter on my soundcard (M-Audio Firewire 410) application should read somewhere between -12 dbFS to -18 dbFS?

pannel2.gif


Moreover... Is the amplitude between my interface input meter's dbFS and 0 dbFS the amount of headroom available (in dbFS)?
 
riznich said:
also- i still dont get when i import some audio tracks from professionaly recorded/ produced cds that they are over the cliping area once imported in audition, and they obviously dont clip... maybe its just a glitch in the program?


Sorry to hijack your thread... Heres the some info on your question from this off-site thread...

"Yes; in fact, they are often mastered louder than the dynamic width of the media in which they are stored! Sounds foolish, I know... but this is done by deliberately clipping the peaks of the waveform in order to exceed the maximum decibel level of a standard Red Book audio CD. Some of the worst offenders have been mastered at Sterling Sound in New York (although not all of Sterling Sound's efforts exhibit this flaw...); for a quick look at an obvious example of such clipping, check out any of the Jimi Hendrix reissues by Experience Hendrix. If you open a WAV file from any of the Experience Hendrix/Dolly Dagger CDs in a visual wave editor, you will easily be able to see the clipping."

:)
 
well that was interesting to read thanks alot.

also..when i've imported other peoples home recorded tracks that they sent me online in mp3 or wma format this has happened as well, where the waves of the files will be clipping when imported to adobe audition (cooledit2)...and they say they werent clipping when they saved the file at their place.
 
riznich said:
well that was interesting to read thanks alot.

also..when i've imported other peoples home recorded tracks that they sent me online in mp3 or wma format this has happened as well, where the waves of the files will be clipping when imported to adobe audition (cooledit2)...and they say they werent clipping when they saved the file at their place.

Here's another site with some information...

"click on Normalize.... Set the options in the following dialog to match this screenshot (the Normalize Left/Right Equally option is disabled if you have a mono recording). I suggest setting to 98% rather than 100% to avoid possible clipping in the decoding phase of the final compressed file. "
-------
Which brings us to the question... How much clipping do you see? Seems like a lead to me...
 
Back
Top