My peaks going into the red aren't distorting

B

Brackish

New member
In the past I've used compression to
keep the peaks out of the red but lately
I've tried cutting back on the compression
such as has been recommended by
engineers such as Bruce Swedien.

Overall, the sound is much better than
with the compression but I'm getting
peaking into the red (i.e. going over
0dbFS) but there is no distortion that
I can hear.

My question is ... is the main problem with
having peaks into the red that the
digital-to-analog convertors will have
a problem with it? Can you go into the
red just a bit when working with digital
and be okay, so long as you keep it
in the digital domain?
 
Do you mean on your individual tracks or on your stereo master???

Not that it really matters because the short answer is: Why would you want to go into the red if you can simply turn it down and NOT go into the red???

(Ok, that's a question, but whatever)
 
Is this while mixing? With current floating point software, you can turn up a track all you want, and it won't distort until you clip your converters on output, which you won't do if you bring down the master fader. On mixdown, however, all those clipped samples will be brought out in all their white noisey glory.
 
Ford Van, most meters tha max at 0dB FS also have a clip indicator.

Brackish, as said earlier, you won't clip while mixing due to floating point operations. Most likely 32-bit float. Your converters will cause clipping eventually.
On mixdown the waves that clip will have the file data truncated and will be distorted with click-pop high order harmonics in your mix.
 
Usually, peaking in your software meters doesn't really matter. I don't usually pay too much attention to the software meters. But you should know, if you hear any digital distortion.
 
*Where* are your meters peaking? On individual tracks, mix, in software, in hardware?

Sometimes the clip light is a warning that peaking will happen soon, not that it is already occurring. How is your system calibrated, or have you calibrated it to make sure everything is as it says it is? There are so many factors that play into your question, that it is hard to give a simple answer.
 
Except for very percussive sounds, your levels shouldn't be anywhere close to 0dbfs. Just turn it down if you are peaking that high, the analog chain leading up to the converters will thank you.
 
Since you're using software to mix, pre-master, etc. DO NOT GO BEYOND ZERO!!! I dont care if you hear the clipping or not, dont do it. Until you start mixing on an analog board, dont go over ZERO. The meters are there for a reason. Software mixing causes more artifacts when you do anything extreme. I learned the hard way.

Now... if this post helped you out, please be sure to listen to my music @ http://www.soundclick.com/raydio
 
Ford Van said:
So, if digital meters only show you values "up to" 0dbfs, how do you know you actually went over that?
Click to expand...
'Display numeric peal values' in Sonar, you get " + n.n ".

The peaks may just be too brief to easily notice. Don't we want to always leave a least a few tenths below zero on the final?
Wayne
 
In many cases going into the red does not hurt on individual tracks or even on the masters, if you have a clipping indicator at the end of the meters, you do need to watch for that depending on the headroom available, sampling rate and what not.

At any rate, when all else fails....trust the ears! :)
 
tarnationsauce2 said:
Ford Van, most meters tha max at 0dB FS also have a clip indicator.
Click to expand...

They do? Show me some examples. And do you REALLY put your trust in a "clip indicator" in mid level digital gear?

The fast is, you DON'T know if you exceed 0dbfs on digital meters. the manufactures just make the "clip" or "peak" light come on when you reach 0dbfs, but you have absolutely NO WAY of knowing if and by how much you exceeded 0dbfs.

It has been debated as to how many consecutive samples in a row that "peak" or "clip" needs to happen for the human ear to actually HEAR distortion. The debate is 8 vs 12 consecutive samples. I never followed up to see who one! :)

Anyway, you can actually "clip" your A/D's and still not HEAR digital distortion IF the clips were less than around 10 consecutive samples in a row.
 
Ford Van said:
Anyway, you can actually "clip" your A/D's and still not HEAR digital distortion
Click to expand...

Is that like 'If a tree falls in the forest....'?

If you can't hear it, does it matter? Or will it only matter if you clip 16 tracks, unable to hear the clipping when isolated, but together, it's the nails on the chalkboard thing?
 
:p
Robert D said:
And stop tracking so HOT ;)
Click to expand...
;)


Square Waves are popular these days, Robert D :D

What can you do!!! , The younguns think you need to get that beneficial warm , slightly compressed sound that comes from saturation of the HARDDRIVE!!!
 
ez_willis said:
Is that like 'If a tree falls in the forest....'?

If you can't hear it, does it matter? Or will it only matter if you clip 16 tracks, unable to hear the clipping when isolated, but together, it's the nails on the chalkboard thing?
Click to expand...

As with all things audio, if you can't hear it, don't worry about it. ;)
 
ez_willis said:
Why ?
Click to expand...
Because there's no reason to. Why rob yourself of headroom and tempt the fate of the clipping gods when everything will sound just as good, and usually better, if you leave yourself a few dBs?

And let's not even start with the tired old "louder sounds better" or "louder is more competitive" myths. Time and again it has been demonstrated - by a virtually unanimous concensus of the most experienced and knowledgable engineers in the business - that those myths are just a bunch of baloney, and the only reason they do it is because those that don't know any better are the ones signing their checks.

And even if they were true, a good engineer can acheive those "louder" goals without having to resort to clipping.

So to me the question is not "why should we not clip?", but rather "why should we clip at all?" There's no good reason for it, and plenty of bad ones.

G.
 
You must log in or register to reply here.
Back
Top