DC Offset

NYMorningstar

Recording Modus Operandi
How much headroom is used up by this?

If you aren't pushing your masters to be extra loud does fixing the offset make any difference?
 
How much headroom is used up by this?

If you aren't pushing your masters to be extra loud does fixing the offset make any difference?

From my unerstanding of it it's caused by a variance in electrical signals between the recording gear and the sound card and can manifest as either increase in noiseflooor at inaudible ranges (ie you've suddenly got a 30 dB increase in the noise floor but its at 0 hZ so you can't hear anything) or you get pops and clicks at file start and stop or at edit points.

I'd say if you can't detect it where you cut a sound file and you're not clipping your meters with sound that is impossible to hear it's not a problem but if you want to see for yourself, most DAWs and editors have a 1 click tool to remove DC offset or you could try a HPF on the offending tracks to remove the crap below 20hz and see if there is any difference

If you have something like sound forge there is some useful stuf in there for detecting if you really have DC offset or if it's just that the waveform is not symetrical, but even if you do find it's there. if the track sounds goood, it is good and probably not that worth worrying about
 
How much headroom is used up by this?
That depends upon the amount of offset you have. Assuming your referring to the digital realm, think of it like this: Let's say You have an offset measured to a level of - 55 dBFS. That gives you only 55dB of effective range. If you're offset measures to -90dBFS, however, you have 90dB of effective range to work with.
If you aren't pushing your masters to be extra loud does fixing the offset make any difference?
It can. It's really best to play it on the safe side and eliminate any offset as part of the mastering phase. It's not just a matter of DAW headroom, but the effect the offset may or may not have on your loudspeakers. Offset voltage will push the bass speaker cone out (or pull it in in the rare case of negative offset) and change it's rest position the same way it changes the rest position of the waveform. This means that it can effect the performance of the speaker somewhat.

In small cases of offset this may not be much of an issue, but considering you have no idea what kind of system your music will be played on (car stereo, earbuds, etc.), you have no idea just how resilient any given playback speaker will be in that regard. Since offset is so damn easy to get rid of, there's really no reason not to check for and filter it out on any song that comes along.

G.
 
Glen - would you always do this in the mastering phase, or does it make sense to fix DC offset on your raw tracks while mixing?
 
Thanks Bristol. I use wavelab so it's a pretty simple procedure to detect and fix.

It's really best to play it on the safe side and eliminate any offset as part of the mastering phase. It's not just a matter of DAW headroom, but the effect the offset may or may not have on your loudspeakers. ...but considering you have no idea what kind of system your music will be played on (car stereo, earbuds, etc.), you have no idea just how resilient any given playback speaker will be in that regard. G.
That's what I usually do but I was questioning it to myself last night because I can never hear a difference. Your point about speaker performance makes sense though so I'll keep on keepin on, thanks.
 
Glen - would you always do this in the mastering phase, or does it make sense to fix DC offset on your raw tracks while mixing?
I haven't seen a DC offset problem in 15 years.** If you have one, the time to take care of it is long before the "RECORD" button is ever pressed.



** Okay, on a few $4 sound cards and what not, maybe.
 
Glen - would you always do this in the mastering phase, or does it make sense to fix DC offset on your raw tracks while mixing?
It's best to nip it in the bud, of course. But to be honest, I usually don't think to check for offset on every single track I make or get. If it's pretty significant offset, it's pretty obvious visually in the waveform and I'll get rid of it right away. But if it's a very low level offset, it's not obvious to see or hear - and frankly most likely isn't going to affect the mixing, so if it slides through the mixing stage, it's probably not a big deal (though it's still best to be rid of it.)

But after summing the tracks to a mixdown, the offset can be additive and get much worse. So between that possibility and the ease of checking it on a single file, there's no excuse not to make that an integral part of (pre)mastering, IMHO.
I haven't seen a DC offset problem in 15 years.** If you have one, the time to take care of it is long before the "RECORD" button is ever pressed.
Well, John, I for one would never send you anything that had offset in it; that would just be rude ;). Kind of like sending your kid to the photographer without washing and combing his hair first. I'm not surprised you don't see a lot of it coming your way.

But you are right, it doesn't show up a whole lot these days anyway. Mostly on laptop soundcards for some reason (IME, anyway). But it does occasionally creep in unexpectedly. A lot of times these days when I see it it's not really true offset, but rather a VLF modulation from somewhere. Same solution either way.

And unfortunately getting rid of it before hitting the Big Red Button is not an option for me when I receive tracks for mixing that are from someone else that ain't hip to the DC situation.

G.
 
I have a habit of checking DC offset on every track. Sometimes for some odd reason, even when completely ITB, some process or processor may misbehave (I've seen this enough times to check for it as a habit).

As a side note, I am working on a little tutorial/sound-design article on creative uses of intentionally induced DC offset (very easy to do in the digital realm). Although this will probably be more useful to synth designers, and owners of modular environments (Reaktor, Max/MSP, Kyma, etc.).
 
I have a habit of checking DC offset on every track. Sometimes for some odd reason, even when completely ITB, some process or processor may misbehave (I've seen this enough times to check for it as a habit).
You deal a lot with soft synths, which IME are sometimes awful at creating reliable or even close to textbook or close to the original waveforms. They may sound fine, but the actual waveforms they generate (in many cases but not all) are sometimes quite off or just plain wrong. There have been cases where this includes inversion or other offest of the waveform. I saw one once that generated a sawtooth wave that was not only DC offset, but was backwards also.

G.
 
Back
Top