6

Let's say I have a sound I like. I start mixing down, and I see that it is all over the place on the phase-scope; that is, the signal looks like it has some significant out-of-phase components. I've been told this is a Bad Thing, especially if you want to get your track pressed on vinyl.

Phase scope is all over the place, but it sounds pretty nice.

OK, so I squeeze the channel's stereo signal until it's all in-phase. Now it sounds flat and dead. But at least now the phase analyzer looks sane, according to what I've been taught.

There we go, all fixed, but now it sounds dead.

My ears prefer the rich, wide sound I get when the phase scope shows 'too much' anti-phase (maybe I need a second opinion). How much anti-phase content is really bad for a track, and how do I figure out where the threshold is, especially if the wider version sounds better to me?

Improve this question
1
  • 1
    Usually you don't need to narrow down the whole mix, find out the instruments that give that much trouble and narrow them down. Usually bass with too many effects on it is the culprit. Make the low end (below 300 hz) mono.
    – Eugene Bujak
    Commented Mar 1, 2011 at 10:49

4 Answers 4

Reset to default
4

Yes, trust you ears but refer to these tools as much as possible. If in doubt, try doing some testing yourself: take to a car, headphones, poor quality "media desktop speakers" mashed up in the corner of the desk and see how they translate.

And more important of all: mono compatibility! BBC still won't allow stuff that is not mono compatible to be played (guess they are still afraid of those folks with a pre 1900's tv or radio set which is mono AND are paying for the TV/Radio license :P)

Don't forget if your track is supposed to play in lounges, clubs, etc chances are they'll be either downmixed to mono and matrixed or just plain separated throughout the venue. You REALLY don't want your side guitars to disappear or worse - your vocals!

To be honest I don't use PAZ much I tend to use the DK audio hardware designed for G+ consoles BUT that second one of yours seems WAY to narrow and your first one tilted to the right, with a lot of -phased material.

This is what I mean about "don't use PAZ" (as in, I'm not used to interpret the data as shown) and use your ears/do some testing: "It is a bit tricky to define when a signal is shown by the Stereo Position Display or heard by the ears as “out-of-phase”. This is not simply when the left and right channels happen to have opposite polarity at any given instant. Such a situation can happen if independent instruments are panned to different stereo positions, and it just happens, statistically, that the signal of each channel has opposite polarities. The meter will show a quick spike in the antiphase area. This can also happen if a stereo delay or reverb is applied to an instrument and the delays arrive to each channel at different time" (from WAVES PAZ manual) very true but such a weird language used

Improve this answer
4
  • Good points. My narrowing of the sound would then probably represent the worst-case downmix. So it might lose a little depth in the club, but the element won't disappear altogether. I'll hopefully get to field-test soon.
    – daniel
    Commented Feb 2, 2011 at 22:24
  • I normally start a mix with my monitors set to mono, then pan and listen and go back and forth, checking mono every once in a while.
    – Sam Greene
    Commented Feb 16, 2011 at 22:36
  • 1
    Sam, it's not uncommon. I spent a couple of weeks with a award mixing engineer and he would do the pan in mono as well (specially things like toms/OH etc). It was noticeable the mono mix becoming clear. But he still re tweaked the image when back in stereo. I personally just check in mono. A lot!
    – jlebre
    Commented Feb 18, 2011 at 6:42
  • Uh, on this subject, I found this pearl fluxhome.com they have 2 freeware plugin their stereo tool is amazing and I can say the same about the transient designer - specially the flexibility of it!
    – jlebre
    Commented Feb 18, 2011 at 6:43
3

I generally say trust your ears unless you're introducing some kind of specific technical hurdle.

So unless you know that you're going to run into technical issues, say, pressing vinyl, why not allow the out-of-phase components? Especially if you think it sounds better with them? And even if that is the case, maybe vinyl isn't the best format for representing the sound you like.

Improve this answer
4
  • 1
    I agree completely. In this business, the ears have it, not the ayes.
    – BenV
    Commented Feb 2, 2011 at 18:08
  • 2
    Another vote to "always trust your ears". When you don't, that's when you turn to the tools.
    – Ian C.
    Commented Feb 2, 2011 at 19:26
  • Another vote to use your ears. I might add that something that is considered technically incorrect my the majority might be that "sound" that you are looking for. This is esp. true for speciallity applications. For example, if you are not mixing for vinyl but for a stage or a computer game.
    – d-_-b
    Commented Feb 3, 2011 at 6:01
  • Also, if a track's being mastered for vinyl -- and it's always going to be (re)mastered for vinyl, usually from the original bounce -- the requirements, and the treatment it's given, varies drastically from a 'standard' CD- or digital-oriented master. Multiband processing and selective narrowing of stereo image across frequency bands is easily achievable so out of phase components can be nipped in the bud before a track ever makes it to the lathe.
    – Chris Woods
    Commented Dec 5, 2011 at 16:37
1

This is relevant.

Article on phase problem when recording. I know that the situation is completely different from what you've described, but the article talks about what is phase canceling, comb filtering and etc.

Improve this answer
1
  • btw: this topic should be a comunity wiki - phase canceling is a common problem and there's still loads of information that may be added, specially due to the fact that removing audio from phase can produce interesting results - this song works on the fact that delays may become filters depending on the latency of the effect.
    – Johnny Bigoode
    Commented Dec 23, 2011 at 18:42
0

While all these answers are correct and equally valid, I thought I'd add some real-world examples to help with the bigger picture.

Through curiosity I analysed a few random songs I found on my hard drive with PAZ and the results were(for me) quite interesting, so I thought the community would appreciate them too.

These are the results(6 second peak holds):

Boston - "More Than a Feeling" (Guitar Track)

Boston - "More Than a Feeling" (Guitar Track) 6 sec PAZ analysis

Creedence Clearwater Revival - "Green River"

Creedence Clearwater Revival - "Green River" 6 sec PAZ analysis

Hank Williams - "Hey Good Lookin'"

Hank Williams - "Hey Good Lookin'" 6 sec PAZ analysis

Lynyrd Skynyrd - "Freebird"

Lynyrd Skynyrd - "Freebird" 6 sec PAZ analysis

Percy Sledge - "When a Man Loves a Woman"

Percy Sledge - "When a Man Loves a Woman" 6 sec PAZ analysis

Scott Brown - "Rock U Softly"

Scott Brown - "Rock U Softly" 6 sec PAZ analysis

Underworld - "Born Slippy"

Underworld - "Born Slippy" 6 sec PAZ analysis

Conclusion

So all these examples have an amount of anti-phase, but the amount and the characteristics vary greatly. Although they seem prominent in most cases, it seems most of the anti-phase peak holds are due to the collecting of transients, rather than sustained events. I think this just reinforces answers like Warrior Bob's. The OP's PAZ analysis would fit into this group nicely, indicating that it is typical to allow this amount of anti-phase, unless there is a specific reason not to.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.