Phase question

[Gelatin]

I have a stupid noob question about phasing. If I'm understanding correctly phasing problems occur when two waveforms cancel each other out (second waveform looking like the horizontal mirror image first one), and using the phase invert button on a mixer/DAW just inverts the waveform (so they look more like identical waveforms).

Ok I'll try to explain. How about if the waveform is not completely out of phase, but like in between - so that if you'd invert the phase you would be just as far off.

I guess my question is, how far does the second waveform have to be out of phase so that it's audible, and are there any techniques as to get rid of phasing? Isn't it just possible to delay either signal on a DAW?

thanks in advance!

 

[bennychico11]

I have a stupid noob question about phasing. If I'm understanding correctly phasing problems occur when two waveforms cancel each other out (second waveform looking like the horizontal mirror image first one), and using the phase invert button on a mixer/DAW just inverts the waveform (so they look more like identical waveforms).

What you mentioned is the result from two signal being 180 degrees out of phase. That is, taking the exact same signal (whether it be a sine wave or a recording of a singer), duplicating it and inverting the polarity of the second track. So now the peaks of the first track are being canceled out by the dips on the second track.

This is very very rare in real life. You hardly ever hear a complete cancellation. When we talk about phase problems it typically has to do with certain frequencies canceling at certain points in our wave form. This creates a weird swirly effect that makes you think, "hmmm, that doesn't sound quite right."

You already kind of understood it when you said this:

Ok I'll try to explain. How about if the waveform is not completely out of phase, but like in between - so that if you'd invert the phase you would be just as far off.

This is extremely common when using two microphones on a source. If they are placed at different distances from the source the one further back has a little bit of delay than the one closer to the source. Sometimes this might sound pleasing to your ear. Other times it might be taking too much away from the sound. We use the polarity reverse button to give us another option to choose from. Sometimes we like it, sometimes we don't. Typically if you've mic'd a snare from the top and bottom, you'll reverse one of the mics. That's because their facing the complete opposite way and therefore recording the wave form in a different way from each other.

I guess my question is, how far does the second waveform have to be out of phase so that it's audible, and are there any techniques as to get rid of phasing? Isn't it just possible to delay either signal on a DAW?

It depends on the sound and placement of mics. You just have to use your ear. To minimized phasing though you'll want to use a combination of close micing techniques utilizing different the polar patterns of the mics and/or the placement of the actual microphones. And yes, since phasing is just a result from a time delay of our audio signal from one microphone to the other, you can physically move the audio signals to line them up. I prefer to move the later one to compensate. However, not everyone uses a DAW and has the ability to visually see the wave forms.

Do a search around here. I've talked about this in previous posts and even give some visual/audio examples

 

[SouthSIDE Glen]

A polarity inversion or a 180° phase change are the most egregious in that they do - assuming no changes in overall amplitude - fully cancel the original wave, but phase is a factor always, at every one of the 360° of phase angle there is. Even a 1° phase change will cause noticeable changes in the shape and amplitude of the waveform.

The only time we care about it is when the results are both audible, and when those results sound bad to us (not all of them sound bad). To my knowledge there is no set formula or even rule of thumb for determining this. A big part of the problem is that phase is frequency Dependant; what is in-phase at 100Hz is slightly out of phase at 101Hz. Unless you are playing only synth sounds with pure or mathematically divisible tone frequencies, no two sound pickups are ever going to be completely in phase across the entire audible spectrum.

The best general rule of thumb is simply to try and keep the amplitude of any unintentional or unneeded bleed into other microphones to a minimum of amplitude so that any potentially harmful phase issues do the least harm.

G.

 

[IronFlippy]

If I have two mikes on a subject facing opposite each other, I'll invert polarity on one of them, see if it sounds better, then switch it back if i doesn't. It's a common misconception that you have to flip polarity on one of the mikes if they're facing each other. Only do it if it sounds good.

To more directly answer your question: it is impossible to eliminate phase issues when recording, even with one mike on one source, unless you're in a perfectly anechoic chamber.

 

[antichef]

And yes, since phasing is just a result from a time delay of our audio signal from one microphone to the other, you can physically move the audio signals to line them up. I prefer to move the later one to compensate. However, not everyone uses a DAW and has the ability to visually see the wave forms.

Thanks for this thread and that answer! I had been sub-verbally/subconsciously about to post this question for some time now.

So for those of us with DAWs that allow for nudging like this, wouldn't the whole phase thing be something of a bugaboo? (talking of course, about the phase issues with two or more mics, not the phase / comb-filter issues caused by unwanted sound reflection, which is still a problem). So far, I've been pretty careful with mic techniques and have never noticed a problem -- but all the time I wonder whether I'm being more careful than I need to be, due to the DAW thing.

 

[bennychico11]

phase is just one of those things you want to keep in the back of your mind. You should know what it is and what it sounds like so when you come across it you aren't freaked out as to why things are sounding weird. No need to stress on it too much though.
DAW's let us visually see where phase problems are. I get them all the time when working with video production. The audio guys on location like to use a lav mic and a shotgun at the same time. Which more times than not ALWAYS come back and has phase problems when I sum them together. But all I have to do is zoom in and double check if it's for sure phasing, then find a common peak and nudge the shotgun back in time so that they line up.

 

[SouthSIDE Glen]

Where a lot of confustion comes in here is that once it gets to the DAW, some phase issues are already locked in and cannot be corrected by a simple time shift.

Benny is absolutely correct in his description and use of the time slide to line up phase issues *caused by simple delay* in time between the two waveforms. This is how many such issues arise, and how many such issues are solved.

But it's important to remember that some phase issues are still going to be built into the waveforms, and no amount of sliding back and forth can help. Even if one compensates for delay, there will be instances where the amplitude at a given frequency will be different between the two waveforms because of the distance/wavelength relationship itself. If, for example, the two mics are placed so that one is a half-wavelength further from the source than the other at frequency x, even when one time-aligns the two recorded tracks, there is still going to be some phase cancellation at that frequency (and at mathematically related frequencies). That's just a simple theorteical example; in reality the phase relationships at many frequencies can lead to complicated distortions because if the mic placement.

The thing to remember about that is that once those waveforms are recorded and in the DAW, everything in that regard is set in stone and no ambout of time aligning is going to help. True time alignment is simply going to reproduce the phase problem. Trying to align the waves so that a problematic frequency no longer gives you phase headaches will only throw everything else out of time sync.

This is not a huge problem to lose sleep over, but it is important to understand that the two waveforms recorded in two mics are not identical and simply shifted in time, they do have variances in phase built-in based upon frequency as well. This means that some phase issues can be encoded right into the waveforms and cannot be resolved by sliding back and forth in a DAW; sliding back and forth of the microphones to get the best phase relationships across the board is the ideal solution there.

Again, it's usually not something to lose sleep over; proper mic technique and mixxing technique from the start will usually keep these kinds of problems to a minimum. But when it comes time where such issues do rear their ugly heads, it is important to understand the whole situation so you can troubleshoot it properly.

G.

 

[antichef]

sounds good -- (and, Gelatin, no hijack intended and hopefully none perceived - apologies if otherwise). Thanks!! Based on this, I'll experiment more when I can, but stick to what I know works (the prescribed techniques, generally) when it counts.